View Full Version : MAF to MAP?

02-27-2009, 12:23 AM
I'm really kinda in over my head here, but let me try to explain what i'm wondering.

First of all, am I correct in thinking that all the 3s engines (well at least mid 80's early 90's) use the MAF system while the 5SFE uses the MAP system?

If this is the case, are their major advantages or disadvantages to either? I know I prefer the MAP system astheticly becasue the MAF is just this big box that clutters up the intake.

The core of my question is, is it possible and/or feasible to replace one with the other? Specificly, if I were to install a MAP system in place of the MAF system, would I be able to easily modify the output signal so that the ECU is able to give the correct air-fuel mix?

I have read that both systems often use the 0V-5V voltage range to send the signal. That said, would there be a large variation on how that signal would be read by the 5s V.S. 3s ecu's?

This has always been a question in the back of my mind and I would REALLY appreciate any information about the two systems that you know.

I would also appreciate a "your a dumbass Eli" if I'm seriously off base with the way these systems work.

02-27-2009, 12:30 AM
You're a dumbass Eli...

Just joking. It's been a while since I've read this thread, but I'm pretty sure with it being a Nuclearhappines thread, there's tons of info about one or the other in it and why you can't just "swap" them.

02-27-2009, 12:40 AM
MAF system is more accurate than MAP. MAF is Mass Air Flow while MAP is Manifold Absolute Pressure. The idea is the same, the execution is different. I forget how the MAP works, but the MAF I do know works by heating 2 wires, incoming air cools them, it reads the value because with heat comes different resistances in the circuit. I'm quite honestly not fond of MAP, but that may be because I know more about the MAF system. It isn't just some big box, I know on my ranger its a housing to hold the sensor and the housing is built to allow proper operation. ANYWAY you honestly are better with the MAF system in terms of gas mileage and achieving stoich

Sorry if I'm rambling, Im tired as can be...

02-27-2009, 01:42 AM
You don't have a MAF system with the 3SFE. You have an AFM. Air Flow Meter. It has a flapper door that tells the ECU how much air in coming into the engine.

02-27-2009, 03:24 AM
Ahhh, I see.
I guess I got confused because my famlies ranger has a MAF and the system looked similar to the one of the 3S.

02-27-2009, 03:47 AM
the MAP sensor works off vacuum dram in the intake plenum. under idle the MAP sensor reads a higher vacuum and at part to full throttle, reads low vacuum. The MAP adjusts your A/F mixture by reading the amount of vacuum.

I know that is a really basic explanation, but thats how it works in a nutshell

02-27-2009, 04:12 AM
I'm quite honestly not fond of MAP,

haha you arent very into turbo cars then lol ;)

joe's gt
02-27-2009, 04:20 AM
A MAP conversion is pointless IMO. The only reason you should do it is if you are going way beyond the capabilities of the stock AFM which would mean your going with an aftermarket EMS with speed density anyways. True, the stock AFM is horrible and would probably be much better with an MAF, but converting to speed density is not worth the effort for the slight increase in driveability IMO. The closest thing to an easy conversion seems to be the 3rd gen conversion, which appears to be a very viable alternative. It ain't cheap tho.

The best option would be if we could actually modify and reprogram the signal in the stock ecu.

Speed density can't compensate for mods as well as MAF. You have to re-tune after changing something that affects airflow. Although you should do that anyway. IMO, its only worth it for high hp applications.

02-27-2009, 04:47 AM
i talked to chris at KO racing about a remedy for this issue its in a wat a cheaper way to get the benefits of a ems and keep the stock ecu and delete the afm(VAF vane air flow) meter and yes you will only need to do this once you have gone over the capabilities of the afm- the VAF meter is a shitty way of metering air flow for any setup IMO. so i do know the feeling of wanting it gone.

that said if you can make a converter to invert the 0v-5v signal from a MAP sensor you can just put said converter between the signal wire from the MAP sensor to the ecu and presto you have VAF to MAP it would be nothing more than simple electronic hardware and very very cheap i just don't know how to put that kind of stuff together.
with a 2 bar map sensor it will end up taking some fuel off the total but still be way safe as the 3sgte ecu's tend to over-fuel a lot.

i guess if you did that you WOULD gain a bit of flow and better fueling for very very cheap like 10-15 hp at best but hey its worth it

the reason for the fuel difference is that the MAP voltage is linear with the rise in pressure. but the VAF is more of a curve when you compare the velocity of the air to the pressure at the manifold. but for most i guess its not worth it. the more you learn about the way sensors work and stock ecu's fuel trim according to the imputs it receives you will get what i'm saying. its simple in concept but the fun part is making the hardware.

02-27-2009, 05:40 AM
One thing we are missing.

Eli is referring to a 3SFE in a 4th Gen Celica drop top.

Not a 3SGTE in any form or fashion.

joe's gt
02-27-2009, 06:06 AM
v8killer, did you read that thread linked above? That is pretty much what Nuke tried to do. What he was doing was way beyond my comprehension, but basically he tried doing exactly what you said. He tried simulating the AFM voltage to the ecu using a MAP sensor using tables and such, and it turned out to be an enormously complicated project. I don't know whether you have the knowledge or not to do what he was talking about, but it would be great to try and start it up again to see if there is some affordable way to eliminate it. The electrical engineering aspect of that was way beyond me, but I get what he was trying to do and the methods he was using.

Your right that its a simple concept, but it is an enormous challenge to try and implement. I agree that the VAF (AFM) is crappy way, if it was MAF it would be a lot more effective and IMO, that is the best method for measuring air flow.

edit: based on what murg just said, it is pretty much pointless.

03-01-2009, 07:16 AM
Well, I'm going to check it out, and if it looks simple enough for me to tackle, I might give it a shot. I have pretty good judgement when it comes to knowing if somthing is over my head so I won't spend too much time with it if it doesn't work.

03-01-2009, 05:05 PM
What are you going to try it on?

03-01-2009, 05:28 PM
It's a vane air meter VAM and it works on 5-0V, it's bassackwards from most of the rest (MAFs). The 7MGE uses the same thing, 5-0V flapper, and you can use the MAFT pro to translate to a 3" or 3.5" LS1 MAF on the 7M, have yet to see anyone attempt it on the 3S. ;)

03-01-2009, 05:52 PM
Well, I'm going to take a look at the voltage outputs on the vane air meter as I rev and then see if I can get a similar reading from a fabbed up MAP system, if it looks way too complicated to be worth it, I'll give up and play some video games.

I'll be trying this on the 3SFE in the vert I just got.

joe's gt
03-01-2009, 07:16 PM
An MAF IMO would be better. That translator would be a much easier task than a MAP conversion.

04-08-2009, 05:41 PM
Well, i've been doing some research and the systems are simple enough that I havn't given up hope yet.

As a reminder, I was ill informed about the semantics of this project. All i'm trying to do is replace the vane air flow unit from the 3SFE with the 5SFE's MAP sensor.

The MAP sensor puts out a 4.5v-5.5v signal which would have to be translated to a resistence value which is what comes from the AFM.

The AFM seems like it might be a little wonky in the resitence values it puts out becasue the BGB says that the expected values vary from 200ohms-600ohms closed to 20ohms-1200ohms open, which doesn't make much sense. My guess is that the ECU just learns the pattern of resistence change. This may afford me some leeway in setting up my system if I can just use almost and resitence values I want.

As always, i'm starving for input.

04-09-2009, 11:28 PM
I did this, with a microcontroller and look up tables. And I sent a prototype to a member on here who got the car to idle cleanly with it but it wouldn't rev.

I dropped the project (this is some 3 years ago)... From what the tester measured and desrcibed, my best guess as to a 'next step' was that the setup i was using was slow at processing (which i knew but i thought it was fast enough to at least run). So I would've had to simplify my code, dumb down my lookup tables a bit (so the translation table would be less accurate but it would process faster), and I would have to up the clock frequency on my chip.

Anyway, leave my project alone and get back to the theory of it.

The main difference between a map system and MAF / MAS / AFM / Karmann system is RPM.

in the most simple of translations your maf system = map * rpm

so that at 3000 rpm and 3" of vacuum the amount of air (or voltage, or resistance) that you are sending to the ecu is different than what you send at the same pressure (3" of vacuum) at say 4000, or 5000 or 6000 rpms

To have a truly accurate transfer function:

Mass air flow = pressure * temperature (density) * volume * volumetric effeciency * rpms

so mass airflow at the same PSI is different as temperature, rpm and VE change
temperature you can measure with an IAT.
VE you can figure out from dyno charts and some standard engine equations
rpm you can measure

but then you're not talking about a simple electronic circuit,

at least you need a frequency to voltage converter to 'capture' rpm, you need an iat input, you need a map sensor, you need somewhere to 'store' VE or to filter the RPM voltage in a non linear way that maps with RPM ...etc

in an analog circuit it's pretty complex
so i tried to do it in a digital circuit.
One of the things i did in my circuit is figure out rpm by counting pulses... so i'd wait say 1/10 of a second and count how many ignition pulses happen in that time... this slowed down my circuit alot because i had to wait and count before i could decide what rpm it was to run that equation i posted above. which killed the speed as i said before.

So if i were to do it again i'd use at least 2 chips
1 SOLELY to read RPM triggers, that sends a voltage to the main processing chip where the math is done, this way that processing chip can run full speed, and when it needs an rpm value it can just poll the other chip, get a voltage back (in 10ms) and plug that into it's equation.

So how does this help you if you're not an electronic engineer that wants to BUILD a product to market ? and just want something for your car...

I have 3 solutions for you...

1- If you're willing to use a MAS (which are better than vane air flow meters) i'd use a bosch MAS which has a non linear flow / voltage chacteristic similar to that of the nonlinearity in the VAF (Except it's upside down) ... and then i'd build a simple inverter analog amplifer circuit so that voltage out = 5-voltage in
this will give a MAS that tracks pretty closely to the stock AFM and can be built in a weekend.

2- If you truly want an ONLY map setup... use something proven like the MAF-T pro (Google it). .. it's a product that does what i was trying to do, when i was trying to do it, they only had specific applications like the Z28 and the 30000GT and the DSM's, but hte MAF-T pro as far as i understand is fully programmable so you can make it work on any car.

3- Get an e-manage ultimate. E-manage ultimate can take in rpm and map
and give you a 16 X 16 matrix for one of two things:
1- Corrections (from the stock fuel table)
OR: 2- Fuel injector pulse width duty cycle where you complete overwrite the stock computer's fueling to be whatever you want it to be having direct control over your injectors almost like a standalone. (WAY COOL).

The e-manage also has a simple temp correction if you wire up your temp sensor...
so that way you have factored in temp, rpm, ve (through the 16X16 table) and MAP (again through the 16X16 table)....

Hope this helps, sorry if some of it is too technical...

04-09-2009, 11:39 PM
If you have an SAFC for fine tuning or whatever (on the car you're talking about) and are willing to try this out for the educative aspect of it, You need:

Frequency to voltage converter ( to convert rpm pulses into a 0-5volt signal that represents rpm).
A 0-5 map sensor.
An IAT sensor wired up to read 0-5 volts through a voltage divider / analog amplifier
Ignore VE for now (You'll do that through your SAFC)

Once you have all your three signals (pressure, temperature, and rpm) represented as 0-5volt signals, use two analog multipliers to multiply them out.

A= MAP * Temp
B = RPM * A

so what comes out of your second multiplier is B = RPM * MAP * TEMP

Now this signal is 0 to 5 volts
You run it through one inverting stage so it runs from 5 to 0 volts.
This gives you a very very rough AFM mimicking signal
You fine tune it through your SAFC later on.

Let me know what i can help you with of all these ideas and we can talk more about it...


04-11-2009, 12:09 PM
You run it through one inverting stage so it runs from 5 to 0 volts.

and there is the key an inverter of this type would be simple and cheap but how to make a circuit board to do this without any time lag is the part i'm dieing to find out just because i want to clean up my engine bay more than anything and we can go from there. the 2 bar MAP sensor on the usdm 3sgte's is pretty close to the opperating range of the VAF. as for other engines i don't know for sure what is close.

04-11-2009, 12:12 PM
1- If you're willing to use a MAS (which are better than vane air flow meters) i'd use a bosch MAS which has a non linear flow / voltage chacteristic similar to that of the nonlinearity in the VAF (Except it's upside down) ... and then i'd build a simple inverter analog amplifer circuit so that voltage out = 5-voltage in
this will give a MAS that tracks pretty closely to the stock AFM and can be built in a weekend.

all i want is that inverter or tell me how to make it

sorry just skimmed through the first time

04-11-2009, 05:43 PM
You'd be taking out a Mass Air Flow meter and putting in a Mass Air Sensor.

The difference between measuring flow and measuring mass is temperature.
Ideally you need 2 stages, 1 to UN-correct the mass for air temp, therefore changing a mass reading into a flow reading, and the other stage to do the inversion....

If you don't un-correct for temperature, you have to make the ECU see a fixed value for IAT so it doesn't perform temperature corrctions twice.

Just to give you an idea going from -30*C to 50*C is a correction of about 35% so it is significant if you live in a place with seasons.


If all the resistors are 10k ohms
and V2 = 5 volts
and V1 = MAS voltage

Vout = 5-Mas voltage

That's your inverter...

I use this place for prototyping circuit boards:

And I use digikey.com for parts

If i were to build this instead of using 1 resistor for R2 ... I'd use two:
1 Resistor of 8k ohms
1 adjustable resistor / potentiometer of 4k ohms.

When you turbo the knob all the way on the pot to maximum position, R2 (the sum of our two resistors) becomes 12k
So the output becomes Vout = 1.1*5 - 1.2 MAP (so you have a gain of +15%)
When you twist the pot all the way to 0 , R2 becomes 8k + 0 = 8k
Vout = 0.9*5 -0.8 MAP (so you have an adjustment close to -15%)

With that you can have an adjustable inverter so you can try to fine tune things if you need to.

Have at it ,let me know how it works out...

04-12-2009, 02:16 AM
ok so that inverter will work for sure right all at 10k and possibly add the potentiometer for r2.

so it directly reverses the signal w/o the potentiometer

joe's gt
04-12-2009, 06:54 AM
Finally I see a real world application of an op amp after 4 yrs. in an engineering curriculum. lol.

04-12-2009, 06:32 PM
op amps are great for lots of things that we do with cars.

HKS Fuel Cut Defender
Greddy Boost Cut Controller
The input buffers on the E-manage, AEM, and SAFC

Yeah V8 killer, if everything is 10k it will be an inverter. IF you want adjustability use an 8k + 4k pot so you will have a rane between 8k and 12k so that your 'gain' varies between 80% and 120% basically (8k/10k or 12k/10k)...

You need a good 5 volt source. You can do that if you have a sensor voltage supply from the ECU, or if you want you can use a 5Volt voltage regulator ....


04-12-2009, 09:20 PM
thanks nuclearhappiness you i was going to do something kind of like this with an ADC and a DAC but it was too complicated and wasn't going to be worth the money if its ok i'll be asking you for info every now and then

04-21-2009, 03:37 PM
yeah that's why my system had, I was using an analog devices 'analog microcontroller'

i had 4 ADC's
1 for IAT
1 for MAP
1 for offest correction to shift the whole curve up or down
1 for gain correction to do +/- 15% adjustments on the top end

and I had some digital inputs

1- Trigger input which i used to keep track of rpm duty cycle
1- Digital input for a switch to switch between a 450cc and 540cc map for 3sgtes that is

and on the output side

1- DAC to create the new airflow signal
the other DAC's were spare
1- Digital output , running to a darlington amplifer, so that i could control TVIS when i wanted based on flow rather than boost or rpm

1- output to an LED, i forget weather this was a DAC that would output a the same flow voltage making the LED brighter , or if it was a digital output that was pulsed to show that it was really reading rpm... don't quite remember at the moment ...but i used it mainly for diagnostics and development...

You could use it for anything, i think the final program i used it to show when I had switched TVIS.

I'm sure if i had a lab with oscilliscopes and capture capability as well as a test mule i could've finished this product ...

inside of it i had 2 lookup tables... one table would take in RPM and look up VE

Then when i had VE and RPM i'd do some complex 8 bit math

HP = Boost (measured from a ADC) * RPM (measured from my trigger) * IAT (measured adn corrected based on kelvin) * VE ...

and i'd come up with an internal 'air flow' number... that number was linear, and so i could linearly apply corrections to it

so output airflow = airflow number * gain adjustment (measured from my gain pot through an ADC) + offset adjustment (measured from my offset pot through an ADC).

Once I had my corrected airflow signal, I had to invert it and warp it to mimic the 3sgte AFM

Since that is a nonlinear relationship with a very complex equation (that i worked out to be some 4th order polynomial)... which i plotted in excel for 5 volts / 256 values or in steps of 0.02 volts

The rather than have the controller have to do that math internally and waste time, i took the excel table of flow vs afm voltage (for 256 values) and loaded it as my second lookup table in the chip

so once you have your corrected airflow signal you can just go and do a lookup and see what the AFM output for that should be... and get that number and just push it out to the DAC and out it goes.

The other thing i do with my airflow signal (before i lookup the AFM voltage) is if the airflow signal is greater than say 50%, then turn on TVIS
If it is less than 45% turn off TVIS...

so that gives you new TVIS control, without some hystorysis in case you are holding power right at 50% it doesn't flicker on and off...

We had some 10 page thread on it here... it was a good idea but i was a bit ambitious with the tools i had available :)

I got your private message, post up a picture or a datasheet of the part you got so i can know what your'e talking about... every opamp has different pins so i'd have to see which part you bought to help you man...

04-21-2009, 03:42 PM
See typing that i knew i was forgetting something ...

I wasn't reading in IAT , since this is a flow meter not a mass meter.

IAT goes directly to the ECU as it always has.

My 4th analog was TPS

if you are at 8000 rpm, 15psi, with 100% VE, but the TPS is at 0% then your airflow signal should be pretty low ... and if 100% TPS then it should be high ...


05-20-2009, 01:05 AM
I would still be so interested in that Nuke! God would I ever. With people that have rom tunes, the AFM is the worst thing in the world!! Please!!! make one.

05-30-2009, 04:14 PM
Clay are you talking about the


or are we talking about the

MAF To Bosch MAS