View Full Version : Megasquirt 4A-GZE

03-16-2010, 10:27 PM
The purpose of this thread is to chronicle my adventures in getting my MR2 to run with a Megasquirt-II engine management system. Obviously this is specific to my car. Iím also posting here in the hopes that if I have some information that is wrong someone will let me know before I do any real damage. Updates here will be slow! I will be taking my time (what little of it I have) and I have to wait until I can park the car for a few weeks before trying to get the MS installed.

So as some of you know I have a MK1 MR2 with a 4A-GZE swap. The engine that was swapped in was a North American model (AFM, no under piston oil squirters, distributer). I also have a spare 4A-GZE that came from Japan that has been sitting for a few years. Initially I was just going to swap the engine in to get a small increase in performance but I ended up holding back on that. Situations in my personal life took priority. However, one day while trying to figure out why my car wasnít performing as I thought it should I took off the supercharger and found that the rotors were all scratched up. I swapped the supercharger from my spare engine onto the car and things were happy again.

Obviously, like any other moving part, the supercharger wears out. Iím sure Iíve still got plenty of time on the one that is in the car now. Toyota will sell me a new one for $4500, and I could buy at least three more complete engines for that price. There are seals available for the outer housing, but no one is supplying the bearings or the rotors. So I have decided that the best route to go is to replace the supercharger with a turbocharger. Iíll be honest and say that this is still a ways off as I simply do not know enough about turbocharging theory to make an informed decision as to what turbo setup to use.

It was along this line of thought that I came to the conclusion that I needed and aftermarket ECU to run the thing. I started looking at the commercial units and seeing what they offered. Like many before I had looked at the Megasquirt but had decided that it didnít do everything I wanted it to. I did look at it to see if there were any updates and found that the thing had make huge leaps since I last looked at it (for reference, when I had last looked at it it only did fuel, no ignition). I started to research it and spent a couple weeks going through all the information of how it works and what it will do.

Finally I came up with a plan. I would get the MS to work on the engine that is in the car now and then rebuild the other engine with a turbo in mind and swap the MS to that engine. I went to DIYAutoTune and placed an order. I ordered the MegaSquirt-II PCB3 Unassembled Kit, the JimStim 1.4 unassembled kit, the 4 bar MAP sensor with Barometric Correction and the Zeal Engineering Daughterboard.

I still have some more to order (like the wideband O2 system), but I wanted to be a little kinder on my wallet and I donít need everything right now. I am planning for this to take some time to install in the car and get it running so I have to wait till summer when I can ride my bike to work so the car can stay parked for a week or two.

I am confident in putting this unit together has I have had many years experience with building circuits in industry. The first thing to do is research. Then build the stimulator, then the MS unit, then get it all working on the bench. Finally install it in the car and start tuning. For me I think the tuning will be the hardest part.

I will try to break the systems down as much as I can. If things seem to be oversimplified it is simply because I want to make sure I understand it.

Fuel delivery

Fuel management is actually fairly simple. Basically the ECU inputs data from sensors on/in the engine and then determines via lookup tables how much fuel should be injected into the engine. Tuning is done by changing the values in those tables. The MS looks at several engine conditions. There is the Throttle Position Sensor (TPS), the Coolant Temperature (CLT), the Intake Air Temperature (IAC), RPM, and an Exhaust Gas Oxygen Sensor (EGO).

On top of the lookup tables there are other variables taken into consideration. For example an engine that hasnít reached optimal operating temperature can have a warmup enrichment. When the TPS is open far enough there can be an acceleration enrichment. The EGO data is used to correct the amount of fuel injected (to a certain extent), and so on.

The MS pulses the injectors on and off to control the amount of fuel injected. The width of the on pulse (how long it is activated or on or open) determines how much fuel gets in per pulse. Changing this pulse is known as Pulse Width Modulation (PWM). An injector that spends half of a cycle open and half closed is said to have a 50% duty cycle. The higher the duty cycle the more fuel is injected. If an injector is at 100% duty cycle then it is open all the time. If this condition occurs and the engine still doesnít have enough fuel, you need bigger injectors.

The version of the MS that I am getting has the flyback circuitry built in so it will run both high and low impedance injectors. I am getting the MS-II which can only batch fire two banks of injectors. The two banks can be fired together or in an alternating pattern. However, sequential fuel injection is not available. The MS-III will have it but it is still in the testing phase.

The firmware that I will load onto my MS will be the MS2extra. This gives the MS a lot of capabilities. It contains code for wall wetting (fuel that doesnít get sucked into the engine but stays on the walls of the intake) and such. I havenít had a chance to play with it much yet as the Tuner Studio software doesnít do much unless you have a MS connected.


I have four options when it comes to ignition triggering:

1) I have a distributor from a 4A-GE that has the standard 4+1 sensor VAST setup. What this means is that there are two wheels in the distributor, one has four teeth and one has one tooth. The four toothed wheel triggers a sensor (signal labelled Ne) at every half crankshaft rotation, which is an ignition event. When cranking or the jumper is inserted into the test connector the igniter fires the coil each time that the sensor detects a tooth. The spark then travels from the coil to the distributor where it is routed to the correct plug by the distributor. After cranking and the engine is running the ECU then adjusts the timing by taking over the triggering of the ignitor based on various engine sensor feedbacks. The other wheel has one tooth on it and as far as I know the ECU just uses this for injector and ignition timing adjustments (this signal is labelled G). According to everything Iíve read the 4A-GE is the only engine that has the 4+1 with the VAST ignition. All the others just have the single four toothed wheel. I donít know if this means that the 4A-GE has sequential fuel injection or not. Using this would be the simplest method of getting the ignition working. I could use the distributor unmodified and send the Ne signal to the Megasquirt. The Megasquirt would use this to determine engine RPM and I would have it trigger the coil directly. Although this is the simplest, I would rather run without a distributor and use a wasted spark, coil on plug, or coil near plug ignition.

2) The easiest and most common way to run a wasted spark with a Megasquirt is to use a crank trigger wheel like one from the Ford EDIS. Although I like the idea of not using a distributor I donít really want to mount a toothed wheel on the crank pulley and a sensor nearby. And whatís the point of having a Toyota if youíre going to put Ford parts on it? I have also read that the Ford EDIS does have its limitations and the Megasquirt doesnít have full control over it. This is somewhat of a moot point now that Megasquirt has evolved a bit in that all that is needed from the Ford system now is the crank wheel, the sensor, and the coils.

3) I also have the distributor that is on the car now which is a Toyota ESA ignition system. It is a more advance system in terms of using the sensors to run a single coil distributor system. Again it has two wheels, a twenty four toothed wheel and a four toothed wheel. The twenty four toothed wheel provides the Ne signal. This is an improvement in that the ECU is sampling the engine RPM more often and as such able to make corrections to fuel injection and ignition more often (particularily the ignition timing in acceleration conditions). The four toothed wheel now provides the G signal. However on my unit the four toothed wheel has two G sensors, G1 and G2. This setup would work great with the Megasquirt in that the twenty four tooth would provide the RPM signal and the four tooth would provide the ignition triggers. With the Megasquirt only one of the G sensors need to be wired up. Both can be wired together, but this is just inviting noise into the system. Now if I wanted to run a wasted spark system and/or coil on plug I would have to modify this distributor a bit. There are two ways I can do this, both involve grinding teeth off the four tooth wheel. I could remove two or three of the teeth. If I remove two I would have to remove two that are opposite of each other and I have to make sure that one of the remaining teeth lined up properly with the sensor to indicate TDC of cylinder one. Because the camshaft rotates at half the crankshaft speed the sensor would indicate a signal once for every crankshaft revolution. If I were to grind off three teeth then the Megasquirt would decode a single camshaft revolution. This might work better for coil on plug as a G signal would only be sent when cylinder 1 is at TDC. Alternatively I could connect the two G sensors in parallel and I would get a signal similar to the removing two teeth method.

4) My last option, and likely the one I will use is to use the cam position sensor that came with the AE101 4A-GZE I have. This works very similarily to the ESA distributor in that it has two wheels, one with twenty four teeth. The other wheel however only has one tooth. I have not yet pulled mine apart but base on pictures Iíve seen it isnít just a small tooth, but rather a tooth that covers 180 degrees of the shaft. Because the VR sensors only trigger on the edges of the tooth then this isnít a problem. This unit also has two G sensors in it. The instructions on the Megasquirt Extra page state that the second sensor should be connected for wasted spark and not connected for coil on plug. However digging through the forums the concensus seems to be that the second sensor need not be connected at all. Considering that the Megasquirt will be counting the teeth on both wheels it should know how many teeth on the twenty four tooth wheel it should count before the next ignition event.

Now option 4 is the favourite so far. Not only does it require zero physical modification it also comes in a neat little package. There are no connectors for spark plug wires and it will gain a little bit of clearance in the engine by (MR2s need all they can get). Instead of the large plastic distributor cover it has a slim aluminum cover. The only disadvantage I can think of would be replacing parts when they wear out. Because this engine was never offered as an option here in Canada parts might be a little difficult to source. Assuming that it uses the same bearing and shaft as the distributor (it bolts right up so why wouldnít it) then getting parts to stop future oil leaks shouldnít be a problem. I donít know how similar the sensors themselves are to those in the distributors, but they might be interchangeable. Also the 7M-GTE engines used a wasted spark ignition with a cam angle sensor. The housings are different but there is a chance that the inner components might be similar enough to be swapped.

I should mention that the Megasquirt comes with the ability to accept one VR sensor input. I have ordered the Zeal Engineering Daughterboard which will give me two more VR sensor inputs (though I will only use one) as well as a 12V tachometer output and a few more on/off outputs. The 12V tachometer output can be modified to run the stock tach in the car. I was planning on using one of the on board outputs as a tach signal for the rev counter in the car. There is a relay coil modification listed on the Megasquirt site that describes how to do this (see below for diagram). Alternatively this signal could be used to drive an aftermarket tach. However the only way I would put an aftermarket tach in my car is if I mounted it in the dash where the stock one is. That would be more work than I think its worth. Itís a matter of personal taste, but I do not like the look of aftermarket tachs mounted on the dash or hanging from the A-pillar. Mind you this only applies if the vehicle already has a stock useful tach. If the car didnít have a tach or the engine has been swapped/modded to run drastically higher revs then I donít mind. Although in that case the old tach wouldnít be of much use so I would remove it and try to put the new tach in its place. Still I would prefer to find a model of the car at the wreckers that has an instrument cluster with a tach option and swap that in. Again itís just personal taste. Do what you want with your own car.

Iíve already mentioned a few options of driving the spark plugs. I will expand on that a bit now.

To understand how the spark plugs are fired a few basic electronic principles are needed to be understood. First there is the sensor which either tells the ECU when to fire the coil or, on older cars, fires the coil itself. The sensors in all my distributors/cam position detector are all Variable Reluctance (VR) sensors. It uses magnetic properties to detect the teeth. These only work when the toothed wheel is moving. The next step up is the Hall sensor. These are similar but they actively supply information even when there is no motion. Finally there are optical sensors. These act like the laser beams in movies where things are stolen from highly secure locations. When the beam is broken the alarm goes off. However in this case when the beam is broken a signal is sent to the ECU. A wheel with holes in it is rotated with a light source shining through the holes. On the other side a light sensor picks up the light. When the beam is broken it is like a tooth passing by a VR or Hall sensor.

The coil is simple, it is a transformer. A simple transformer works with two coils (or windings) of wire. If the coils in a transformer are the same (have the same number of windings) and an AC voltage is applied to one coil (the primary coil), the same AC voltage can be measured on the other coil (the secondary coil). (This is often used as a DC filter and as an isolation technique in signal circuits). What is happening is the primary coil is creating a magnetic field with alternating current. This field induces a current in the secondary coil. Interesting things start to happen when two different coils are used. Most people these days have charged a cell phone. The cell phone charger has a transformer in it that ďsteps downĒ the voltage from 120V to something much smaller that the phone can use. This is accomplished but having a much smaller secondary coil than the primary. (Yes I know that charging circuits have more parts, Iíve built a few, Iím just trying to keep it simple). The exact opposite happens if you have a larger secondary coil than the primary (the voltage is ďstepped upĒ). This is how the ignition coil in the car takes twelve volts and creates a spark of tens of thousands of volts.

But a carís electrical system is DC, not AC and transformers donít work with DC. So how is the coil fired? Well one end of the secondary coil is connected to ground and the other to a spark plug or distributor. The primary coil is connected to twelve volts on one side and ground on the other. The ground side is connected to ground through some kind of switching mechanism. This switching mechanism is used to ďbreakĒ the circuit. When this happens the primary coilís magnetic field collapse very quickly. This rapid change in magnetic field around the secondary coil induces a very high voltage current. The secondary coil sends this voltage to either a plug or distributor.

The switching mechanism on older cars are physical switches driven by a cam in the distributor. This is what is called a points system. Because this is entirely mechanical the parts would wear out and need to be replaced as part of regular maintenance. The next step was to remove the points and use position sensors to activate a switching transistor. This is the basis of electronic ignition control modules. As cars became ďsmarterĒ more control was excersized over spark timing. Instead of simply triggering the coil the sensors would send the signal to the ECU which would then use switching transistors to trigger the coil. For the most part these transistors are located in the engine bay in a protective case (usually aluminum for heat sinking) near the coil. Auto manufacturers call them ignitors. Bear in mind that this is an over simplified explanation as most ignitors have some more components in them to protect the ECU from voltage spikes and some filtering.

A wasted spark system does away with the distributor all together. It uses engine position and speed to determine when to fire a coil. It also uses one coil for every two cylinders. It is wired a little differently in that instead of the secondary being wired to the distributor on one end and engine ground on the other, each end is wired to a spark plug. When the coil is triggered the circuit is completed by the current travelling down one of the plug wires, arcing the plug, into the cylinder head, arcing the other plug and returning to the coil to complete the circuit. The term wasted spark comes from the fact that only one of the two cylinders is actually starting its power stroke and will ignite the air fuel mixture. The other cylinder being fired is just finishing its exhaust stroke and as such the spark will do nothing for/to it. One of the big advantages here is that there are no longer distributor caps and rotors to replace. Spark can be stronger as well because there no longer is a gap for it to jump from rotor to distributor cap. Each coil only fires half as often as a single coil/distributor set up. And finally the ECUís control over spark timing is no longer limited to what works in the range of the rotor alignment in the distributor cap. This is better for high revving engines. These are fairly simple to wire up as they only require two wires to the coil (12V and switching ground). However, if one of the two spark plug wires come off, you will loose both spark plugs.

The next step is the coil near plug applications. I believe that some Chevrolet high performance engines use this set up. This uses one coil per cylinder with a short spark plug wire to the plug. There are some different coils out there that can be used with this. The simplest has three wire (not including the spark plug wire) connections. One wire connects to 12V, one to the control circuit and one to the cylinder head ground. As with a single coil setup the ground wire is connected to the secondary coil and completes the circuit with the spark plug. There are also four and five wire coils. These usually have the igniters built in and can be driven with lower input voltages. For example a four wire coil would have a 12V in to the primary coil, a 5V switching signal fom the ECU, A ground to the engine on the secondary, and a signal ground to the ECU. The megasquirt can control up to six ignition outputs, so each cylinder can be driven individually on a four banger. Alternatively two ignition outputs can be wired to four drivers (two drivers each) to run the coil near plug in a wasted spark configuration.

Finally there is the coil on plug setup. These are usually wired similarily to the coil near plug, but do not have a spark plug wire going to the plug. Instead the unit itself plugs onto the spark plug. Like the coil near plug this can be wired to run in wasted spark or individually triggered. The advantage here is that there are no plug wires to wear out. Thatís about the only advantage I can think of. They would be closer to the engine heat and have to put up with more vibration. However, there certainly is a ďcool factorĒ in running coil on plug.

My plan is to use the camshaft position sensor from my AE101 4A-GZE and I will probably use the stock wasted spark coils from that engine as well. The coils arenít very large so I will be keeping my eye open when I next visit the wreckers for something a little more powerful. The transistor drivers for the Megasquirt are actually Bosch units that have been used by auto manufacturers for years. The Megasquirt information online states that if using more than two a new connector should be put on the box as the traces on the board just arenít up to handling the power. I plan on not mounting the transistors in the box at all. Instead I will mount them on a heavy duty heat sink and put them in the engine bay somewhere (in a weather proof container of course). This should keep dangerous currents out of the MS ECU and keep the ECU box temperatures down a bit too. Iíve been thinking that I would like to try and do a coil near spark individually fired system. The advantages being longer spark times possible and each coil is only fired a quarter as many times as the stock coil and dizzy. However thinking about it I believe that a coil on plug set up might be better. I came to this conclusion after thinking about running the plug wires and mounting the coils. First off the most logical place to mount the coils would be on the side of the engine facing the front of the car. This puts them right above the exhaust and thus trying to keep them cool might be an issue. The next thought would be off the engine but near the fire wall. Again it would be near the exhaust. Not to mention that I donít think I would like the way that the plug wires would look coming over the exhaust cam cover. The coil on plug would look a lot neater. And if I could find some small enough to fit between the sides of the head I could even cover them all with a plate for a very clean looking upper part of the engine. But I am getting ahead of myself.

What I will likely do is start by using the cam position sensor and the stock coils. Once I have that up and running I will start looking for some coil on plug units. I will then connect those up essentially the same way as the wasted spark coils and get that running. Because the spark outputs should be the same it should be straight forward. Then I will connect another set of drivers and get the individually firing coil on plugs working.

General Purpose Inputs and Outputs

The Megasquirt has eight user configurable inputs/outputs. At first glance this looks great, however it isnít quite that straight forward.

First off these inputs/outputs get used for different features one might use. So depending on what the Megasquirt does for your engine determines how many outputs you can use. For example, the ignition outputs for a four cylinder wasted spark system requires two connections, one at D14 and one at D16 (these were originally the injection and acceleration indicator LEDs). If I were using the stock coil and distributor I would be able to use the purpose built on board ignition driver and still have D15 and D15 configurable in the Tuner Studio Output Port Settings.

The options in tuner studio allow you to set seven outputs. They are labelled PM2 Ė Fidle, PM3 Ė Injection LED, PM4 Ė Accel LED, PM5 Ė Warmup LED, PT6 Ė IAC1, PT7 Ė IAC2, and PA0 Ė Knock Enable. Each output can be enabled or disabled and be set to a power on and trigger value of 0 or 1. It sounds complicated but itís not. Power on value determines whether the output is 0 or 5 volts when the MS is first turned on before any conditions are applied. So lets pretend it is turning on an LED once the RPM rises above 5000. If power on is set to 0, then when the key is turned to the on position the LED would be off until the RPM is above 5000. If the power on is set to 1, then the LED would be on when the key is first turned for as long as it takes the MS to figure out that the RPM is below 5000. In my case I will set this to 0. The triggered value is the logic output. Lets take the same example with the LED coming on above 5000 RPM. Setting the triggered value to 1 activates the LED just as expected. Setting it to 0 would ďinvert the logicĒ making the LED come on until the RPM rose above 5000 where it would turn off.

Now the conditions for the port to activate can be set. The port can come on with one or two conditions. With one condition it is simple. For example: when coolant temperature is higher than XX turn on the radiator fan. Setting two conditions is pretty simple too. There is the option for using AND or OR logic. If two conditions are set with an AND then both conditions must be met before the output is activated. For example: When coolant temperature is higher than XX and RPM is higher than YYYY, turn on the fan. The other combination is the OR. In this case only one of the conditions must be met and the output will activate.

Also hysteresis must be set. This tells the MS the amount that a variable must change before the output can be reset. For example, letís consider a fan that is turned on when the coolant temperature is above 50 degrees. So the engine is started cold and driven. The temperature rises to 50. As soon as it is higher than 50 the fan comes on. Now if the fan is good enough to cool the coolant almost instantly to 49 degrees the MS will turn off the fan again. If the engine is producing enough heat to instantly get it back up to 50 then the fan would come on again and the cycle would continue with the fan rapidly being switched on and off. But if a hysteresis of 5 degrees is set then the fan would come on once the coolant is above fifty and stay on until the coolant drops to 45 degrees.

Now in my application I had to decide which outputs I would use. First I had to go through and figure out what I needed to connect. This is what I came up with:

PT6 and PT7 are unavailable to me because I will be using the MSIIís stepper motor idle air control. So right there I lose two of my outputs.

PM3 and PM4 are used as ignition outputs for my wasted spark set up. So there goes another two.

PM2 will be configured as my tachometer output. The stock tachometer in the car is set up for a single coil system and will not work with the wasted spark setup. Even getting this to work will require a circuit and some experimentation. The following is a circuit diagram from the MS-II extra webpage, I am told it will drive the stock tachometer:


That leaves me with two outputs: PM5 and PAO. It is important to note that if I were to go with and individual coil for each cylinder and not waste spark, I would lose these two outputs.

So what am I going to use these for? Well I donít currently know if I will need both of them. I need to find out if the stock ECU is responsible for turning on the radiator fan or if it is on itís individual circuit. If it is on its own circuit then I will only need one output. If it is controlled by the ECU then I will have to wire up a transistor to drive a relay to turn the fan on when the coolant temperature is above a certain level. Not wanting to have a feature that I donít use, I will probably mount an LED in the dash to indicate a condition from the Megasquirt if I donít need the fan control. The two conditions that come immediately to mind are WARMUP and KNOCK. I plan to use the coolant temperature based rev limiter function and having an LED next to the water temperature gauge would be a good indicator that the driver shouldnít push the car as it is still cold. The knock indicator would be more useful when I swap the supercharger for a turbocharger and am experimenting with boost levels. Of course once the supercharger is removed I could run both as I wonít need the other output:

The other output is for the supercharger. The supercharger is one of those things that make it harder to install an aftermarket ECU. Many people have put a switch in the cabin of the car so they can turn the supercharger on and off Mad Max style, but this isnít the greatest setup in my opinion. One would think that fuel economy would increase with the charger off, but then you are running a weak lower compression engine that has to work harder (use more fuel) to accelerate even a little. Leaving it on all the time is also inefficient as it takes a fair bit of power from the engine to rotate so that would equal an increased load during idle.

So I plan to set up the PA0 output to give a five volt output when the manifold absolute pressure (engine load) reaches a certain value OR the throttle opened a certain amount. It will require some adjusting while driving the car to find the correct point to have the supercharger come on. The idea is that when cruising at part throttle and the road starts to go slightly up hill the load on the engine is increased and then the supercharger is engaged to give the car the power needed to maintain speed on a hill at part throttle. The other condition is simply when I want the car to go faster I press the pedal. If I press hard enough then the ECU engages the supercharger to give me the power I want/need.

The obvious issue here is shifting. If Iím trying to accelerate quickly I will put my foot to the floor. When it comes time to change gears I will press in the clutch, release the accelerator, shift, release the clutch and press the accelerator again. When the clutch is in and my foot is off the gas the engine has no load and the throttle position is closed. The Megasquirt will take this information and determine that the supercharger should be off. Then when Iím in the next gear it will switch back on again. This is a little harder on the components and can affect acceleration times.

Toyota addressed this issue by having a supercharger disengagement delay. This means that when the conditions are met to turn the supercharger off it will still wait a few seconds (I think I read somewhere that it is three seconds) before actually turning off the supercharger. So I designed a circuit to mimic this behaviour. I went through a couple of prototypes until I came to what I think is the most efficient answer. The circuit is made up of six components: Two diodes, a resistor, a capacitor, a darlington transistor, and a relay:


And here is the prototype I built as a proof of concept:

Without getting too technical the darlington is set up as an on/off switch which is turned on by the five volt output from the Megasquirt. When it is turned on it activates the relay which in turn activates the super charger. The darling ton Iím using is the TIP122. It has a built in resistor from the base to the emitter. With this in mind I built a simple rc circuit as a time delay. The on signal is fed through a current limiting resistor before entering the base of the transistor. Prior to this resistor is a capacitor connected to ground. When the Megasquirt sends this output high, the transistor switches to its ďonĒ state and the capacitor charges. When the Megasquirt switches this output low again the circuit remains active for as long as it takes the capacitor to discharge. In order to make sure that the capacitor does not discharge back into the Megasquirt a diode is placed between this circuit and the MS output. The last diode is used as a quenching diode on the relay to protect the circuitry from voltage spikes cause by the sudden collapse of the electromagnetic field when the relay is turned off. A resistor value of 27k and a capacitor valued at 470uF gave me a two and a half to three second delay on my prototype on the bench.

The reason to use the darlington instead of a standard transistor is that a standard transistor would discharge the capacitor more quickly.

Since this circuit is so simple it would be pretty easy to put all of it (minus the relay) in the proto area of the MS. This way only one wire on the DB37 connector need be used to drive the relay that turns on the supercharger.

Wll that's all I have for now. I am still waiting for my stuff to arrive in the mail. Once some of the parts actually show up I will take more pictures.

If anyone has any questions, comments, suggestions, tips, or want's to tell me I'm wrong on something, please do. I only ask that you are polite about it.


joe's gt
03-16-2010, 10:55 PM
Berg, you gonna tune it yourself?

03-16-2010, 11:08 PM
Can I have yur 4agze dizzy? I'll trade you a usdm one! Nah but very good job there

03-16-2010, 11:10 PM
I would like to, but as of right now I don't have any experience. I have to read up on tuning quite a bit before I try it. There isn't much around here in terms of accessable dynos. Off the top of my head I can only think of one and they are expensive and don't like imports. However there is a whole bunch of information on tuning the Megasquirt online. I'll be spending much time going over that.


Can I have yur 4agze dizzy? I'll trade you a usdm one! Nah but very good job there

I'd like to keep one of each, just in case. I'll be making frequent trips to the wreckers in the comming months, I'll keep my eyes open for one for you (though these are fairly rare in my neck of the woods).


03-17-2010, 12:06 AM
Holy technical write up Batman

03-17-2010, 12:26 AM
Holy technical write up Batman

Just trying to make sure I understand what's going on. ;)


03-17-2010, 12:33 AM

03-17-2010, 12:47 AM

Putting in a megasquirt.
Figuring out what I need to make stuff work.
Think I got a big portion of it figured.

joe's gt
03-17-2010, 02:04 AM
I would like to, but as of right now I don't have any experience. I have to read up on tuning quite a bit before I try it. There isn't much around here in terms of accessable dynos. Off the top of my head I can only think of one and they are expensive and don't like imports. However there is a whole bunch of information on tuning the Megasquirt online. I'll be spending much time going over that.

Yeah, its a risk, but its also a black art and no one likes to share information on it. So I hope you end up trying it out. You easily have enough knowledge to try it out. Jeff Hartman's book and the tuning tips section of the efi101 forum have a great info in addition to the megasquirt site. Good luck man.

03-17-2010, 02:15 AM
Thanks! I'll definately be doing a lot of reading on the subject. The megasquirt will not be installed in the car until I have the wideband O2 sensor and gauge is installed. It will definately be an adventure. I'll keep posting as things develope.


03-17-2010, 05:28 PM
A word on sensors

In order for the ECU to inject the proper amount of fuel, it needs to know a few things about how the engine is behaving. It achieves this by using several sensors to determine what the engine is doing. I already mentioned RPM measurements when I was talking about the ignition system.

Throttle Position Sensor (TPS)

The TPS is actually a fairly simple device. It is a potentiometer. Usually the TPS will have a three wire connection. A potentiometer has three connections. Think of it as an electronic resistor. There is a connection at either end and one in the middle. Measuring the resistance from one end to the other would give the maximum resistance value. The middle connection is on an arm that can be moved from on end to the other. When the arm is moved to one end and the resistance is measured from that end to the arm is will give a value of near zero (a short). If the arm is then moved to the other end and the measurement is taken again then the value would read maximum resistance. If a voltage is applied to one end and a voltage measurement made on the arm it would change depending on where the arm is (how much resistance is applied). So if ten volts is applied at one end, the measurement will read ten volts when the arm is at that end and decreases as the arm is moved to the other end. This is essentially what the ECU is doing with the TPS. One end is tied to ground, the other has a reference voltage (Vref) applied to it. In the case of MS this Vref=5Volts. The last connection is called the TPS signal. This is simply a voltage reading that will vary from a less than 5 to 5 volts depending on throttle position. (I say less than five volts because not all TPS will give a reading of 0 volts because the arm doesnít always travel all the way to the ground connection, similarily the arm doesnít always go all the way to the Vref). Different car manufactures wire the TPS differently. Some have a TPS signal equal to Vref at closed throttle position and some have the TPS signal equal to Vref at Wide Open Throttle (WOT). One thing that should be true for all TPS systems is that the TPS voltage signal should increase/decrease linearly across the entire range of motion. This works well for the MS because it calibrates the TPS signal by measuring the Vref at closed throttle and again at WOT. It then interpolates between the two to determine how far open the throttle is for any given TPS signal.

Coolant Temperature (CLT) Sensor

For an engine to work at its peak efficiency it needs to be warmed up. The ECU needs to know if the engine is warm or not. If the engine is not then warmup enrichment is applied and the ignition timing is adjusted for the sole purpose of warming up the engine faster. A good way to measure how an engine is doing temperature wise is to measure the temperature of the coolant. The sensor used is very simple. It is simply a temperature dependant variable resistor or thermistor. When the temperature is -20 degrees it has a resistance of XX and when it is 30 degrees it has a resistance of YY. Thatís pretty much it. There are two main variations of this sensor, a one wire and a two wire. The one wire (like the ones that come with autometer gauges) connect that one wire to the device reading the temperature (gauge, ECU) and then the circuit is completed by the sensors physical connection to the metal of the engine as a ground. The two wire sensors donít rely on the engine ground and use the second wire as a ground returned to the device reading the temperature to complete the circuit. These sensors are said to be better as they are independent of all the other systems that use the engine as a ground. The MS comes pre-calibrated for a specific GM coolant temperature sensor. It can be recalibrated to other stock sensors but I plan on just using a new GM unit. I have several locations I can install the sensor into. There is the stock ECU sensor, the cold start sensor/switch (MS doesnít use a cold start injector but instead increases the amount of fuel injected by the main injectors), a couple of temperature dependant vacuum switches (for devices like the EGR which has already been removed), etc. This is one of the ways an aftermarket ECU can help clean up the engine bay. Many of the coolant temperature sensors are for pollution control which is usually removed with a new ECU.

Intake Air Temperature (IAC) Sensor

As most people already know injecting the correct amount of fuel into an engine requires that the amount of oxygen in the engine that is available for combustion be known. This is determined with a couple of sensors. One is the IAC Sensor. The temperature of the air determines how dense it is. The density of the air directly relates to how much oxygen there is in a particular volume. The colder the air is the more oxygen there is to burn. The more oxygen there is the more fuel can be injected and the more power can be produced. If the air is hotter then it can not burn as much fuel so less fuel has to be injected to maintain the proper Air Fuel Ratio (AFR). This is why the ECU needs to know the temperature of the air entering the engine. The sensor itself operates exactly the same as the CLT Sensor. The difference being that it isnít installed in the coolant. The MS is again set up for a GM sensor. This sensor is an ďopen elementĒ sensor in that the actual material that measures the temperature is exposed to the air. This is the sensor I plan to use. It has been said that these sensors are somewhat fragile so some people use oxygen sensor safe silicon on them to hold the element in place. What I havenít yet decided on is where I will mount this sensor. The recommended place is just before the throttle body on the intake. I was considering placing it where the cold start injector is now as a convienant way to plug up that hole, but have read that the vibrations and heat from the engine would cause it to wear out quickly. On naturally aspirated engines it doesnít really have to be placed in the intake track at all, just near where the air is sucked in. On forced induction engines it is better to place it nearer the intake manifold (after the intercooler) to account for the temperature increase caused by compressing the air. The stock IAT sensor on my car is located in the Mass Airflow (MAF) Sensor housing.

Manifold Absolute Pressure (MAP) Sensor

The MAP sensor is the other sensor that is used to determine how much air is entering the engine. The car came stock with a MAF sensor that has a ďdoorĒ or ďflapĒ in it. When the engine needs more air the difference in pressure would force this door open. The sensor itself is attached to this door and works very much in the same way as the TPS. There are many arguments on the internet as to whether the MAF or the MAP is a better system. The MAP is simpler in that it only requires a vacuum line connection to the intake manifold. However the MAF can be beneficial to certain engine modifications such as larger cams. Aggressive cams can fluctuate the pressure in the manifold causing a MAP equipped ECU to constantly be making corrections and thus create a poor idle. With a MAF system the air going in is automatically averaged over all the cylinders and so a smoother idle can be achieved. For the most part though, people who have MAF systems donít like them. Many claim that they are restrictive and thus ďchokeĒ some of the engines potential. They have more wires that one must deal with. And a problem that I have experienced, if there is a leak in the intake after the MAF that air is being sucked in (or blown out under boost) that isnít accounted for resulting in incorrect AFRs.

MAP sensors have taken me a little time to get my head around, and Iím still not 100% confident that I fully understand them. I am used to having a boost/vacuum gauge in my car. When the engine is off this gauge reads zero. When the engine is started and there is no boost there is vacuum. This is where part of my understanding problems come from. ďOf courseĒ I think, the engine is sucking in air, there is going to be a vacuum. When the supercharger engages it creates boost and the gauge goes from the vacuum side to the psi above zero side. What I had to learn is that what the boost gauge is reading out is the Manifold RELATIVE Pressure, in that it is relative to the atmospheric pressure. So an engine boosting at 5psi has a MAP of 5psi above atmospheric pressure. MAP sensors do not give a ďbelow zeroĒ value. If a MAP sensor gives a value of zero then the manifold is in absolute vacuum. The only way this is ever going to happen is if your car is in space. When the throttle is closed the pressure in the manifold is decreased below atmospheric because the engine is sucking the air out of the manifold and the throttle is preventing more from getting in. When the throttle is opened the pressure in the manifold increase toward the atmospheric pressure value. When the engine is at WOT the pressure in the manifold is near atmospheric (on a naturally aspirated engine). It will be a little less than atmospheric just because the engine moving air out of the manifold. On a forced induction setup the pressure in the manifold actually increases above the atmospheric pressure.

When I purchased my MS I ordered a MAP upgrade kit. This kit has a MAP that is capable of reading a higher boost level than the original one. Considering that I am planning to eventually switch to a turbo and that I currently do not know how much boost I will be able to make I wanted to cover all my bases. The kit comes with a second MAP sensor on the board for barometric corrections. The MS compares the value of the MAP to atmospheric pressure to determine the engine load. With the stock setup the MS can be told not to do that or to take the MAP reading when the car is turned on but not started as the atmospheric value. This usually works fine, but if one were to drive up a mountain (like the ones an hourís drive west of where I live) the atmospheric pressure would change significantly from bottom to top. Having a second sensor open to the air provides the ability to make real time corrections. It isnít essential by any means, but I think its useful.

In the tuning software the MAP value is used on the y-axis of the tuning tables. Older versions of the software actually displayed the kPa value. However, newer versions have done away with that and replaced the kPa with fuelload %. This is a representation of MAP. Unfortunately I do not yet know how to use these values. For example, does 100% fuel load equal atmospheric pressure? Does that mean that I have to input values above 100% for under boost conditions? Or can I adjust the fuel load values such that 100% is above maximum boost conditions? I am still trying to find these answers.

Exhaust Gas Oxygen (EGO or O2) Sensor

This is a sensor that is placed in the exhaust system of the engine. It really doesnít do much in terms of adjusting the amount of fuel injected into the engine. This surprised me a little. However considering that the O2 sensor wire broke on one of my Celicaís and I didnít notice a change in driving (other than the check engine light being on) I can believe it. The O2 sensor is instead used to make corrections. The MS allows the user to set how much priority it has to fuel injection. I believe the recommended value is around 7%.

The EGO sensor is used to provide feedback as to how much oxygen remains unburnt in the exhaust. The MS tuning software has an AFR table in it that has looks like the other tables (has fuelload% on the y axis and RPM on the x) but the values in the tables are targets for the EGO sensor reading. The MS then will try to adjust the amount of fuel injected to give the oxygen sensor the reading specified in the AFR table. It will only do this to a certain extent depending on how much priority you give it. It is even possible to run without an O2 sensor all together.

There are two types of O2 sensors that the MS can use, narrowband and wideband. Narrowband is typically what vehicles are equipped with stock. The basic ones have a one wire setup. When the O2 sensor heats up it creates a voltage based on how much oxygen is in the exhaust. The one wire units are ignored until the engine has warmed enough to create the proper voltage in the EGO sensor. The next step up is a heated O2 sensor. This is the same as a one wire, but has three or four wires. On the three wire one wire is the sensor, one is ground for the heater, and one is power for the heater. The four wire is the same, but the sensor has a signal ground running to the ECU as well. All the heater does is make he EGO sensor reading valid sooner. The narrowband sensor will output a voltage between zero and one volt. Essentiall all it does is tell the ECU "lean" or "rich". That's it.

The other type of sensor, the better one, is the wideband O2 sensor. These are typically more accurate that the narrowband. Instead of the varying 0-1 Volts that simply let the ECU know if it was lean or rich the wide bands use a broader voltage range and work with a linear scale. As such a specific voltage output translates to a specific AFR. With a wideband O2 sensor the MS can be trusted to make better corrections using the AFR lookup table. One of the things that I will be purchasing when it comes time to install the MS in my car is a wideband sensor and controller. They can be purchased from the same place I bought the MS and even offer a gauge that can be mounted in the dash for real time outputs. Everywhere I have researched state that tuning should not be done without a wideband EGO sensor.

Again, if anyone wants to correct me, please do.


03-18-2010, 09:16 PM
There's a good 5-6 dyno's in Calgary.

But yes, the time is pricey. I had my 165 tuned at http://www.dynomotive.ca/ by Toma. His rate was very reasonable in comparison to some that I saw. I could get it sorta close by street tuning, however it is definitely a lot easier on the heart to have a professional do it. Plus you can see and hear a lot more that is going on with the dyno.

Good luck! I know this has been a big project in the making for you.


I would like to, but as of right now I don't have any experience. I have to read up on tuning quite a bit before I try it. There isn't much around here in terms of accessable dynos. Off the top of my head I can only think of one and they are expensive and don't like imports. However there is a whole bunch of information on tuning the Megasquirt online. I'll be spending much time going over that.


joe's gt
03-18-2010, 09:45 PM
There's a good 5-6 dyno's in Calgary.

But yes, the time is pricey. I had my 165 tuned at http://www.dynomotive.ca/ by Toma. His rate was very reasonable in comparison to some that I saw. I could get it sorta close by street tuning, however it is definitely a lot easier on the heart to have a professional do it. Plus you can see and hear a lot more that is going on with the dyno.

Good luck! I know this has been a big project in the making for you.


Chris said it better than me. I think it would be awesome for you to do your own street tuning. But in all honesty, you should take it to a dyno after your done on the street to extract more and iron out any mistakes or things you may have missed.

03-18-2010, 10:23 PM
For sure I absolutely agree. I didn't mean to imply that I wasn't going to a dyno. There is no possible way to get the best tune without using a dyno. I just meant that I am planning to try and get it tuned to the best of my ability on the street before taking that step.

Thanks for the replies guys. My MS is still sitting in customs according to the tracking information. I'll keep this thread updated as things develope. Hopefully I get the details right.


03-18-2010, 11:23 PM
fyi, the JDM 4age map sensor dizzy is a 24 tooth bottom and a 4 tooth top

03-19-2010, 01:52 AM
fyi, the JDM 4age map sensor dizzy is a 24 tooth bottom and a 4 tooth top

So they use the Toyota ESA ignition system. That's good to know, it would work well for a MS setup.


03-19-2010, 04:08 PM
In the tuning software the MAP value is used on the y-axis of the tuning tables. Older versions of the software actually displayed the kPa value. However, newer versions have done away with that and replaced the kPa with fuelload %. This is a representation of MAP. Unfortunately I do not yet know how to use these values. For example, does 100% fuel load equal atmospheric pressure? Does that mean that I have to input values above 100% for under boost conditions? Or can I adjust the fuel load values such that 100% is above maximum boost conditions? I am still trying to find these answers.

The software has simply relabled the kPa as fuelload %. The numbers themselves are still exactly the same. So if I were t enter a value of 100% fuelload it would simply mean 100kPa.

Tracking still says that the MS is in customs, I guess I wont get to play with it this weekend.

There's a good 5-6 dyno's in Calgary.

But yes, the time is pricey. I had my 165 tuned at http://www.dynomotive.ca/ by Toma. His rate was very reasonable in comparison to some that I saw. I could get it sorta close by street tuning, however it is definitely a lot easier on the heart to have a professional do it. Plus you can see and hear a lot more that is going on with the dyno.

Good luck! I know this has been a big project in the making for you.


I checked out the dynomotive website. $75 for a baseline, $100/hr for me to tune and $150 for them to tune. I admit I'm kind of tempted to take the car down to get a baseline before I swap the ecu, just to see what kind of difference can be made.


03-19-2010, 04:24 PM
Cool. Toma's a bit less formal than the site, chances are he'll cut ya a deal on the tuning - the shop is just something he does in his spare time for fun, he has a day job too. I believe he charged me for like 2 hours and we were there tuning for about 4.

The only downside is that it is a dyno dynamics dyno, so it's a bit hard to compare to the typically posted dynojet's on the 'net.


03-19-2010, 04:46 PM
Hmm... the site advertises that they have two dynos now. I admit I do not know much about different kinds of dynos. A quick google has me reading that the mustang dyno is a true measure of torque and can load the car to simulate the work the engine has to do to move the car around. I now have another thing to look up :)


03-19-2010, 10:48 PM
Just a note, I was wrong in stating the MS-II can't do sequential fuel injection. So long as one uses the MS2 Extra firmware and Tuner Studio the unit can be expanded with various daughter boards from jbperf.com. The site also has an I/O extender board to increase the number of inputs and outputs. This board requires either some modification to the MS or to be mounte in its own case. It communicates using a CAN bus serial connection. The inputs and outputs can't be used for everything because of timing issues with the CAN bus (so injectors and ignition is out). However things where timing is less critical, like boost control, launch control, fans, etc would work. I don't need any of these for my installation but its worth a look to see what they can do. There is a youtube video of someone who used this board to bring in data from an accelerometer for data logging g-forces. It's not in a car (just on the bench) but it seems to work.


03-31-2010, 07:52 PM
Well I figure I’m overdue for an update.

The Megasquirt arrived in a box last Monday. Well they tried to deliver it, but I wasn’t home. So I had to wait until Tuesday after work to go and pick it up from the local post office. My first thought when I picked it up was that it was too small and that there was no way that everything I ordered was in there. I didn’t open the box until the next evening when I finally had a chance to work on it. I decided to put it together in my office after hours as I have better equipment for soldering there and there is no cat to “help” locate components.

This is what was in the box:


I started by assembling the JimStim. This is the device that simulates some of the engine’s signals for testing purposes to ensure that the MS is working before installing in a car. I followed the instructions here:


My JimStim had a few differences than the one on the site above. Nothing too difficult to figure out though. The voltage regulator is rotated ninety degrees anticlockwise and folded down to fit against the board. There is also a polyfuse added for protection (this is a good thing). I assembled the Jimstim and it tested fine.


Next it was time to assemble the MS. There are plenty of instructions online for putting this thing together. I followed these two:



The first link above is a very good step by step assembly with testing along the way to make sure things are working properly as you go. The second one has you install everything and then test. If there are any problems then with the second method you have to check everything as opposed to knowing where the problem occurred. What I did was write down all the modifications that the msextra instructions called for and then went through the megamanual instructions and write them in there. So I made my own set of instructions that were a combination of the two. This worked well for me and I was able to put everything together and have it all work right away.

First I checked the board for shorts and fitment as per the instructions and then assembled the power supply. At this point the Jimstim is plugged in to the connector and a multimetre is used to test both the twelve volts and five volt power supplies.


The next part that was constructed was the RS232 serial interface. This is the communication between the MS and a computer. A serial cable is connected to the MS through an interface chip that translate commands between the computer and the MS processor. The first step is to make sure that the computer is talking to the serial port. This is done by shorting the Rx and Tx pins in the connector and then using hyperterminal to send characters to the Tx pin and receive them on the Rx pin. All received characters should appear on the hyperterminal screen. There are two pins on the processor which are used to communicate to the serial chip. Again these are Rx and Tx. These two pins are shorted for testing purposes. Once shorted the Jimstim is used to power the MS board (the serial chip is the only thing that requires power at this point in the build. The serial cable from the computer (in my case a USB to serial adapter) is plugged into the board. With the pins on the processor socket shorted characters in hyperterminal should go from being input on the keyboard, through the serial cable, through the chip on the board, to the Rx pin on the processor socket, to the Tx pin on the processor socket, back through the chip, back to the computer and displayed on the screen. If this works then computer-processor communication is possible.

Next the processor is mounted and programmed with the firmware according to the instructions. Now the processor is capable of communicating with the computer, but nothing works for the inputs as that part hasn’t been built yet.


This is how the assembly went, put a little together, test it, and continue. Most of the parts that the msextra stated to leave out were left out. The exception being R37 and R38. The megamanual states that these are clamping resistors used for protection. I do not know why the msextra says to leave them out and was unable to find an answer so I put them in. Should it cause problems later I can always remove them later. I also did not install all the circuits. I know that I will be using VR sensors for RPM inputs so I didn’t bother installing the Hall sensor circuit. I also didn’t install the ignition driver, or the Fidle circuit because I am planning on using wasted spark and a stepper motor for Idle speed control. Before long I had it all together. The average expected time to put it together is stated to be eight hours. It took me a lot less time than that, but as I said I didn’t put all the circuits in and I have a fair bit of experience putting boards together (it used to be my job).


I then tested it with the Jimstim to make sure everything was working.


Now that that was done, it was time to start doing the modifications to make it work with my setup. The first thing that I did was to install the jumper wires to run the Idle air control stepper motor and the second MAP sensor.


Next I put the supercharger control circuit on the board. Half of the circuit is mounted on top of the board and the capacitor is mounted underneath. After testing I found that the 470uF cap didn’t hold it on long enough and I install a 1000uF (it was all I had on hand). The circuit now holds a relay on for almost exactly three seconds. Because I have no way to adjust the input to the MAP sensor right now I just set it up in the tuning software to activate the circuit at 60% throttle, which I can simulate with the stimulator. It works fine.




Next task was to put together and install the Zeal Engineering daughter board. This board is designed to fit in the case lid of the MS. It has two VR conditioners and several outputs to drive relays and such. However these outputs just connect to the outputs on the board for the general input/outputs. This does not add any outputs, it just conditions the ones that are already there. For this reason I only installed one (I only have one output left). This output is connected to the “warmup” LED on the board and then configured in the software. For testing purposes I set this to turn on when the coolant temperature reaches 60 C. This simulates turning on a radiator fan. Again this worked first try. To test these outputs I’m connecting them to outputs on the DB37 connector and then jumping the corresponding pins on the JimStim to the LED pins. This way when an output is activated (goes low or is grounded) the LED on the JimStim turns on. In the case of the supercharger circuit I actually have a relay connected.



Unfortunately this is where something went wrong. Nothing major. The Jimstim can simulate a dual wheel setup like the Nippon Denso cam sensor I’m planning to use, but not properly. The second tach signal is a square wave, which will not trigger the chip on the Zeal daughterboard. After discussing online with members of the msextra board I was told that a useable signal could be created by putting a small capacitor in series with the second tach signal. The next day I decided to do this. Somewhere during this procedure I must have shorted something. It was a stupid mistake as I was moving wires around while power was applied. All of a sudden the power LED on the Jimstim dimmed. I turned the power off and on again. Things worked alright for a few seconds then the LED dimmed again. When the LED dimmed the MS turned off. At first I thought I had damaged my MS. I unplugged the JimStim and powered that up by itself. The power supply showed a higher current draw and the voltage regulator on the JimStim quickly heated up. So did the chip that provides the tach signals. I removed this chip and then checked for shorts using this schematic: http://jbperf.com/JimStim/jimtim_v1_4_schem.png

The board checked out fine and I reconnected it to power. It stayed cool and all the voltages checked out. I connected it to the MS and tested again. Everything worked but the tach signal (obviously, the chip was on the desk next to the board). I then checked the chip itself. Pin three is VDD and pin 4 is GND (power supply and ground). There were 21 ohms of resistance between the two. That’s pretty much a dead short. When the chip was plugged in it shorted the board and drew too much current from the power supply. This caused the polyfuse to heat up and open. Once it cooled down again it would close and heat up again. This is why the LED was dim and not off, the fuse was opening and closing quickly. Also I had the power supply current limited so it could pull a maximum of just less than an amp. So I have ordered a replacement chip and a spare and am waiting for those to arrive.


03-31-2010, 09:51 PM
On Saturday my girlfriend went to the garden show with one of her friends. I had the afternoon to myself. I took this opportunity to do one of my favourite things: wandering around the local Pick n Pull. Initially I was just looking for a set of coils that might be a little better than the ones that came with the AE101 4AGZE. I found a Pontiac that had a DIS module for a V6. It had three coils and the connector. I figured that I could take this and have two coils and a spare. So I cut the wires about four inches from the plug and removed the unit. I then decided to wander around the imports section looking for nothing in particular.

Well as luck would have it there was a MkIII Supra there with the 7M-GTE engine. The turbo, exhaust manifold and intercooler were long gone, but the rest of the engine was there. I was only thinking of ignition so I removed the cam position sensor and the coils with the plug wires. The coils on this engine look exactly the same as the ones on the 4AGZE and I suspect that they probably share part numbers.


After I took the coils off I was getting ready to leave the car when I thought that I should maybe take the injectors. I didn't know at the time if the injectors flowed more than my 4A-GZE ones but figured that for sure they wouldn't flow less, so worst case scenario they would just be spares for what I have. I looked it up after I got home and the 4A-GZE flow 365cc/min and the 7M-GTE flow 440cc/min. I cut the wires going to the injectors so I could wire them up and use the stock plugs. I need to find a place that will clean and test injectors to ensure that they are all good. Incidentally the gentleman at the register failed to charge me for them so thats a free injector upgrade! Well worth soaking my mechanix gloves in gas.


In getting the injectors out I had to remove the throttle body first. The throttle body is much bigger than any of the ones that I have on any of my engines. I know that a 3S-GE throttle body will bolt right up to the 4A-G(Z)E engine no problem (I've done it), but I do not know if the same bolt pattern is shared with the 7M-GTE. If it is, or if the getting the 7M-GTE throttle to fit is simple then I will probably use that. If not I will use a 3S-GE throttle body.

I started to play with the TPS of the 7M-GTE throttle body. It has four wires instead of three. As far as I can tell the two at either end of the connector are the ends of the potentiometer (as I described in an earlier post) and the one nearer what would have been the top of the throttle body is the connection to the arm. The extra connection only seems to indicate when the throttle is fully closed. I do not know if this had something to do with idle control on the Supra, but for my installation I will only need the three wires. I'm also assuming that TPS connections are similar across 80s era Toyotas so what I learn with this one will be relavent to any of the ones I have available to use.


I then decided to play with the coil packs. The first thing I did was try and get it apart. Removing the connector was simply a case of unbolting it and pulling it off. The same with the coils. With the coils removed the mounting plate on the bottom also came off. Now I was left with the plastic shell with a metal cover where two of the coils were and a a metal back plate that was between the plastic and the mounting plate. It took quite a bit of time with small screwdrivers, a knife, and a scribe (dental pick) to get the plates off. Under these covers is the stock electronics for controlling the coils. This was covered by a protective goop that was a real pain to get off. After that I used the multimetre to determine which connector pin came into the case. I then cleaned everything up and mounted my coil drivers to the backplate to use it as a heatsink. Finally I connected the driver transitor through the old wiring of the case to the connector. It looks really good but until I can simulate a tach input I can't test them. Although I could try hooking up the Supra cam position sensor and spin it with a drill.... That would test both the VR conditioner circuit on the Zeal board and the ignition outputs...



I'm still undecided on what to do with the connection for the third coild. One of the pins has a constant twelve volts on it so I cnt leave it exposed. I can't cut the baseplate off at two because one of the mounting bolts fits under the coil. I could cut the pins off and cover the exposed metal with silicon to prevent any probems. I'll have to explore my options.

Feedback is always welcome!


04-05-2010, 11:01 PM
Well I am still waiting for the new chips for the JimStim to arrive, but everything else is lookig good. I spent some time going through the electrical schematics here: http://opc.mr2oc.com/

I have managed to figure out a lot of the wiring I'm going to have to do from this information. Firstly the two fans (radiator and engine bay) are on their own circuits. The schematics include both models that have AC and those that don't (like mine). The circuits I am looking at are on the first two pages of this document: http://opc.mr2oc.com/online_parts_catalog/1988AW11_complete_wiring_diagrams2.pdf.

The first page has the Engine bay cooling fan circuit. It is an entirely self contained circuit. There are two external connections, one to the ECS and one to the speed sensor. The one to the ECS was only used in 1989 MR2s and the speed sensor was only used in 1988. Mine is a 1987 so the circuit stays completely independant.

The second page has the radiator fan circuit on it. I do not have A/C on my car so not allof this circuit is present in my car. I have the shorting pin in place of the high pressure switch. This is a completely independant circuit and doesn't look to be connected to the ECU in any way.

So I do not need the extra output to run a fan as described above. I could use it to run an extra fan. For example I could put a fan on the intercooler. However when the car is moving there isn't much point in having a fan running as it wouldn't be able keep up with the air entering the engine bay. I could set it up to only activate at a high intake temp and a no load (low MAP) condition. I dont think I will be going this route, this is just an example.

The tenth and eleventh pages have the charging and starting circuits schematics displayed. The charging circuit is very simple and has no connections to the ECU. The starting circuit has one connection to the ECU and that is the cold start injector. For the most part the cold start injector is controlled by a temperature switch and the ECU has the ability to turn it on for a short time. Essentially all it does is inject a SMALL amount of fuel to aid in starting the engine only when the starter is engaged. Thats all it does. The MS doesn't use this as it just uses the main injectors and uses a cold/cranking enrichment. The MS also has a priming pulse which is designed to get rid of any air in the fuel lines. This pulse occurs before the engine is started but afterthe ECU is powered up.

I also spent a fair bit of time looking at this: http://opc.mr2oc.com/online_parts_catalog/1988AW11_complete_wiring_diagrams4.pdf. It's a little harder to read, but it has a lot more information. Figure 1 on page 3 has the radiator fan relays on it. Since I do not have A/C I don't know how many of these relays there actually are on my car.

Figure 5 on page 7 has one of the main relay blocks located in the engine bay as well as the cold start injector and starter circuit. Looking at this I realised that I have a couple of options. I could wire in the MS to use these stock relays or I could do all my own wiring. The MS instructions state that the injector should be powered by the fuel pump relay. Looking at this diagram it can be seen that there are seperate relays for the EFI, Injectors, Engine, etc. I believe that it will be easier for me to reduce the number of relays and install the system like the MS sites instruct to. I will probably mount a fuse panel with whatever relays i need in the rear trunk of the car. I have a wiring harness from an NA MR2. My plan is to use this to connect the MS to the car for the circuits I need (power to the MS, starter, signal to tach, etc) and remove the rest that I don't need. I will leave the relay block where it is (I still need the starter an cooling fan relays) and mount what I need in the trunk.

Finally Figure 7 on page nine has the wiring for the stock ECU for the 4A-GZE engine. This schematic gives me an idea of how much simpler the wiring will be with the MS. Firstly the Aiflow meter will be gone This has the stock IAT sensor in it and that will be replaced with a GM unit. The circuit opening relay is what turns the fuel pump off when no air is passing through the airflow meter, this too will be removed. All the Vacuum switching valves (VSV) will be removed. This will cut down on the Vacuum lines quite a bit. All that I will need will be the vacuum line to the brake booster, one to the MS, one to the fuel pressure regulator, and one to my boost gauge.

Other things I learned from this diagram is that the stock super charger relay is switch on when it receives a ground from the ECU. This works perfectly with the circuit I built for it on the MS. The fuel injectors are run in two banks from the stock ECU. I can only assume that this runs the injectors in a semi-sequential format. This is exactly as the MS is set up. The ignition coil/ignitor/distributor will be removed in favour of a wasted spark directly driven coils.

Lastly I've been looking at this and toying with the idea of keeping the diagnostic port. From the schematic: FP is the fuel pump. It is connected after the relay so it can be used to either determine if the fuel pump is on or power it through this pin. E1 is a ground. OX is the oxygen sensor reading. I don't know if I can connect this with the wideband inovate sensor I'm planning on using. It has two outputs, one for the MS and one for a gauge. I don't currently know what T is. It is a direct connection to the ECU. The same holds true for VF. I know that in order to set the timing two pins have to be shorted in order to defeat he ECU's control over the electronic advance. I think one of these pins has to be shorted to ground. To be honest I never memorized which pins need to be shorted, I always have a Chilton or Haynes book with me when doing maintenance. +B is twelve volts from the EFI Main relay. Finally IG- is connected to the negative side of the ignition coil. This can be used to drive a tachometer while adjusting timing and idle speed. It shouldn't be too hard to wire this up to work similar to stock. Of course this isn't really necessary as the MS allows you to simply plug in a laptop to see what it is doing. I have a spare from the NA MR2 harness.

Well that's where I'm sitting now. I am trying to find some DB15 connectors to mount in the MS case for my ignition outputs. I will then mount everything on a piece of wood and use all the sensors and devices that I will use on the engine and make sure that all works before diving into the car's wiring.

I will update as things happen.


04-12-2010, 07:50 PM
I'm still playing the waiting game on the chips for the JimStim. Sometimes waiting is the hardest part.

04-16-2010, 07:39 PM
Alright, so the chips arrived yesterday.

I plugged in a new chip and fired up the JimStim to ensure that it was indeed the chip that was causing issues. Everything worked fine.

I wanted to test the second VR conditioner on the Zeal daughterboard so I set the JimStim to nippon denso mode. The second tach output from the JimStim is simply a squarewave with no zero crossing so it will not trigger the VR circuit without some modification. I have to put a capacitor in series with the signal and add a pull up voltage of twelve volts it should work. Instead of just throwing the parts together and hoping for the best I was planning to use the osciliscope in my office to compare the two tach signals. Unfortunately the scope is being used for another project that is work related so it takes priority (I'm doing most of the work on the MS in my office after hours and during lunch). I probably wont get the scope back until next Wednesday or Thursday.

I hit another limitation, although this one was expected. I set the JimStim and the MS up for a single toothed wheel and tested everything using Tunerstudio. I then connected the GM ignition coils and tested that at low rpm. I did get a spark but it was weak and inconsistent. My power supply simply did not have the juice to power the coils. Just after firing the coils would charge up again and this sudden demand for current activated the over current protection on the power supply. Although it deactivated before the voltage dropped to zero the coils never got the juice they needed to charge. As such only one coil fired sort of consistently and the other had a very weak spark. The simple solution to this is to use an actual car battery and a battery charger to power everything. Before I even think about doing that I have to go through all my wiring again to make sure it is good and GET FUSES.

I went to a couple of stores here in town looking for fuse blocks last night. I need to fuse all the systems that are controlled by the MS via relays. The MS has an output that connects to ground when active to drive a relay to turn on the fuel pump. I will be using this output to drive three relays. One will power the fuel pump and O2 sensor. The fuel pump will have a 10A fuse and the O2 a 5A fuse. One wll power both injector banks, each with a 5A fuse. And the last will power the ignition coils. I do not yet know what fuse rating I need. The MS only turns on the fuel pump when the engine is either cranking or running, so controlling the power to the igition and the injectors with this signal will ensure that nothing will get powered or burn up while the engine isnt running.

Unfortunately the selection of fuse blocks in town is somewhat limited. In fact the only non-glass tube fuse block I could get was a six position block with a common bus down the center. This will work fine for testin purposes, but I was hoping for a little more circuit isolation. To that end I ordered three four position fuse blocks made by Painless from jegs.com. This will allow me to make a slightly neater package. The MS itself requires a 2A fuse.

At this point I feel that it is probably easiest to bolt everything down to a peice of wood and use that for testing purposes. Once I get that done I'll post up some more pictures. The plan is to then start replacing signals from the JimStim with the actual sensors that wll be on the car and make sure that all runs on the test board. At that point I will be confident that it will work in the car and can finally start the process of removing the old wiring.

Many people who install these end up wiring a connector to plug into the stock harness so the MS can be plugged right in without rewiring. I have decided not to go that route. I have a spare harness from an NA MR2 that I will use to provide the power from the ignition switch and the signals to the stock gauges. Everything else will be removed. I will make my own harness from the MS to the engine. All the vacuum lines will be removed except the fuel pressure regulator, brake booster, and boost gauge. One more will be added for the MAP sensor. The stock wiring to the oil pressure and coolant temperature sensors will remain, as well as th wiring to the starter and alternator.

It will be quite a job, but I am looking forward to it. As always, comments and critisms are welcom.


04-16-2010, 08:41 PM
Well my supervisor gave me the call to tell me to go home early today. I work on a university campus for a research group. It is the last day of classes so around sunrise the students start to drink. We were sent home to avoid the craziness. I just grabbed a scope and locked myself in my office.

I don't know exactly why, but after trying different trigger wheel settings te MS is now happily running off the JimStim in nippon denso mode. The second VR conditioner is now confirmed to be working.

I would start playing a bit more with it, but it's Friday, its sunny, and there is cold beer in the fridge!


04-21-2010, 04:32 PM
Not much of an update. I've spent the last couple of days in the field for work, and I have to unload the trucks today. I did pull the throttle body off my spare 4A-GZE. I have discovered two things. One: the 7M-GTE throttle body does not share the same bolt pattern as the 4A-GZE. Now I could make (or ask a buddy of mine to make) an adapter for it, but that would be part of the "down the road turbo project". Right now I just want to get the MS running the 4A-GZE. I know that the 3S-GE has the same bolt pattern and is a little bigger than the 4A so I might just end up using that.

The second thing I found was that the stepper motor for the idle air control on the 4A-GZE is wired differently than I expected. I was expecting a four wire setup (which is what the MS is currently setup for), but instead it only has three. I'm only guessing here but I would think that one is a common (ground or power) and the other two are switching. Switch one wire to goe one way and the other to go the other way. I'll have to experiment with it when I get some time. If I can't get that to work with the MS then I would probably just run an external idle air control. Diyautotune has a mount for a Jeep one that could be used on virtually any vehicle.

Edit: Found this: http://www.glensgarage.com/catalog/wire-idle-valve-driver-board-p-79.html
Schematic is here: http://www.glensgarage.com/3wire/Idle-Coil_Board_schematic.pdf
So now the question becomes do I build another circuit to control the stock idle control or do I use a four wire one off of another vehicle? Right now I'm leaning towards the using another vehicle's four wire idle air control. Specifically the Jeep one just because diyautotune has everything for sale that I need to get it to work (except for the stepper valve itself). The reason I am leaning this way is that the 4A-GZE is the onlythrottle body I have that has the built in stepper valve. It is also the smallest throttle body I have. If I were to upgrade to the 3S-GE or even the 7M-GTE throttle body I would still need the external IAC. FYI, I haven't been able to confirm, but the 2nd gen 3S-GE and 3S-GTE might have the same stepper IAC as the 4A-GZE. Also using the 3-wire circuit means that I have to use the Fidle output which means I'd have to use my last spare output as a tach driver leaving me with no flexibility.


04-22-2010, 06:15 AM
If you do need that 3sge TB you gave me, you are definitely welcome to it. I have have the st205 tb on mine, and already have a spare st185. You are also welcome to the 185 one if that helps.

For IAC, I would definitely use the Jeep one. It's early on, chances are at some point you'll want that output for something.

04-22-2010, 06:34 PM
If you do need that 3sge TB you gave me, you are definitely welcome to it. I have have the st205 tb on mine, and already have a spare st185. You are also welcome to the 185 one if that helps.

For IAC, I would definitely use the Jeep one. It's early on, chances are at some point you'll want that output for something.

I'll take you up on the offer for the 185 TB. I have a couple more 3S-GE TBs lying around the garage, but from what I hear the 3S-GTE is a little easier to install.

I'm pretty sure I'll be using the Jeep IAC. Both for the reason you mention and because it's easier to use any TB with that setup.


04-26-2010, 10:12 PM
Well thanks to ChrisD I now have an ST185 throttle body. The bolt holes to mount the throttle body to the j-pipe (this is where the 4A-GZE throttle mounts to the intake system, it is pre-supercharger) line up perfectly. I will have to do a little work to the throttle body as there are vacuum lines that interfere. I'll also have to modify the the linkage for the cable. The only other issue that I can think off off the top of my head is that the bolt pattern for the intake on the other side of the TB is rotated ninety degrees and is a different size. This means that I will have to fabricate an intake pipe here. This might be enough to cause me to wait to put this on until after I install the MS as then I don't have to try and connect the intake to the MAF, but instead just an air filter.

Here are the two throttle bodies side by side:

Here is the stock 4A-GZE:

Here is the 3S-GTE:

And just for fun here is the 7M-GTE:

When I went to ChrisD's to get the throttle body we took a look at his car. As he stated he has the ST205 throttle body on his engine. He showed me the adapter he had made to mount it to his intake manifold. As far as I know the 7M-GTE and the ST205 throttle body have the same bolt pattern. So in the future I may try to get one of these adapters to mount the 7M-GTE TB.

In other news I decided to test the ignition. I am still using the stimulator to drive the tach inputs. The signal seems a little unstable when I turn the RPM up and down, but I suspect it has to do with the stimulator not producing a perfect signal. I will know for sure when I use the cam angle sensor.

I borrowed a car battery from the lab and connected it to a charger. I then used a fuse block and powered the MS through that. I tested everything again and then turned it all off. I connected the two coils of my wasted spark system. My initial plan was use some metal clips to test the spark. I would connect the clip to one of the posts of a coil and then rotate it so it was near the other post. This would create a gap for the spark to jump. It would be larger than a spark plug gap, but I wanted to test it like that because it would be a definate indication that the spark would be strong enought to spark any plug. I decided that before I started to let sparks fly that I would turn the unit on with the coils connected but without the clips t make sure nothing is shorting out.

This was fine in theory, but not what actually happened. It gave me a bit of a scare only because I wasn't expectng it. Here's a picture:

The spark was strong enough to jump the 1 1/8" gap between the posts! This is a much stronger spark than I ever thought I would get. I am very happy with this result. It should have no problem at all igniting a boosted engines combustion mixture.

My next move is to start looking at using the cam angle sensor driven by a drill to provide the tach signal.


11-09-2010, 05:14 PM
did say that progress was going to be slow...

Several things have held up this project. There was a vacation, a move, work, etc etc. However the biggest hold up was the fact that this is the only car I had to drive. That has now changed. I have been given a 1987 Oldsmobile and a 1977 Chevy truck. I have to decide which one I want to keep and scrap the other one. Then I can finally start to work on getting the Megasquirt into the car. I haven't even tried the cam position sensor yet.

A few things have changed with the car too. First I received a water injection system as a gift, so I installed that. It is a Cooling Mist brand unit and came with everything I needed to install (tank, pump, nozzle, activation switch). It took me a while to figure out where to mount the nozzle. With a supercharged car there is some debate. The idea of the water injection is to suck heat from the intake air by using that energy to convert the water mist to steam. For that reason it is best to mount it as close to the intake as possible. However, it has been said that injecting the water mist before a roots style supercharger actually helps seal the rotors against each other thus helping create boost. However in this case the cooling effect is less efficient. I decided to mount it in the neck that connects to the intake manifold post intercooler.
The way that the kit is supposed to work from the factory is that at a certain boost level the pump is activated and keeps going until the boost drops again. I didn't want to run the full current of the pump through the switch so I decided to use a relay to activate it. One of the things that you are supposed to avoid is running the pump dry. To avoid this one should check the water/alcohol level in the tank regularily. I decided to take this a little further and added a level switch. It is simply a float switch that when the water drops is makes contact. The instructions state that this switch should be used to light up an indicator light that you are out of water. I decided that I wanted to make sure that I wasn't running the pump dry and added another relay such that the pump will not turn on in the event that the water level is low as indicated by this switch. Finally, after going for a ride in ChrisD's car, I decided to add an indicator LED that lights up when the pump is on. The circuit I designed is this:
The blue light comes on when the pump is on and the orange one comes on when the water is low. I mounted the LEDs in the A-pillar:
Lastly I welded a frame together to mount the pump, tank, and relays to. This frame bolts to some of the holes that originally held the trunk liner in.
Yes the wiring is messy and needs to be redone, but this was just my first pass to see how it worked. It took some trial and error to adjust the switch. It wanted to come on as one or two psi, so I had to adjust that to about nine. Things worked pretty good after that.

After installing the water injection I realized that I didn't like the way that the gauge pod for my a-pillar fit. So I took it and the a-pillar off and glued them together. It then used fibreglass and body filler to make them appear as one peice. In my opinion this looks much better than before. Pictures taken with cell phone usually while paint still wet. The black is actually flat, and matches the rest of the black dash better than I thought it would.

After this I noticed that the clutch started to slip. So I needed one of those. I started to collect the parts. I have sitting on my workbench a clutch kit (aftermarket oem), a slave cylinder and a master cylinder. I also came to accept that the synchromesh on first and second is pretty much gone. So I started looking at tranny rebuilds. After realising that that would be too much money I started looking at doing it myself. After realising that that would be too much money just for the tools I started looking at used transmissions. The first one I looked at was the one that came with my spare engine out of the AE101. It is the heavier supercharger transmission and has an lsd. I'm not crazy about the heavy part but it is pretty much bulletproof. Unfortunately there really isn't a cost effective easy solution to get axles for it. So I gave up on that and started to call some local engine importers. The 4A-GE 20V is a popular swap for AE86's and the original 20V transmission doesn't get used, so there are a handfull of 20V transmissions sitting around. I found one and bought it for a good price. It is a silvertop transmission so the only thing I need to do to make it work in my car is drill a hole for the gear selector to come out the front of the transmission and press in a seal. The blacktop would need an adapter made.

Because I need this car to get to work and I've had a whole bunch of stuff going on I haven't had a chance to do any of this work yet. I am hoping that as soon as I get my other vehicle on the road I will be able to drop the engine and transmission out of the car, fix some of the notorious oil leaks (distributor, oil cooler hoses, k-box has everything needed for this), and wire in the megasquirt all at once.

Street tuning the megasquirt looks like it is a little easier now. The Tuner Studio software now has a VE Analyze Live function that will correct the VE table as you drive. It's like automatic tuning but you still have to keep an eye it and make sure it doesn't throw everything out of whack. The more difficult tuning with be the Spark Advance table. This is a little more obscure and I can't find as much information on it. This is also where power gains are to be made. I will probably break down and take it to a dyno to have it tuned. I will update as things happen.


11-12-2010, 09:09 PM
Another victory!

I pulled my spare instrument cluster off the shelf this morning and brought it into work with me. I decided that it was my goal at lunch today to have the Megasquirt drive the stock tach. I followed the information about driving a tach that was previously driven by a coil found here: http://www.msextra.com/doc/ms2extra/MS2-Extra_Hardware.htm#tachoout.

The Zeal daughterboard I have running the second VR conditioner and an output through has a tach driver on it as well. It is the first input/output. The Zeal board is designed for an RX7 where the tach requires a twelve volt input to trigger. I removed the 1Kohm resistor between the collector of the transistor and twelve volts and replaced it with a relay coil. Unfortunately the only one I had on hand was a sealed unit so it still had the switch part of the relay in it. This isn't a big deal except for the fact that the switch clicks whenever the coil is energized. I'll have to find a relay I can pull apart just for the coil.

I connected the output of the Zeal tach driver to the "IGN-" on the back of the instrument cluster. I then connected "IGN+" to twelve volts an "EARTH" to ground. I turned the tach input signals on the Stimulator to zero and turned everything on. It all powered up and the MS seemed to be happily talking to the computer. Then came the moment of truth: I turned the RPM signal up. The fuel pump and injector lights came on on the stimulator. The relay also started to make it's clicking noise. The needle on the tach moved. It moved up and down happily. It feels like it might be a little slow to respond, but I believe that it is behaving just like it would be if connected in the car as stock. I wanted to see how accurate it was so I set it to 5000RPM as read by the gauge and then went to look at it on the computer, which showed 5048RPM. Here are some pictures of the setup (bear in mind that this is just proof of concept and not final wiring):

That makes me pretty happy. Now I just have to stop dragging my feet and see if I can get the CAS working with the MS. I probably won't have a chance to do that until mid December. Work ramps up next week. On the plus side I have a little overtime comming my way.


11-24-2010, 06:06 PM
Well it's time for another update.

I took the 7M-GTE and started to investigate it a little further. It has three VR sensors in it. One sensor monitors the twenty four tooth wheel and the other two monitor the one tooth wheel 180 degrees apart. I started by removing one of the sensors on the one tooth wheel by simply removing the screws that hold it in. As this will be used for my test rig only I will not completely remove it, just leave it hanging outside the case. I then started to figure out the wires. Comming out of the connector are four wires, one for each of the sensor and one for a common ground. The black wire was the common ground, the blue wire was the twenty four tooth sensor, the red wire was the one tooth sensor that I left in and the yellow wire was the one tooth sensor I removed. I determined this by connecting the wires to my scope and spinning it by hand. I could see the signal on the scope jump when the one tooth passed the sensor on the red wire. The yellow wire had no signals as it was outside of the unit. And the blue wire jumped around alot indicating it had the multiple teeth.

I then mangaed to get the gear off the 7M-GTE cam position sensor. This allowed for me to easily put the shaft into my electric drill. I connected the wires to the connecting block I'm using for testing and soldered the appropriate wires to the connector. I then plugged it in and turned it on. I checked to make sure that the air intake temperature sensor and the cootant temperature sensor were still working. Both read twenty one degrees celsius so I was good to go. I then decided to connect a tachometer to make sure that the tach output driver circuit was working with the new tach input. I used an autometer tach so I wouldn't have to deal with a ticking relay.

Once it was all connected I started the drill.

It worked! It's still got some bugs to sort out but it works! I get a nice steady tach output when the drill is held at a contant speed. However if I'm increasing or deacreasing the speed it "glitches" a bit and the tach needle jumps around a bit. This isn't a big concern yet as I am using cheap wire to connect it instead of the recommended shielded cable. Also there are some adjustments that can be made on the VR input circuit. I will try both of these ideas before I start to scratch my head too much.

Here is a camera phone video I made of the experiment. You can see how I have the cam sensor in my drill and the tach is sitting on top of the power supply. The glitching is very obvious by watching the tach.

http://i150.photobucket.com/albums/s117/Berg9987/Megasquirt/th_MOV20101124_001.jpg (http://s150.photobucket.com/albums/s117/Berg9987/Megasquirt/?action=view&current=MOV20101124_001.mp4)

By the way, I found out my drill has a maximum RPM of 5000.


11-25-2010, 08:04 PM
Well I used Tuner Studio's composite logger and took a look at the signals it is getting for the trigger events on the cam position sensor. The signal from the twenty four tooth wheel is good. The signal from the one tooth wheel, not so much.

On the data logging screen there should be twenty four pulses on the main wheel to one pulse on the secondary. Zooming in and counting the pulses I've been able to determine that I am getting the one pulse on the second wheel for every twenty four pulse on the primary wheel. However, I am getting extra pulses in between these. This throws the timing off and from what I can tell is causing the fluctuating revs.

I posted on the megasquirt forum and have been told that this is a symptom of electronic noise. Because the one tooth wheel spends so much time in a non triggered state it is more prone to noise than the other wheel. The noise could be caused by a number of things, equipment around the wires, the wires themselves being insufficient, the drill, etc. So I have a few things to check and to try.

Here is a screen capture from the composite logger showing the extra pulse:


11-25-2010, 10:33 PM
So I followed some of the advice from the MS forum and installed a jumper on the Zeal daughterboard that connected one of the pins of the conditioner chip to ground. The tach now moves smoothly throughout the rev range of the drill, with the exception of below 1000 rpm. It is still a little jumpy there.

I checked the composite logger again and found that I am still getting too many triggers on the one tooth wheel, however they are more consistent in that they happen every eleven to thirteen pulse on the twenty four tooth wheel. In some spots it cleaned right up to where it should be, so I'm getting closer all the time.

Here is a pic of what it looks like when its behaving:

And here's a question: Is anybody still interested in this? I'm feeling a little lonely in this thread.


12-17-2010, 04:12 AM
Just read your thread - lots of great info here as far as i am concerned.
Not sure if i ever want to mess with a Megasquirt myself but you've certainly done the work to make it run correctly.
I am contemplating going to water injection, possibly a WTA I/C & found that info you posted very valuable.

Looking forward to seeing the car up & running & seeing what dyno numbers you are able to achieve.
Are you running a larger "cross" pulley or the smaller "star" pulley on the s/c?
Are you running a larger diameter crank pulley? Upgraded I/C or larger I/C piping?

My husband & I dyno'ed our swapped 4agze in 100*F weather (summer in Tx) & got 157rwhp & 145rwtq with the following mods:
Cusco 175mm crank pulley
Smaller AE101 "star" pulley SC12
2.25" diameter chrome I/C piping
Ebay China special bar & plate A2A I/C
Ceramic coated 4-1 custom made header
2.25" b-pipe to 2.5" MKII Tanabe exhaust
Apex-i Air Filter w/2.25" piping
E51 S/C tranny w/TRD 1.5 way clutch-type LSD

Like I said, planning on possible WTA I/C upgrade + water/alky injection
I have a set of Toda 256* camshafts, looking for a set of adjustable cam gears & thinking of going shim under buckets.
I also just got the stock 4agze TB bored out, didn't want to fool with going to the 3sgte TB.
Planning on tuning on the dyno to see how far we can get on the stock ECU + the SC12.

Looking forward to updates on your project. I don't post on mr2oc.com except to sell things or buy things... too much arguing & very few people with any knowledge left over there these days.

12-17-2010, 04:53 PM
Thanks for the kind reply!

I am running the smaller pulley on the supercharger. The supercharger that is currently on the car is the one off of my late model 4A-GZE out of the Levin that came with the smaller pulley. I have an NST crank pulley driving the supercharger. I also have the ABV blocked off and haven't seen any negative effects of doing so yet. I used the 4 to 2 header that came with the late model 4A-GZE to build my own exhaust:

The water injection system was easy to install and everyone I have talked to who has one have had only good things to say about them. Many times the benefits seem to come not from the water injection cooling the intake charge, but simply by it cleaning out the inside of the engine. Especially on older engines it seems to soften the carbon buildups and with repeated use it can remove them. I'm biased, but I would say go for it!

Your HP numbers sound like they are about what I expect to the wheels of my car. (Ok, your's are probably higher than mine :cheesy:) And I'm jealous of your LSD.

I don't know if I mentioned it but I have the water to air intercooler system from an ST185 I'm planning to fit. The system has an upgraded radiator for the front of the car. It was formerly used on ChrisD's Celica. As for the IC itself, I will likely have to cut the in and out ports off and weld on new ones, otherwise it would take up too much room in the engine bay. Every single instance I have read about changing the A2A intercooler to an A2W in an AW11 have all shown a decrease in air intake temperature.

I could go on and on about the "plans" for my car, but I'd rather keep it real and discuss only the parts that I currently have. When I switch to a turbocharger I will be using a JDM big port 4A-GE intake manifold (it's exactly the same as the US except it doesn't have the hole for the EGR) and I hope to get an adapter made to put on the 7M-GTE TB.

About the MS:
I picked up some shielded cable from the local electronics shop in the hopes of cleaning up the CAS input signal. The shop only had three wire shielded wire so I will try running both signals in the single shield and also each signal in their own shield wire. If I can get away with both in one it would make for a much neater install. I have also started mounting everything on a plank of MDF to make everything neater for on the bench testing (got sick of having wires running all over my bench). I'm not quite done mounting it yet. I will probably be taking a break from this until next year as my holiday's start tomorrow and I plan on spending as much time with my family and friends as possible (cars are cool, but they're not people). That and with dinner's at my parents and my girlfriend's family's houses I'll be too full to move lol.


12-17-2010, 05:27 PM
Correction, you have the ST205 IC. ;)

Just catching up on your posts since I was out on vacation through November.

I have always wanted a low level warning for my WI. I should pick up a float level switch and install that this winter too. Hmm. I will have the wiring apart anyway to clean it up, so maybe. I forget - was that switch included in your kit or did you pull it from a junkyard car? Have to figure out where to get a switch...although I'm sure even new ones are cheap.

Great thread Kevin - awesome documentation on the MS system. I think the reason not many responses is b/c there aren't a whole bunch of people here running that, so it is a bit foreign to many. Also electronics scares a lot of people....yet to others, it's just another thing to do. :D

12-17-2010, 08:55 PM
Correction, you have the ST205 IC. ;)

I knew that. I don't know why I wrote 185. I'm blaming this being my last day in the office for the year (can you tell I'm working hard?).

The float switch I am using came as an accessory from coolingmist.com: http://www.coolingmist.com/detailmain.aspx?pid=float01
You could probably find one for much cheaper at the wreckers, any car with a low windshield washer fluid warning light will have one. At this time of year I'm not sure I would try and order something to be shipped. Other things have them too, maybe an acquarium store would have on? I know that my coffee maker has one. If you are using a circuit similar to the one I described above you can pretty much use any switch as very little current is passed through it.

Maybe I'm strange, I think electronics are fun!

We still haven't managed to find time for that beer.....


12-20-2010, 10:01 PM
We should definitely meet up for that beer. I know Steve, Tara and Ryan were hoping to get together some night for a drink at 1410 during the Christmas week. Maybe then?

(also, I could bring you your tool)

12-23-2010, 04:56 AM
Do you have any more pics of the circuit and the wiring for the low water and LEDs for the Water Injection? That looks awesome. I have dabbled with electronics and all that jazz, but I am never really built a circuit for anything.

03-16-2011, 07:37 PM
Do you have any more pics of the circuit and the wiring for the low water and LEDs for the Water Injection? That looks awesome. I have dabbled with electronics and all that jazz, but I am never really built a circuit for anything.

Sorry for such a late reply. I'm atually in the process of designing a circuit board for the water injection system. The idea is to mount it in a box and have it as a self contained unit. I don't know when I'll have it built as I keep comming up with ideas, but when I do I have to have a minimum number of around ten boards printed from the local electronics company, so I will probably have some for sale. The circuit is pretty much identical to the one already posted. Here is my initial thought for a circuit board layout (quck ten minute jobbie):

Well it’s been a while, but I am still working on this. I grew tired of having all the parts scattered about on my work bench so I mounted everything to a piece of MDF. The MS sits at the bottom in the middle, to the right is a connection block where I can bring out the wires to the thirty seven pin connector. In the top right are the two temperature sensors and the throttle position sensor. At the top is the power supply input and fuse panel. To the left is the distributorless ignition system. I currently have the ignition connected straight to the fuse panel. I have not installed a fuse yet so it hasn’t been fired up. I will not running ignition power like this. I will be installing a relay activated by the fuel pump output of the MS so as to avoid overcharging and burning out the coils (the MS turns off the fuel pump after a few second of no RPM signal).

In the second picture you can see a modification I made to the case of the MS. I have put in a DB15 connector. The reason I did this was to be able to run the ignition outputs through the other side of the MS away from the signal inputs and I have simply run out of room on the DB37 connector. I used a dremel and a file to make the hole the right shape and then mounted the connector with some screw in hardware.

You’ll notice that I mounted the connector on the inside of the case. I did this because the connector will be soldered to wires inside the MS and this way I can take the case apart for trouble shooting without having to desolder or cut these wires.

I have shielded wire connected to the CAS now. I also removed the unused VR sensor completely. I still have false triggers on the one tooth wheel, although now it only happens at higher RPM. The shielded wire I have is the three wire stuff (one wire for each sensor and one common ground). The shielding is grounded at the MS only. It seems that the MS is ignoring the false triggers as the RPM reading in tuner studio seems smooth and happy. There is some noise filtering capability in the software, but I haven’t played with that yet. I haven’t read enough about it yet but apparently I can tell it to only look at the on tooth trigger that lines up with every twenty fourth trigger on the other wheel. This is more of a band aid solution and I would rather stop the problem at the source. I still have to try running each signal in a separate run of shielded cable and it may be my drill that is making the noise. Unfortunately I haven’t access to a pneumatic drill at the moment. I was going to try a cordless, but it didn’t have any charge in the battery.

Another problem I have right now that has me a little concerned is that the map sensor is reading low. According to the weather website the pressure in my area should be about 99.5 kPa. My MS is reading 87 kPa.

I am using the Mapdaddy 4 bar Map senso with barometric correction found here: http://www.diyautotune.com/catalog/mapdaddy-bar-map-sensor-with-barometric-correction-p-117.html

I have it set up as per the instructions. Their instructions (in Megatune):

My setup (in Tuner Studio):

So I don’t know what is going on there. I will post up over at the MS forums and see is someone there has some advice. I might be able to put in my own values in the settings to correct it, I might have a bum sensor, or I might have a problem in the circuitry. I don’t think the sensor is bad. I was able to find some hose that fit loosely over the MAP connection and by blowing in it I was able to get it up to 90kPa, so the reading does move.

If anyone has any thoughts I’d be happy to hear them.


03-16-2011, 07:40 PM
After posting on the MS forum I took a measurement of the voltage output of the MAP sensor. It was a little low at 0.998 Volts. One of the owners of DIYautotune replied to my post saying that it sounds like a faulty sensor. So I am going to need a new one.


03-17-2011, 01:35 AM
All of your pictures disappeared =[

03-17-2011, 02:59 AM
Oops! I organized my photobucket album and I guess I must have broken the links. I still have all the pictures, give me a day or two to fix all the links.



03-17-2011, 03:21 AM
Ok, so it only took a few minutes to fix it. All the pictures should be working again, let me know if they are not. My apologies for the inconvenience.


03-17-2011, 04:04 AM
Very interesting thread.

Your knowledge of electronics is way above my head but im trying to follow.

Very excited to see your car running on the MS

03-17-2011, 08:44 PM
Very interesting thread.

Your knowledge of electronics is way above my head but im trying to follow.

Very excited to see your car running on the MS

Thanks me too! Electronics started as a hobby for me when I was a kid. I would get various kits for Christmas' and birthdays. Things like the "build a crystal AM radio that fits in a Tic Tac box" and the like. In high school I started to design some of my own circuits to impress my friends. It mainly revolved around circuits to control lights to correspond to music, making my parent's rec room look like a disco. Eventually I found my way to a technical school where I got my diploma in Electronics Engineering Technology. Since then electronics have been a part of my job.

Hooray for diyautotune! I emailed them with my MAP sensor issue and they are going to replce it under warranty, even though my warranty technically expired nine days ago. They are great to deal with. I highly recommend!


03-17-2011, 09:30 PM
Last post reminded me of this, read it long time ago, kinda goofy.


03-19-2011, 02:21 AM
Well I feel stupid now, but I'm man enough to admit my mistakes. So here goes:

I was comparing my MAP sensor readings to the local weather reports. The weather station was reporting about twelve to fourteen kPa higher than what I was reading. I was stumped and posted on the MS forum and it was thought that the sensor my have been out of spec. I removed it from the MS today and found the sensor that came with the MS kit. After I installed that I set the sensor calibration to that sensor and tried again. I got the same reading (around 88-89 kPa). So I tried the other sensor on the MapDaddy (used for barometric corrections) and again got 88-89 kPa. So I went through the MS and checked the rest of the MAP input circuit (two capacitors C2 and C3 and one resistor R2). They all checked out fine. So now I knew that the sensor was good (or I had three bad ones-unlikely) and that the rest of the circuit seemed to be good. So I was stumped.

I started looking for more accurate pressure readings in Calgary. Everything I found was around 101 kPa. I did find one that was only forcasting that listed the next few days at around 660mmHg. Convert that to kPa and you get around 88 kPa, which is pretty much what I was reading. So now I was thoroughly confused. I checked back with the megamanual (http://www.megamanual.com/v22manual/mtune.htm) and it listed Calgary as being around 89.15 kPa.

Wikipedia actually gave me the answer:

Mean sea level pressure (MSLP) is the pressure at sea level or (when measured at a given elevation on land) the station pressure reduced to sea level assuming an isothermal layer at the station temperature.

This is the pressure normally given in weather reports on radio, television, and newspapers or on the Internet. When barometers in the home are set to match the local weather reports, they measure pressure reduced to sea level, not the actual local atmospheric pressure. See Altimeter (barometer vs. absolute)

So the internet weather reports are all corrected to see level! Everything seems to be working fine. Boy do I feel silly.


Ah crap, its even in the manual: "Note that weather reports usually report the barometric pressure 'corrected' to read as if 101.3 kPa was the 'normal' for your elevation. Do not expect these reports to correspond to what you get on MegaSquirtģ unless you are at sea level."

03-19-2011, 02:22 AM
Last post reminded me of this, read it long time ago, kinda goofy.


It's an oldy but a goody!

06-27-2011, 04:17 PM
Long time no update.

With work sending me out of the country and the weather finally turning nice I havenít had much time to work on this project. However things are still progressing.

I am still working on some circuit designs for the interface between the car/engine/Megasquirt. A relay board is offered by the same people who sell Megasquirts, however I want mine to be more ďpurpose builtĒ for my car. Iíll update on those when I am a little closer to the final design.

One thing that has bothered me about my setup is the fact that I have the coils from a GM distributorless ignition from a V6 for my ignition. My first thought was that I would be able to cut the extra coil part off. This would leave me with an edge I would have to recover because of a contact with 12V on it.

Yesterday while I was at the local auto wreckers I decided to look for a four cylinder engine with these coils. I found one in a late nineties Chevy Cavalier. Getting it out was a little ridiculous. The four spark plugs were on the exhaust side of the engine towards the front of the car. I followed the plug wires over the cam cover and down between the first and second intake manifold runners. From there I couldnít see where they went but by reaching down and feeling around I determined that the coils were mounted to the back of the engine. It looks like it would be easy to get to them if the car was up on a hoist, but it was raining and there was a six inch deep puddle under the car that I wasnít terrible enthused to crawl around in. I only know how deep it is because I dropped a socket in there that I had to retrieve.

I donít know who came up with this engine design, but it obviously wasnít designed with serviceability in mind. To get the coils I ended up taking the alternator off to loosen the drive belt so I could remove the power steering pump. Then I took the intake manifold plenum off and then the intake runners. From there I had a clear shot to the coils (provided I had my knees on the front bumper and was lying across the engine.

After all that I had this:

Notice that it only has four wires connected to the coils. This is a much neater package than the V6 unit. I havenít had a chance to pull it apart yet but if Iím lucky then it only has power, ground, and direct connections to the coils. The V6 unit had a bunch of circuitry inside for dwell control etc that I removed to install my coil drivers. If the I4 unit doesnít have any of the circuitry inside I can connect my coil drivers up to the wires that are direct connections to the coils and Iím laughing. I will have to built a box to house the drivers though. Iíll probably use some aluminum heat sink material I have at home.


06-28-2011, 03:37 AM
I pulled the new ignition apart and found that it does have circuitry inside. This is a good news/bad news scenario. The bad news is that now I have to pry another one of these apart and try to remove that goop that is impossible to work with. The good news is that I get to install the drivers right into this unit so it will be a nice neat little package. The mounting plate for it has four holes tapped for the bolts that hold the coils down. It should be very simple to make my own mounting plate for a neat install.

The next item on the list are to modify the MS2 slightly. I want to redo the wiring from the main board to the Zeal daughter board. I also want to do the wiring modifications to get table switching and launch/flat shift control. The table switching will be activated by the water injection so if I run out of water or am not using it for some reason it will run a more concervative map. The launch/flat shift won't do much for me until I have a turbo install and I feel the need to drag race. Still, I'd like to have it there. There is a switch on the clutch that was originally tied into the cruise control that I will use for launch activation (through another toggle switch so I can turn it on and off of course).


08-10-2011, 07:21 PM
I haven’t had much time to work on this with so much work needing to be done around the house and other projects. Plus I have had several work related trips and a few more coming up.

Anyway, since I had the tachometer drive circuit working and an Autometer tach in the car I decided to button up the wiring for the daughter card in the MS. If you look at my previous posts you can see how I ran the wires for the tach to the daughtercard. These were the orange wires between the two boards. I wanted to make things as neat and tidy as I could so I used another piece of ribbon cable and redid all my wiring to the card. I now have ten connections to this card and it looks really clean:
And this is what the back of the daughtercard looks like:

What isn't pictured is that I populated another one of the general outputs on the Zeal daughtercard. This was done just to future proof the project. Since I'm interested in swapping the supercharger for a turbo there will come a time when I no longer need the custom supercharger circuit. When that happens I can use the output I'm using for the supercharger for something else and I will likely have no need for the turn off delay. So when that time comes I can just remove my supercharger circuit and move its activation output to the Zeal. I will probably remove the ribbon cable again and put in a twelve wire cable when that happens. As for now, I knew where the parts were and I can't say for sure I will later down the road so I installed them instead of losing them.

I’ve been doing some thinking and I want to use a couple more features of the MSII. One is the Launch/Flat Shift control and one is the Table Switching. The Launch/Flat Shift is mainly used for drag racing to get consistent launches and to build boost with turbo cars. The flat shift allows for shifting with the gas pedal to the floor.

The table switching is used to switch between two ignition and fuel tables. This could be used if you had two different fuels. I will use it with my water injection system. My water injection already sends a ground signal when it is turned on to light an LED on the a-pillar. I will use this signal to trigger table switching in the MS to run a little more aggressively when the water is being injected. With my setup it doesn’t even try to inject if the water level is low so when there is not water it will run more conservative tables. This will require more time to tune in the end but I think it will be worth it.

In order to run these two option some modifications need to be made to the MSII. The information on that can be found here:

I started these mods but I don’t have wire thin enough to solder directly to the processor. What I have done is remove R1:

And installed the jumper for the Launch Control:

I still need the four diodes and two resistors to build the two active low inputs to use these functions. Next time I’m at the electronics store I will pick them up along with some thin wire.


08-10-2011, 07:26 PM
PS: I should mention that after all the features I've decided to make use of, the DB37 connector is pretty much full. The only pin I have left for an input/output is pin 36 (originally intended to be the Ignition output). So I can bring in either the Launc/Flatshift or Table Switching through the DB37. The other one I will bring in through the DB15 I added for the coil drivers. If I had decided not to use the stepper motor idle air control I could have another three connections on the DB37.

08-10-2011, 09:19 PM
Wow. Very nice and clean work on the MS. I just hope to one day have half of your knowledge of the MS. Impressive.

08-10-2011, 09:36 PM
Thanks! I can't claim that any of the knowledge is actually mine though. Everything I'm doing has been done before, it's just a question of spending time searching. Feel free to ask me questions if you think I haven't covered something properly.

I spent my lunch today at the local speed shop. Spent a little more than I usally do:

I was debating whether I should put this in this thread or my gauges thread. One can not tune using a narrowband A/F gauge so I got this wideband. It is Autometer (PN 4378) so it matches the rest of my gauges (car no longer has any non autometer gauges in it). I was considering getting the the analogue one to completely match the rest of the gauges but decided on this instead. The reason I decided on this is that I plan to do some tuning on the street before booking dyno time. If I do something wrong with my VE tables then it will be easier to see on this gauge than the analogue one. This one also has a warning function (provided I have a WOT switch installed). This unit will output a 0-4V signal that can be interpretted by the MS for a wideband EGO input. I can customize the bar graph display around the outside to work with my engine specifically. The number on the display will always show what the A/F ratio is but if the lowest my car ever goes is 11.5 I can set the low value to 11, and similarily for the highest value. That will keep the bar graph display relevant at all times with the best resolution. However this will also change the output to the MS. It's not a big deal, I just have to go into my EGO settings in the MS and change it from "Autometer 0-4" to custom and define the low point (0V) and high point (4V) to correspond with the A/F values on the gauge.

The warning function can be set up to flash all the LEDs if the mixture gets too lean. I can adjust the point at which it does this. It should do well to catch my attention. The warning function is designed to be connected through a WOT switch that goes to ground at WOT. Instead I will likely just wire this to ground without a switch so it will always flash if I get too lean.

And obviously I'm going to have to extend the wires to the O2 sensor. It's just one of those things that MR2 owners get used to.


08-11-2011, 09:40 PM
Small update.

I found some smaller gauge wire:

That completes the modifications I have to do to the MSII daughtercard. Now I just need those four diodes and two pull up resistors and I should be done with the insides of the Megasquirt. Then it's on to final bench testing and finally getting it wired into the car.

I think I've decided to use the one programable output I currently have to run the engine bay fan. Right now it has its own little control computer and temperature sensor but I have this bypassed so it is on all the time. This works well in summer but when the temperature drops it's harder for the engine to get up to operating temperature. I will likely use an OR argument with both intake temp and coolant temp to turn it on.

- Modify two coil GM wasted spark coils for use with Bosch coil drivers
- Acquire Jeep Stepper IAC
- Acquire housing for IAC (diyautotune sells these: http://www.diyautotune.com/catalog/custom-idle-air-control-body-for-use-with-jeep-40l-iac-valve-p-69.html)
- Modify 4A-GZE CAS to one VR sensor per wheel
- Install wideband O2 sensor
- Wire in Megasquirt with custom fuses and relays (prototype stage)
- Get car running and conservatively street tuned
- Redo fuses and relays as custom circuit boards (after configuration has proven to work)
- Get car dyno tuned

I'm looking forward to just getting it idling. Every success story I've heard states that they've never had the car idle so smooth on stock electronics after getting the MS installed.


08-11-2011, 10:28 PM
This is a list of common pin uses for various inputs and outputs on the MSII running the extra code (from here: http://www.msextra.com/doc/ms2extra/MS2-Extra_Hardware.htm#mods_daughter):

Here is what I am using them for (I put the letter of the column corresponding to the use I'm using the pads for as per the table above):
JS0 - A - Stepper Motor
JS1 - A - Stepper Motor
JS2 - A - Stepper Motor
JS3 - A - Stepper Motor
JS4 - B? - Not yet used
JS5 - A - Constant Barometric Correction
PIN 15 - F - Table Switching Input
JS7 - C -Launch/Flat Shift Input
JS10 - A - Single Tooth Wheel VR Sensor Input
JS11 - D - Supercharger Engagement Output (My circuit is on the first page)
D14 - A - Spark A - Coil connected to cylinders 1 and 4
D15 - B - General On/Off Output - Most likely engine bay fan
D16 - A - Spark B - Coil connected to cylinders 2 and 3
FIDLE - B -Tachometer Output

I've rethought some of my wiring. Since both the table switching and launch control inputs are both solid on/off signal as opposed to a more sensative input it should be safe to bring them both in through the DB15 conntector with the ignition outputs. Again this will leave me with one pin on the DB37 that it unused. Can't have that. So I'm thinking it might be an idea to bring the knock sensor signal through there. I can try and get the stock one to work or I can try an aftermarket unit. There is even one that I know of that is designed for the MS.


08-12-2011, 12:58 AM
Using my advanced MS Paint abilities I modified an existing MSII external wiring diagram to show how my unit will be wired into the car:

Please note that both the supercharger relay and the fan relay need fuses. Pin 1 is at the top left of the DB37 looking at the socket on the MSII from ouside the case. It has 1 through 19 on the top row and 20 through to 37 on the bottom. Here is a list of what each one is connected to:

1 - Intake Air Temperature Sensor Ground
2 - Coolant Temperature Sensor Ground
3 - VR Sensors Ground
4 - VR Sensor on the One Tooth Wheel
5 - Supercharger Relay Activation
6 - Fan Relay Activation
7 - Wideband O2 Ground
8 - Throttle Position Sensor Ground
9 - Not Connected
10 - Not Connected
11 - Not Connected
12 - Not Connected
13 - Shielding for VR Sensor Cable
14 - Engine Ground
15 - Engine Ground
16 - Engine Ground
17 - Engine Ground
18 - Engine Ground
19 - Engine Ground
20 - Intake Air Temperature Sensor Signal
21 - Coolant Temperature Sensor Signal
22 - Throttle Position Sensor Signal
23 - Wideband O2 Signal
24 - VR Sensor on the Twenty Four Tooth Wheel
25 - Idle Air Control Stepper Motor
26 - Throttle Position Sensor Five Volt Supply
27 - Idle Air Control Stepper Motor
28 - Twelve Volt Power In
29 - Idle Air Control Stepper Motor
30 - Tachometer Output
31 - Idle Air Control Stepper Motor
32 - Injector Bank One
33 - Injector Bank One
34 - Injector Bank Two
35 - Injector Bank Two
36 - Extra Pin, Possibly for Knock Input
37 - Fuel Pump Relay Activation


- I may not need to use seperate grounds for the VR sensor ground and shielding. I'm not sure if it is best to keep them seperate or to use the shield itself as a ground.

- All the pins that have "Not Connected" next to them are grounded inside the MS. It is how the board is printed and I can not change that.

- The Autometer wideband has two outputs to go to the ECU. One is positive and one is negative. The negative one will go to pin seven. http://www.autometer.com/productPDF/1143A.pdf

- The DB15 connector has the activation lines for Launch and Table Switching through it.

- The DB15 connector has the outputs of Spark A and Spark B going to the coils. The Bosch ignition drivers are mounted below the coils and have a seperate ground to the engine to avoid high current travelling through the MS.

- The fuel pump relay will activate more than just the fuel pump. This will ensure that when the key is in the ON position but the engine isn't running that none of the devices connected to it will run. Hopefully this will avoid things like burnt out coils.

- Pin thirty six is an extra pin and at the moment not connected to anything inside the MS. This was originally intended for use with the single ignition coil drive inside the box. I did not install this driver as my intent was to go wasted spark from the beginning. To use this as a knock input I will have to install a jumper wire from the IGN output on the board to the JS4 pad thus connecting pin thirty six to JS4.


08-13-2011, 03:15 AM
My girlfriend decided to spend some time with one of her friends this evening so I spent some time with the MS. IT'S DONE! :biggrin::bigsmile:

I had to go to the electronics store for work so while I was there I picked up the resistors and diodes that I needed. I then built the input circuits to be used with an active low:

This circuit diagram is from the MS2-Extra link that I put up a few posts ago.

This circuit is simple and fairly standard in electronics. The MSII chip uses what is called transistor transistor logic (TTL). This means that it has set rules for inputs and outputs. They must be digital, either on or off, 1 or 0, in this case five volts or ground. TTL chips with expected logic inputs do not do well when these inputs are left floating (not 1 or 0). That is why there is the five volts and the resistor there. The resistor acts as a current limiter and when the external switch is open the input to the chip "sees" five volts. This is called a "pull up resistor". When the switch is closed it grounds the pin and the pin side of the resistor so the chip "sees" a ground. If the resistor wasn't in place when the switch closed the five volt supply would be shorted to ground and cause damage. The diode on the left makes the switch one way, so if anything other than a ground gets onto that input it wouldn't affect the chip (within reason). The diode from ground to the input is there for protection.

All the iternal wiring and modifications are done. I put the case back together and hopefully won't have to open it up again until I start the install. Then I might have to open it for trouble shooting and/or adjusting the pots on the VR sensor input.

Next up: Getting the GM DIS coils running.

For now though, there is a beer in the fridge calling my name! If you're reading this, have a good weekend!


Edit: I should mention that I tested these inputs before closing the box. I put two indicators on the screen in tuner studio. I grounded the inputs and nothing happened. I checked the circuit and everything was fine. It turned out to be that the launch doesn't work unless the throttle value is above zero. Also the table switching requires a tach input before it is activated. It worked fine once I connected the jimstim and gave it the inputs it needed.

08-16-2011, 04:44 AM
Tonight was probably the last chance I had to work on this project until mid September. So I decided that I wanted to get the ignition module working.

I started with the two coil module that I got at the auto recyclers and pictured in a previous post. I removed the coils and saw this:

After spending some time with a pick and small screwdriver I managed to get the metal cover off. I wasnít careful enough for it to be reusable so it got thrown out:

This is the stock circuitry for controlling dwell and other such timing operations. Iím not using any of it, so I cut all the connections to the thick contact plates. I then used a long thin screwdriver to pry the back plate from the plastic part:

I used the same screwdriver to pry the circuit boards off the metal then cleaned it all up with a wire wheel in my dremel:

There isnít as much space in this one as the three coil unit. I still managed to drill two holes and fit my ignition drivers (transistors) inĒ

I checked all the pins for continuity and shorts and then plugged it in to the MS without the coils. I used a scope to ensure that the output looked alright on both drivers using the Jimstim to drive the MS. It all looked good so I plugged the coils in and it sparks!

To finish this part I have to figure out where/how I will mount it to the engine and build a bracket to do so. I also have to re-seal the base to the plastic part with silicone. I will also have to cut some metal or lexan and silicone that over top of the drivers to make the unit water tight again. Other than that the ignition module is finished!


08-17-2011, 03:34 PM
And the award for 'most technical project thread on CelicaTech' goes to... :P

I wish I could wrap my head around this stuff like you.. I feel like an idiot reading your posts.

08-17-2011, 03:56 PM
Thanks, I like awards!:laugh:

Seriously though, me doing this is nothing next to your ability to shape and weld metal. The body work in your thread is nothing short of amazing. If I could do that all the rust on my MR2 would be gone...

I'm a little disappointed that I have to leave town for two weeks on Monday for work. I'll be back for the long weekend but I suspect that I will be busy with things around the house (always seems to happen when I'm away for a while). Then I'm out of town for another two weeks on another job. There is nothing more frustrating that being ready to do something and not have the time :wiggle:


08-17-2011, 10:57 PM
I love this thread. This is probably the most well documented MS build thread I have seen. Cant wait to see the end product.

01-19-2012, 10:23 PM
I'm still alive!

I haven't had much time to work on this as my job required a significant amount of my time over the past few months, I decided to spend the holiday season with friends and family, and any time I have had to work on the car I spent on getting the winter beater up to snuff. Also I no longer have a garage available to leave a car in for days (weeks) at a time so with the temperature outside hanging around -30C doing any work outside would probably be a bad idea.

What I have done is started to look at how I would connect everything to the engine/car. There is a relay board made specifically for this which is sold by the same people who sell the megasquirt kits. Info can be found here: http://www.bgsoflex.com/mspower/mspower.html. This would work very well for most installations and can be made to work with mine, but I want something that is actually build for my unique pin out. I've started to sketch on paper where I want connections to go and will soon be putting it into a circuit making program.

I am planning on running all the connections to my custom relay board with the exception of the VR sensors inputs and ground. I will run those straight to the cam position sensor in the shielded cable in the hopes of keeping noise out of the rest of the inputs/outputs. I will be able to have pins 14 through 19 connect to a single trace on the board and use one wire of a heavier gauge from the board to go to the engine ground making everything a little neater. My board will also have four relays on it instead of three.

The fifteen pin connector will not be connected to the relay board. Instead it will be directly connected to the coil drivers, the launch control switch and the water injection control board.

I still don't have as much time as I would like to work on this, but I will keep you guys posted of any progress.


01-21-2012, 12:25 AM
Well I threw a quick design together for my relay board this afternoon. It is heavily influenced by the official relay board. Mine has four relays instead of the usual three, I am using two of them for the supercharger and for the engine bay fan control. The other two are the main relay that is turned on by the ignition that then powers the megasquirt as well as all the other relays, and the fuel pump relay which is used to send power to the fuel pump, the ignition coils and the injectors. I did this because the megasquirt will initially turn on the fuel pump when the ignition is turned on but will then turn it of if the engine isn't spinning. If I connected the ignition coils to twelve volts when the key is in the run position I run the risk of powering the coils for much longer than it is designed for and burning it up. If I had a short in the injector control wiring if I had the injectors on all the time in the run position I could flood my engine.

Each item that takes power off this board has a fuse to protect it. The official relay board has a polyfuse to protect the MS, but I decided to use an actual fuse. There are two other polyfuses on the official board and they protect the fast idle and vref (five volts to the throttle position sensor). I don't have a fast idle solenoid (I will be using a stepper motor instead), but I did put the polyfuse in on the vref. All of my relays except the main one have kick back diodes installed to protect the MS from any spikes during switching. I also have two extra pins and a diode installed to give me the ability to add a switch or jumper that will allow me to turn on the fuel pump without turning on anything else (so long as the key is in the run position). This was added in the event that I want to pump all the fuel out of the tank or if I get an aftermarket FPR I could set it without running the engine.

I plan to use a DB37 connector like the one on the MS to connect this board to the MS, and then solder teh wires to this board and use weatherpack connectors to connected the harness to the engine. The thinking behind this is that it should reduce the vibration seen on the DB37 connectors as the relay board and MS will be mounted to the chassis.

Here is my first rough draft of the schematic:

Until I order all the parts I need from digikey I won't be able to make a proper board layout as I need to confirm part dimensions. Having said that I will make a rough draft of the PCB next week sometime.

Having done this I'm still not sure if I will go with a relay board. I could just as easily do this with a couple fuse blocks and four relay holders, I just think that a board would make it look neater.


01-26-2012, 05:21 PM
So I took a look at digikey.ca to start pricing things out. I'm not buying enough of any of the items I need to qualify for a bulk discount. So the DB-37 connector costs $7.63, the relay sockets cost $3.18 each and the fuse holders cost $2.899 each. So for four relays, eight fuse holders and one connector the total price comes to $43.54. Considering that a relay board kit only costs $64.00 I would recommend that if you aren't doing unique things like me and can use the relay board kit, do it, it will be less expensive.

Once I have the parts in hand I can start making a board layout.


03-20-2012, 04:14 AM
Well time for a long overdue update.

I am still waiting on some electronics parts and until I have them I don't want to finalize my relay board. Also my daily driver is acting up so I'm currently driving the MR2 to work. That means I can't rip out the engine harness at the moment.

So I turned my attention to some of the other bits I'm going to need to get the engine running on Megasquirt. I decided to revisit the ignition coils.

I am happy with the way I have the drivers mounted below the coils as shown in a previous post. Now I just need to find a way to mount them to the engine. I would rather have them mounted to the engine than on one of the firewalls just to keep the plug wires shorter and neater.

The stock coils on the spare 4A-GZE from the AE101 are mounted to the cylinder head by a bolt above the exhaust manifold and a stud on the coolant neck by cylinder four. I mocked this up on my spare big port head:

Excuse the mess, basement renovations are happening and contractors have taken over the garage.

Now using these coils would be simple and would no doubt do everything I need them to do. However there are few reasons I have decided not to use these. First off, I would have to build a box to house my drivers much like the ignitor box on the stock car. Second, I don't like the way that the wires plug into them. They require a special boot to clip them in. Third, it is heavy! And finally, I can find hundreds of spare GM coils at the pick n pull. They are cheap and available. Since I've not tried something like this before I'm sure at some point I will burn up a coil or two.

So I started by tacking some scrap metal together to make a frame for the new coils:

And then spent a few hours welding and grinding and I had this:

Now at this point all that is left is to drill some holes, run a tap through four of them and then clean it up and mount it. It is also at this point I realized that I do not like what I have built and think it looks like crap.

So this little bracket is getting junked. Put this in the "I learned something box" and then make a better one.

First of, someone with my welding skills should do as few welds as possible. So making this frame out of five bits of angle iron and flat stock was a bad idea from the start. That and I did it free hand.

I will probably try again in a week. This time I will use a larger piece of flat stock that I can cut the entire mount out of. The only part that I will need to weld will be the part to reach the bolt on the exhaust side of the engine.

Live and learn.


03-20-2012, 04:53 PM
In order to get the Megasquirt installed on the engine there are a few changes that I need, and would like, to make to the current configuration. One of the biggest advantages to running a standalone is the elimination of various systems that arenít necessary to the function of the engine. A lot of emission control stuff falls under this category. Luckily where I live there is no penalty for removing these parts. I would like to try and make my engine bay as clean as possible. Iím not talking about doing a wire tuck. Yes they look great and I might do it one day, but for now I just want things as simple as possible and easy to troubleshoot.

An advantage to having a spare 4A-GZE is that I can prepare parts on that for the MS install. The intake system on this little engine is fairly complicated. I already have most of the intake track before the filter removed. Originally it started in the engine vent on the passenger side, up over the rear wheel well, across the back of the car in the trunk, to the filter mounted in the driverís side of the trunk, then into the engine bay, through the AFM, up to the throttle body, down through the J-pipe, into the supercharger, out of the supercharger to the intercooler, then to the intake neck, and finally the intake manifold.

There are vacuum lines at almost every point past the AFM and it looks very cluttered. My goal is to remove everything before the throttle body and simply have a pipe with a cone filter on it. I have not yet made the decision to bring a catch can line to this or to simply have the catch can vent to atmosphere. Next is the throttle body. I will be using a slightly larger throttle body that I got from ChrisD off of a 3S-GTE. It uses the same mounting holes as the 4A-GZE but is 5mm bigger in diameter. I need to do some work to the throttle body before I put it on. There are several vacuum ports on this throttle body that I need to plug up. I have already removed some of the hard lines that used to connect to the rubber vacuum lines. I will find some set screws that fit the holes left behind and tap the holes so I can install them with an allen wrench. This will seal up the throttle body to any possible leaks.

3S-GTE Throttle body bolted to J-pipe:

This hard line needs to be removed and plugged:

One hard line already removed and another to be removed, both holes need to be plugged:

I have already removed the stock idle air control. This is the standard Toyota three wire unit which does not connect directly to the MS. Yes there are circuits out there that can be used to make it work, but the unit is rather bulky and not that appealing to me. Also this is where the coolant lines attach to the throttle body. There is a little hollowed out cavity that would fill with coolant for a reason that I donít know. I really donít know why Toyota has coolant flowing to their throttle bodies in many of their engines. I will cut a plate out of aluminum to cover where the idle air control used to be to seal it up. I will use the stock gasket that is there.

Stock idle air control unit:

Where it bolts to the throttle body:

The next part in line is the J-pipe. There are a few more vacuum lines on this. The one that is in the car has a few lines that are connected between them thereby disabling their original function. This was done by the previous owner. The engine that came out of the AE101 has a slightly different J-pipe in that it came out of a front wheel drive car so the vacuum line to the brake booster comes off this piece. It is probably more efficient here because it never sees boost whereas the one in my car comes right off the intake manifold. However, I have no interest in moving it as mine works fine so I removed the adapter that threaded into the pipe. I also removed the large pressed in hard line vacuum port. What I am left with is one unthreaded hole on the side, one small vacuum port on the back and thro threaded ports on the back. I also have the gaping hole where the air bypass valve for the supercharger connects. Since I am running a larger crank pulley to increase the boost out of the supercharger I first modified the vacuum feeds to the ABV so it allows me to run the higher boost. I eventually made a block off plate to defeat the ABV altogether and have been running it like that for several years. In the interest of keeping things clean I will not be reinstalling the ABV when I put in the new J-pipe. I will either cut a block off plate to bolt over the hole or put a frost plug in there. I will also have to make the diameter of the intake on the J-pipe bigger to take full advantage of running a larger throttle body. Looking around online several 4A-GZE owners say that a bigger throttle body is one of the best modifications they have made.

The ABV hole that needs to be plugged:

My current thinking is to plug one of the threaded holes to seal it, use the other one for idle air control, use the small vacuum line to connect to the charcoal canister (if I keep it), and drill and tape the hole on the side for water injection. The idle air control Iím planning on using is the external adapter for a Jeep idle valve that is available on diyautotuneís website. I donít yet know if I will tap the other line into the intake pre throttle body and post filter or just install a small filter onto the valve housing. I am planning to use AN fittings and lines for this, just because they look good. I am currently running water injection with a nozzle into the intake neck. I am planning on changing it to a two stage system. I will run two pressure switches and three relays. I will also put in two solenoids. I will have to use two smaller nozzles instead of the one larger one I have in there now. I will replace the one in the neck with a smaller one (probably 1GPH), and put a larger nozzle (probably 2GPH) in the J-pipe. Now for this engine that is a little much, but the larger nozzle wonít do much for cylinder cooling. Instead its purpose is to increase the efficiency of the supercharger. It should keep the supercharger cooler and the water (theoretically) helps seal the rotors against each other making it work better. This should not be done on a turbo or centrifical supercharger. I will probably set the 2GPH nozzle to come on at around 5psi and the 1GPH at around 10psi. Now according to various calculators Iíve used the recommended nozzle for my engine is around 2GPH. I am counting on the fact that I am injecting pre supercharger to allow me to use a total of 3GPH. If it proves to still be too much I will run another 1GPh nozzle in place of the 2GPH.

The two threaded holes and the small vacuum line:

Where I want to install a water injection nozzle:

My only concern with putting the water injection nozzle here is clearance to the humps above and below it. I might have to try a few different nozzle holders. Again I would like to run AN fittings.

If you look at my signature you should notice that I like the look of polished aluminum. I considered polishing the J-pipe and throttle body, however I have decided against it for a few reasons. First of it takes a pretty big time commitment. Second, the shine needs to be maintained. Which is fine for easy to remove parts like the valve covers, but the J-pipe doesnít fall into this category. Thirdly the J-pipe isnít really all that visible. Itís one of those parts that you almost have to look for to find. So I will instead degrease it really well, then make my modifications (drilling and tapping) then I will media blast it and paint it. This should give it the clean look that I am after.

Any comments welcome.

Edit: For the sake of interest here is a shot of the throttle body fully open on top of the J-pipe. This shows how much material I need to remove from the J-pipe. The mechanical idle control (screw barely visible at the top left of centre) is very close to the outside of the J-pipe, but I am confident that a good gasket will prevent it from leaking.

You may also notice the oily inside of the J-pipe. This is because the PVC on these engines just vents back to the intake in front of the throttle plate. I am running the inexpesive DIY catch can made from instructions found on this site on the engine in the car now, but the more I think about it the more I think I want to run a catch can vented to atmosphere. Any thoughts?


03-20-2012, 11:26 PM
I dont care what anyone says, you can vent PCV to atmo with no problems. You dont need to be sucking the gases out of the crankcase. Engines vented to atmo LONG before PCV was required by the EPA and they had no issues.

03-21-2012, 03:25 PM
I dont care what anyone says, you can vent PCV to atmo with no problems. You dont need to be sucking the gases out of the crankcase. Engines vented to atmo LONG before PCV was required by the EPA and they had no issues.

This is true. Many older engines simply had/have a breather on the valve cover(s) to deal with the positive crank pressure created by blowby. Problem with these is an oily mist sometimes comes out of the breathers creating the need to clean under the hood more often (if you're concerned by such things). On my engine this breather is connected right to the intake track before the throttle body. This means that it functions just like a breather, except into the intake track. It has taken a little time to get my head around manifold pressures and such, but just before the throttle body will only see slightly below atmospheric pressure, and only because the engine is moving air out of the intake track. The only reason it is hooked up to the intake here is to prevent the oily mist entering the atmosphere and to filter any air comming into the crank case through the main engine filter.

Some more advanced newer engines have PCV systems that have multiple hoses between the crank case and the intake in different parts (ie after the throttle body) to try and create a negative pressure in the crank case. But as stated in the first page of this thread at WOT the intake manifold on a naturally asperated engine will be near atmospheric pressure so negative crank case pressure by means of a PCV system is highly unlikely at any time other than idle.

Negative crank case pressure is an advantage through. As such some high performance/race engines actually run a vacuum pump on the crank case. I believe around 15" is what they aim for. This is also an advantage to running a dry sump system as they often create negative pressure in the crank case. But I don't feel like spending more than I spent for my car on an oiling system.

With me cleaning the parts i am putting on the car I don't really like the idea of putting oil into the intake track. This can lead to increased carbon buildup and various minor things such sensors and vacuum ports getting plugged up.

An oil catch can will help prevent this, but some of the oil mist will make it's way through to the other port on the can. So it comes down to a choice as to whether you want the oil vapours to get into your engine or just into your engine bay. I think that the engine bay is probably easier to clean. And cleaning the catch can an filter on it regularily should help reduce this anyway.


03-21-2012, 06:54 PM
Glad to see some updates! So what are you doing for idle control? Aftermarket unit? Which/what? I've got to figure out my solution as my new 7M throttle body doesn't have an IAC valve.

03-21-2012, 09:08 PM
Hey Chris, been a while!

For idle control I am taking advantage of the four wire stepper control built into the MSII. I will use the external housing found here: http://www.diyautotune.com/catalog/custom-idle-air-control-body-for-use-with-jeep-40l-iac-valve-p-69.html. It uses a Jeep stepper valve (part number is on the page in the link) and is pretty much a universal fit. For the 7M-GTE body there is a nice place to hook up such a device. There is a large hard line that can be connected to a heater hose type pipe on the top of the throttle body. In this picture that line has been removed but the large hole on the top of the throttle body facing towards you is where it used to be:
I've taken a liking to AN fittings so my personal taste would be to tap that hole and install a fitting in there.

What options do you have with your ECU for controlling idle?

I will be routing one line from the idle control valve housing to the J-pipe as described above and the other will either be routed to the intake before the throttle body or simply have a small filter on it. I haven't decided which yet. Since I'm planning on using AN fittings I might mount the housing in the trunk and run a bulkhead fitting through the firewall for a cleaner look. If that is the case I will put a filter on the unit to keep the lines passing through the firewall to a minimum.

Speaking of updates, here's a small one:

Removed the last two vacuum hard lines from the throttle body:

This was done by placing the hard line in a bench vice and then gently rotating the throttle body to break it free. Then I rotated back and forth while gently pulling and they came out.

I dropped by the local speed shop witht he J-pipe and asked about AN fittings. We tried a 1/8th NPT thread but this didn't work. We then tried the metric equivalent and this too did not work. So instead of trying to find the weird thread that it actually is (I'm guessing some foreign pipe thread) I bought a package of two 1/4 NPT plugs and one 1/4 NPT to -6 AN fitting. I grabbed a drill and used a drill bit just big enough to remove the threads. Then I increased the drill bit size slightly and redrilled the hole. I repeated this five times until I had the correct size for the tap (In my case the tap called for an 11mm). I then carefully tapped the holes.

This requires a little patience to do properly. You can't just crank the tap in and call it done. The tap has to be screwed in a bit, then backed out, then in a bit further etc. Also a cutting oil should be used. Because these are tappered threads I had to keep backing out the tap and then test fit the plug and AN adapter to ensure that they went in the proper amount. If I ran the tap all the way in then the AN fitting wouldn't seal properly because it would run out of thread before getting tight. I did the lower hole first and then found out that the original hole wasn't straight. So my fitting is angle very slightly upward. This will not be visible in the car and won't affect the function in any way so it doesn't bother me much. The other hole is straight:

I don't want to tap the water injection hole until I have the nozzle with me to again insure that I tap the threads correctly.

I will probably take the throttle body to the hobby shop to find set screws to plug the vacuum lines. RC vehicle parts are usually pretty good for these. Hopefully I can make it there before the weekend.

The weekend project will be trying to change the timing chain on the daily driver/beater:


03-21-2012, 11:45 PM
My plan for PCV is to hook up to the factory location on the valve cover and simply run a hose down somewhere near the bottom of the tranny. This way, hopefully, the mess will stay down low with all the other road grime.

03-22-2012, 03:09 AM
Love the progress and keeping an eye on the emission removal and such since I would like to do that on my trac when I use the 3sge mani.

My plan for PCV is to hook up to the factory location on the valve cover and simply run a hose down somewhere near the bottom of the tranny. This way, hopefully, the mess will stay down low with all the other road grime.

Why not to a catch can so you can keep on eye on if the motor is running well?

03-22-2012, 03:54 AM
And it should always have a filter on it as it is a source of air getting into the engine.

Looking at emmissions I think I have it down to this:
-PVC to catch can vented through a filter to open air
-Charcoal cannister

That's about it! The EGR is already defeated. The header on the exhaust has no provision for it and I blocked off the port on the intake. All the vacuum plumbing has already been removed.

Vacuum lines should be simplified to this:
-One line ot the Charcoal Cannister (probably the one on the J-pipe
-There are two connections to the intake manifold. One is the larger one to the brake booster. The other is a "standard size". This will supply reference to the MSII, the FPR, the Boost Gauge, and the water injection (planned to have two pressure switches). The only one of these four items that are in the engine bay is the FPR, so I will likely tee this connection and run one to the trunk where the rest of the gear is and use a vacuum manifold. I'm all about simple!


03-22-2012, 03:58 AM
That valve looks nice. Since my EMS relies on OEM IAC valve, I'm thinking the 4 wire may pose a challenge. I might just add a plate to a stock IAC, and run a hose to it from the manifold. Not sure 100% yet...

03-22-2012, 04:02 AM
Love the progress and keeping an eye on the emission removal and such since I would like to do that on my trac when I use the 3sge mani.

3S-GTE as far as I know has a 24-1 cam position sensor in the distributor so I would think that running an MS would be a good fit... of course I'm biased... and it would let you eliminate a lot of the clutter in the engine bay. 3S-GTEs always look cluttered in stock form to me.


03-22-2012, 04:10 AM
That valve looks nice. Since my EMS relies on OEM IAC valve, I'm thinking the 4 wire may pose a challenge. I might just add a plate to a stock IAC, and run a hose to it from the manifold. Not sure 100% yet...

I have a few extra valves that I'm not going to use if you need one or two to play with. The 3S-GTE one I have should be pretty easy to set up. It has the same size hard pipe on it as the one in the 7M-GTE I believe. If you want I can cut a piece of aluminum and tap it for an NPT fitting and you can try that. Just let me know.


03-22-2012, 04:44 AM
Sure! Any chance we could do that in the next couple weeks? I have to finalize that before I send my manifold parts away for welding...

Sent from my Transformer Prime TF201 using Tapatalk

03-22-2012, 05:05 AM
I should have some scrap aluminum lying around that I can use to make a plate for the valve. I'll take a look tomorrow and let you know.

03-23-2012, 10:33 PM
Well I made a mistake. The vacuum line to the master cylinder on the engine in the car does come off the J-pipe just like the one I have on the bench. So having removed that nipple from the J-pipe I can no longer use it for the brake booster... in stock form. Two very easy ways to fix this:

1 Ė Use a ľ NPT to a hose barb fitting with the appropriate diameter for the hose that is currently on the car.
2 Ė Use an AN fitting and use the same style hose that I plan to use for all the other hoses in the engine bay.

Speaking of hoses I have to start thinking about what I want to use. The first though that comes to mind is stainless steel braided lines (Aeroquip or Earls are the most popular I think). Iíve been going through Aeroquipís catalogue and they have a hose called ďStarlite Racing HoseĒ. It does pretty much the same thing as the steel braided hose, but it looks like it has a cloth type braid with a red tracer. Now I do like the look of the steel hoses, but the black hose definitely would give a more subdued less in your face look. My mind isnít made up. Thoughts?

I also stopped by the hobby store yesterday to try and find set screws to put in the vacuum lines in the throttle body. No luck! Everything was too small. I made it to the hardware store at lunch today, but again couldn't find anything I liked. 1/4 inch machine screws look like they will work, but the heads on them are huge. As a last resort I may go back and get some and then cut the heads off after they have been installed. I still haven't tried a dedicated nut/bolt shop to try and find these set screws, nor have I looked online.

With a little luck I'll have the daily driver fixed tomorrow and try making another bracket for the coils on Monday.


03-23-2012, 10:39 PM
What about a rivet to plug the holes? Then you have a nice clean plug. Just a thought. Maybe use some sealer on the rivet before install to make sure it seals up good?

03-24-2012, 02:24 AM
I'm not sure a rivet would do. I've only ever used them for sheet metal, not for filling a hole. I found some set screws online that will do, but I would have to cut them down and minimum quantity to buy is 100. I could use a piece of aluminum rod that is tight enough to pound in, but I sitll like the piece of mind of having something threaded. I might try the bolt supply house and even tech services at work next week. I'll keep you guys updated!


03-27-2012, 02:52 AM
So the weekend did not go as planned. My plan to get the daily driver running fell flat on its face. The good news is that the problem with it is much simpler to fix than the problem I thought it had. The bad news is I didnít (and still donít) have the part I need.

However the weekend was not a complete waste. There was some more J-pipe tom foolery.

The first thing I did was to put the 3S-GTE throttle body on and then scribe the diameter of the throttle body outlet on the J-pipe:

This gave me a guide as to how much material I had to remove. I put an 80 grit flap wheel in my Dremel tool, filled up my water bottle, grabbed my dust mask and settled in for about an hour and a half of high pitched metal work. The end result is a hole large enough to take advantage of the larger throttle body. I also smoothed the inside of the pipe just behind the throttle body a bit. I expect this to have no effect on the performance whatsoever, I just like the way it looks:

Here is a ďduringĒ picture showing how the hole is a little bigger but still needs a little work:

As I said earlier, I donít intend to polish this part, but I do want it to look good. Well I had just over $200 in gift cards for Princess Auto (I think itís the Canadian equivalent of Harbor Freight), and they had a bench top sandblasting cabinet on sale for $150. So I picked up one of those and some glass beads for it. I started with some brushes and the parts washer and then cleaned the J-pipe with brake cleaner. This dries pretty quick and as soon as it was dry I used the blaster for the very first time. I did tape up the gasket surface first. This is the result:

This part probably hasnít been this clean since it left the factory!

The surface is clean, but unprotected. Since it is cast and rough (and roughed up a little by the blasting) dirt will stick to it. So as a preventative measure I decided to paint it. I started with some self-etching primer:

And then some engine enamel. I followed the directions of two light coats then a thicker one. Paint is still a little wet in the pics:

All thatís left is to tap the hole for the water injections nozzle (which I will do when it gets here which should be some time after I order it), and I can call this part done.

I would like to do something similar with the throttle body, but I need to get the set screws in first as well as take off all the throttle linkage and do some really good masking. Iím not removing the throttle plate because the screws are a nightmare to get out and replacements run the risk of getting swallowed by the engine.

Iíll try to get some pictures of the dry paint tomorrow. I might even use a real camera (gasp!).


03-27-2012, 03:59 PM
As promised, some pictures taken with a real camera:

I put the throttle body back on for a ďbefore and afterĒ look.
No flash:

With flash:

Remember I have done nothing to alter the surface other than blast it with glass beads and paint it. I did not grind any of the casting marks down or anything. If I were trying to build a show car I probably would have, but for this part which is mainly hidden I think that this is fine.

There are still a few things I need to do before putting this on the car. I need to wash the part really well to ensure that all the bits of glass are out of it. Even one could ruin the engine. Also the instructions for the paint state that I have to cure it at 300F for two hours. I donít have an old oven for this so I have to wait until everybody is out of the house... And I still need to tap the hole for the water injection.

While I was playing with my bead blaster I decided to try and clean up one of my spare cam pulleys. I put some self etching primer on it but didnít go any further than that. I donít know what colour I want to paint it. Not that it really matters, I still run with the timing belt cover.
No flash:

With flash:


03-27-2012, 06:28 PM
You will find that the glass beading cabinet will be one of the best purchases you ever made. I couldn't live without one.

03-27-2012, 07:09 PM
A few years ago I picked up a parts washer cheap from a garage sale. I thought that was the best solution to cleaning parts, but now I've got a new favourite. It's not complete though. It doesn't have a light and I need to rig up some kind of shop vac dust extraction system. All in good time...

But you are right, everyone who doesn't have one needs one!


04-02-2012, 09:38 PM
Well I have some more work done. I'll be posting pictures tonight or tomorrow.

I've been looking over the emissions control and I'm still undecided what to do with the EVAP system. I have learned that I likely can not use that vacuum connection on the back of the J-pipe for this (in STOCK configuration). The stock system uses one of the vacuum ports on the throttle body that I have plugged. This is called the purge port. It is one of the very small ones that open to the throttle bore near the butterfly valve. The charcoal canister has three check valves in it that open under various conditions. The filler cap also has a check valve in it. The charcoal canister has three connections on it. One to the fuel tank (on top as a vent tube), one to the throttle body, and one open to the air. When there is pressure present in the gas tank a check valve in the canister opens allowing the pressure to release. This vents out the bottom of the canister while the actual vapours are captured by the canister. When the tank has a vacuum the previously mentioned check valve is closed and another one opens allowing air into the tank through the charcoal canister. Also the check valve in the cap opens allowing air into the tank that way as well. This is important so the gas cap doesn't get suctioned onto the filler pipe preventing one from filling the car with gas.

The connection to the throttle body purge port first goes through a vacuum switching valve. This valve is a temperature controlled valve. When the car is cold (below 35C) this valve is closed. So everything acts as described above. When the temperature is above 35C then the switching valve opens. With the throttle closed the purge port is in the intake system before the throttle plate, and as such sits at atmospheric pressure. When the throttle is opened the purge port is now on the other side of the thottle plate and the purge port has a vacuum applied to it. This vacuum is enough to pull open the last check valve in the canister which in turn sucks all the fuel vapours into the intake. So if I were to connect that pipe on the J-pipe to the vacuum canister through the switching valve then everytime the engine coolant is above 35C there would be a vacuum leak through the charcoal canister which would likely raise my idle a bit.

Here is the diagram from the BGB for the MR2 EVAP system:

Now I don't particularily like the idea of just venting the gas tank to the air as it would tend to make the car smell like gas on hot days. I could remove the charcoal canister and plug the vent line but this could lead to high pressures in the gas tank while the car is sitting in the hot sun and this can lead to problems, some of which could even be dangerous.

I also don't think the canister on my car has been serviced for at least twenty years and have no doubts that it isn't functioning properly.

So my current options:

1 - Reinstall the vacuum line to the throttle body and run it as stock. Replace the charcoal canister.

2 - Remove the charcoal canister and install a check valve on the vent line from the tank. This will vent pressure to the atmosphere only while the tank is under pressure. When tank is under vacuum the check valve in the cap will allow air in. This is the simplest/easiest option.

3 - Replace the charcoal canister. Run the line that goes to the temperature controlled vacuum switching valve to an electronically controlled vacuum switching valve. I could then use port on the J-pipe, but I would have to find some way of running the valve. I could use one of the outputs on the megasquirt (the one I'm thinking of using for the engine bay fan) to open the valve when the temperature is above 35C and the throttle is >25% open.

4 - Same as 3 but have the valve manually controlled with a button in the car. I would have to remember to press it now and then in order to remove the fuel vapours from the canister though.

5 - Replace the vacuum canister and plug the port to the VSV. I would be relying on the pressure change in the tank alone to suck the vapours back in. I don't think this would work.

6 - Replace the charcoal canister. Connect the VSV as stock and connect it to the J-pipe. Adjust the mechanical idle control for desired idle with the car warmed up (VSV open). Set up the MS2 to keep the idle control valve open until engine temp is above 40C. This will account for the vacuum leak caused by the canister. I would eventually like to have the MS2 doing closed loop idle control. Meaning that instead of having the temperature of the coolant dictate how open the valve is, it dictates the desired idle RPM. This would automatically account for the vacuum leak and adjust accordingly.

Option 2 is by far the most tempting just for it's simplicity. Option 6 is probably the best in terms of trying to hide the vacuum lines down low and being the right thing to do. However it might pose two problems:
1 - If too much air gets into the intake this way then it might be impossible to get the idle low enough.
2 - Every time the aire filter is serviced the charcoal canister needs to be serviced as well.

What do you guys think?

I promise some pictures are comming.


04-02-2012, 11:22 PM
Gah. Me and you are so different. I just ripped out the EVAP system with no regard to smells or functionality. Leave the gas cap loose when the car is parked.

04-03-2012, 04:10 PM
Well isnít doing things differently half the fun? How do you find the gas smell on hot days? I imagine that so long as the engine is running and the fuel pump is moving fuel out of the tank the pressure shouldnít stay too high in the tank. The more I think of it the more I think I might just go with a check valve to release pressure.

Well I did promise some pictures, so hereís an update:

While rummaging through some parts bins at work I came across a box of set screws. I just happened to find some 6mmX1 screws. I could not have found a better match for what I was doing:

So I grabbed the tap and die set and found the 6mmX1 tap and first tested it on a scrap bit of aluminum. It was a match to the set screws so I drilled and taped the holes in the throttle body. I then test fitted them:

In the above photo these are screwed in to the end of the threads, meaning this is their final position.

With that out of the way it was time to take the throttle body apart. Now I didnít completely disassemble the throttle body. The small screws that hold the plate in place are a nightmare to get out. They are usually a soft metal and are either loctited in place or pressed (squished) after they are inserted. I have tried to remove these from other throttle bodies and failed, so I didnít even try here. I did remove all the springs and linkage bits though.

After it was apart I gave it a good cleaning with the parts washer. I then used brake cleaner to wash the solvent off and for its quick dry properties. Once the part was dry I masked off the gasket surfaces with tape. As this was headed for the sandblaster I also used a hot glue gun to protect the bearings from the sand. The set screws were wrapped in masking tape and inserted into their holes. The throttle body was then ready for bead blasting:

I found that my blasting cabinet needs a few modifications. First off it gets far too dusty in there and this makes it almost impossible to see what Iím doing. I had to lift the lid a couple of times to check my progress. Every time I did the dust would waft out of the cabinet into the garage. So I need to rig up some kind of dust extraction system. Nothing too complicated, just remove one of the filters and put a shop vac connection point on it. Secondly, there isnít enough light in there. I have a trouble light shining through the top, but it isnít enough. There is a switch built into the cabinet and some mounting clamps for a tube light of some kind, so I will build something to go in there.

Anyway, back on topic. The throttle body was blasted and then washed again. I reapplied the masking tape and gave it a quick wash with the brake cleaner. This was just to get all the oils from my hands off it. Just like last time I applied some self etching primer following the instructions on the can:

At this point I got into the painting and neglected to take any photos. But here are some taken the next day:

The finish on the throttle body is much smoother than the J-pipe. This made the paint look really good. For reference in the above photos the flat top bit isnít painted, it is just flat aluminum. I almost regret not spending the time to flatten the surface of the J-pipe. Almost. Remember this part is hardly visible when in the car. Iím picking my battles.

All that is left is to install the set screws:

Okay, so that one isnít in permanently, I just wanted to see how the black looked in the throttle body. I still have to apply some thread sealer before this is installed for the last time.


04-04-2012, 03:18 PM
All right, so not really an update, just some more pictures. I painted all the hardware that goes with the throttle body and put the TPS rotated 180 degrees (hopefully without the ABV Iíll have room for the wiring this way):

And some numbers I donít know the meaning of:

What do you think?


05-04-2012, 06:22 PM
Been quiet for a bit. A lot of stuff has happened since my last update. Unfortunately none of it involves the megasquirt project.

I got the daily driver fixed and it has been put back into service.

Oh, and this is the major reason nothing is happening:

So yeah, a much larger build of a completely different kind is now taking all my money...


05-04-2012, 07:54 PM
Nice!! Congrats on what I assume will be a new home............

And the TB looks great

05-07-2012, 08:30 PM
You assume correctly! The builder is targetting for a completion date sometime in December. I hate winter moves. I would expect that the date will more likely be somewhere in the January/February of next year.

Sadly the current state of funds do not allow for the building of a garage. However, it is in the plans. I am having the builder run gas and power to the garage location. With the local bylaws and permitting I can build a 28' by 24' garage. The lot we selected is two feet wider than the neighbouring ones for just this reason!

With everything that is happening I am forced to change my plans a bit. I will still be doing a megasquirt system, but I'm putting off the turbo plans for now. It is likely that I will refresh the AE101 4A-GZE, drop in a set of big port cams, and roll with that for a few years.


05-07-2012, 09:26 PM
A fresh 4AGZE is still not a bad motor to roll with lol. And a worthy sacrafice.

Hope money changes before the house is done so maybe you can squeeze the garage in right away.

05-08-2012, 02:22 AM
Yeah I like the 4A engines, and the instant boost of the GZE is nice. I'm still going to use the 3S-GTE throttle body. Hopefully I don't need to do more than hone/gaskets/rings/bearings. New pumps will find their way on as well. An advantage is I get to wire everything up before putting it in the car.

Money isn't going to change I'm afraid. The builder was going to charge me $14,000 just to put the foundation for the garage in. I could get a 20 by 20 garage foundation for $8,000. I know a few contractors so I'm sure I can get this done for less money. I'm hoping to have the foundation down in a few years. After that the garage should go up pretty quickly.

05-08-2012, 03:03 AM
You could all ways put up a metal frame building, basically one of those metal carports cept put walls and a door on it. That way you can get the slab poured and stick the metal building on it till funds allow a proper garage. You could even do the slab yourself.

05-08-2012, 02:35 PM
I plan on trying to do the slab myself. I know people who know people in the concrete industry. Unfortunately with the climate here I do have to dig the footings down around 4' - 5'. I had the option of having the builder pour a parking pad, but this wouldn't have the footings for a structure on it, and because I'm having the builder run power and gas to the garage location they can't pretend that I'm not putting a garage there. And since it's a new community there are some strict architectural controls in place. Any structure I build on the property other than a shed must match the house's siding in both colour and material as well as the roof.

Anyway, I don't want to get in trouble for having the thread go off topic.

I am thinking that I'm going to put the spare 4A-GZE on an engine stand. If I can do it next week I will. I will then start looking at the stock wiring harness and making notes on what I need to use from that.


02-15-2013, 09:34 PM
How plans change...

Nothing has happened with the MR2. It sits in the back yard under a cover and under snow. Once or twice a month I fire it up to keep the oil circulating, but other than that it sits untouched. I was planning on doing work on the spare engine in my parent's garage, but since they are being kind enough to let me store furniture for the new house in there there isn't any space to do that. All I've been able to do is scribble ideas down on paper as to how I'm going to wire it in. I'm thinking of simplifying the whole car's wiring by installing a Painless Performance kit. As the wiring is already questionable (and has caught fire) and has several generations worth of owners hacking it I think it is the safest thing to do. This will take a little time as money is tight.

The only work I've been doing on cars is routine maintenance and a heater core replacement on he dd.

As for the house, well it now looks like this:

Naturally the builder is behind schedule but we have been given a possession date of just under three weeks from today. Then I will be able to move, clean out my parent's garage and hopefully get started on this again.

Oh, and I should mention that I got my very own shop vac for Christmas, so that blasting cabinet should have a dust extraction system pretty soon.


02-16-2013, 07:55 AM
Congrats on the progress of the house!!! That should be nice to move into.