PDA

View Full Version : Options for eliminating the AFM



Hookecho
10-05-2011, 05:15 PM
I really want to get rid of my AFM. It''s old, unreliable, and replacements are just as old.

I know there is the option of upgrading to Gen3 Electronics but that seems to be as expensive as some EMS's. Especially once I buy the turbo to compliment the system. I can buy a Gen3 motor for that cost. Thing is this, I really dont have any higher power goals for my motor. I will likely upgrade the turbo when the CT26 gives out but that is it.

What I would like to do is keep the upgrade as simple and cheap as possible. If I can get away without buying an EMS that is complicated to tune then that is what I want to do. If there is some kind of piggyback that I can use to run a MAP based system then that is great. I just don't want to spend a lot of money if I don't have to.

What are y'alls recommendations on EMS's. What do you use now, have used in the past, and what is the cost?

I've read a little on the MAP ECU. It seems to do what I need it to do for $300. I don't know what is involved as far as tuning but I'm quite capable of figuring it out. What about SMT6?

T-spoon
10-05-2011, 05:34 PM
For your stated goals the original mapecu should be more than sufficient, and can probably be found pretty cheap (supra guys are selling them pretty frequently after upgrading to standalone). Is it the best option? I really don't know. I can tell you they're easy to install, and since you're running stock a good base tune file should be fairly easy to come by, so it shouldn't require any tuning - although you would definitely pick up power if you had it dyno tuned. I'd compare that to the other options, but I'm not personally experienced with'm. I will say though that if you don't have one already, a wideband o2 is going to make your life a lot easier whichever route you go, but you still want a working stock o2 to regulate cruise and idle unless you want headaches with the mapecu.

I'm using the mapecu2 on the supra right now.. I will most likely go to an aem standalone eventually (so much money!!) for fine tuning reasons, but I'm far from stock, so it's really a different scenario for me.

Hookecho
10-05-2011, 05:46 PM
Well, a dyno tune is not a problem. There is a Dyno Jet a mile up the road from me. That is something I am willing to drop a couple hundred bucks on once I get the system running on my own.

vip09
10-05-2011, 05:53 PM
mapecu, vpc+gcc, standalone, or gen3 efi are your options pretty much.

Since reliability is a priority and power is definitely not, go gen3 efi. Even if it costs as much as a standalone, you won't have to deal with all the quirks that come with the standalone.

Luni
10-05-2011, 06:08 PM
In the end it does NOT cost anything close to a standalone.

Ive told you time in and out, I can piece you together a Gen3 EFI swap for not that much.

Overview:

Gen3 ECU - 150 ish
Gen3 injectors - 125-200ish
Gen3 MAP sensor - 150 ish
Gen3 igniter - 30-50 (there are others that work that can be found)
ATS Bored rail - 75 ish
Repin of harnes (youd just do this yourself, free) OR Rev1 harness to Gen3 adapter from Berk - 100ish
ATS Exhaust cam gear - 75 ish OR Gen3 exhaust cam (set for 125-150 ish)

So around 800-900 gets you the full thing (maybe even cheaper). You cant touch an EMS for that, and even if you do, youre going to be in it much more for dyno time and tuning.

It doesnt even compare pricewise IMO.

Hookecho
10-05-2011, 06:16 PM
The MAP ECU is cheaper though. $400 new and I can download a fuel map. Cheaper if I buy used.

I'm liking it the more I read about it. It's simple and wires up the same way as my SAFC did. I already have the connectors in place for it.

Can I use the temp sensor from a 5sfe with this and put in in the intake pipe?

Hookecho
10-05-2011, 06:22 PM
and it's much less work than the Gen3 upgrade. I don't look forward to replacing a cam and fuel rail/injectors.

T-spoon
10-05-2011, 06:39 PM
FWIW, I have a VPC+GCC I'd sell you for next to nothing, but it is chipped for supra 550cc injectors and the wiring is old. I doubt that's the way you want to go, but it's here if you want it as a cheap option.

Luni
10-05-2011, 06:50 PM
My problem with the MAP2 ECU is it still uses the Gen2 ECUs resolution, fuel maps, and processors.

The MAP2, as good as it is, in the end, is still only as good as the stock ECU is IMO. You do what you want Bruce. Youd be one of the first people to make the MAP2 work properly on a 3SGTE if you did it. Knowing what I know about the 2 EFI systems, I wouldnt run one though. Just my .02

Hookecho
10-05-2011, 07:26 PM
Alright then. I'll just hold out for Gen3 or live with what I've got.

If MAP ECU can't be tuned to work then that's that. I thought it was simpler.

vip09
10-05-2011, 07:37 PM
My friend is selling everything you need for the gen3 efi minus the injectors for like $450 or so. He bought my old AEM since I went with a Haltech so he doesn't need the gen3 stuff any longer.

T-spoon
10-05-2011, 09:29 PM
My problem with the MAP2 ECU is it still uses the Gen2 ECUs resolution, fuel maps, and processors.

The MAP2, as good as it is, in the end, is still only as good as the stock ECU is IMO. You do what you want Bruce. Youd be one of the first people to make the MAP2 work properly on a 3SGTE if you did it. Knowing what I know about the 2 EFI systems, I wouldnt run one though. Just my .02

Not that I really care either way, but I don't understand this here stuff you're sayin, Luni. Obviously using a factory gen3 setup is going to work more like factory, but I don't see what would be unique about a 3sgte that causes the Mapecu (or Mapecu2) to not work on it. If it's just a matter of "a piggyback will always be a piggyback" well of course that's true, but I wasn't aware you couldn't tune a 3s with it. If all you want to do is eliminate the AFM and basically change nothing else, I don't see why not use it.

If you want something else, like genIII efi, well that's something else.

Hookecho
10-05-2011, 09:33 PM
The website lists the dyno results of a 3sgte tuned with the MAP ECU.

Luni
10-05-2011, 10:20 PM
Not that I really care either way, but I don't understand this here stuff you're sayin, Luni. Obviously using a factory gen3 setup is going to work more like factory, but I don't see what would be unique about a 3sgte that causes the Mapecu (or Mapecu2) to not work on it. If it's just a matter of "a piggyback will always be a piggyback" well of course that's true, but I wasn't aware you couldn't tune a 3s with it. If all you want to do is eliminate the AFM and basically change nothing else, I don't see why not use it.

If you want something else, like genIII efi, well that's something else.

There isnt anything unique about a 3SGTE that wont allow you to use a blackbox on. However, all blackboxes do are interpret ecu instructions, and mislead the ECU. In the end, the ECU is still running the show. And in the end, the Gen2 ECU is a primitive pile of shit.

T-spoon
10-05-2011, 10:26 PM
Oh, yah, fair enough, it is.. as they say, what it is :hehe:

The Captain
10-05-2011, 11:15 PM
And in the end, the Gen2 ECU is a primitive pile of shit.

That was a sugar coated, candy ass answer. How do you really feel?

JDM SNUKUMZ
10-06-2011, 12:01 AM
Autronic smc ems. Can be had for 500 to 600 used and hell of an ecu.

Angel

celica9303
10-06-2011, 04:33 AM
Greddy e-manage ultimate??

SMILEYOSO
10-06-2011, 07:37 PM
yea, greddy emanage ultimate ^^^. if you call greddy and talk to a tech, they can tell you how to eliminate the afm and use a map sensor. i had called them and asked that same question last year and they gave me instructions on how to use the AFM tables as pressure, but i really dont remember how tho. i know i wrote all the instructions dwn somewhere, i jux gta look fr them. seemd pretty easy.

The Captain
10-06-2011, 08:10 PM
money

Luni
10-06-2011, 09:04 PM
Emanage Ultimate is just a blackbox too. Do. Not. Want.

temperacerguy
10-07-2011, 01:25 AM
Here's an option I used on an MR2 about 10 years ago... we were going to market it, but my business partner at the time and I had a falling out.

Mustang Cobra hotwire MAF (Very little restriction)
Split second programable signal controller
5S-FE temp sensor mounted somewhere in the intake piping before the turbo.

Our cost on everything was around $300.00, the signal from the MAF was not as dirty as the AFM, and had far less restriction, which caused a bump in turbo efficiency and response. We were seeing SIGNIFICANT power increases, and smoother dyno curves. Also as the signal converter was programable for more than 20 points, you could tweak your injector pulse width through the converter.

joe's gt
10-07-2011, 04:33 AM
You might also want to check out the progress of incorporating the honda ecu onto the 3s. I remember karl, don'r know what his sn is over on the mr2 boards was in the process of doing a write up on it on the mr2 boards, I don't know whatever came about it.

Found it:
http://www.mr2oc.com/showthread.php?t=381639&highlight=honda+ecu

ChrisD
10-09-2011, 07:22 PM
Since you asked, i did experiment with an smt6 for a while. In the end I found it finicky and a bit tricky to get first set up. But, it is possible to eliminate the afm. It acts like any other piggy back in that respect. I aborted that project in favour of the power fc and am very happy I did.

The gen 3 setup is buttery smooth though. Such a great setup in comparison. I guarantee everything you hate about the gen 2 is better on the gen 3. No question.

CollapsedNut
10-09-2011, 08:20 PM
Pair of wire cutters will eliminate it pretty quick...

Luni
10-10-2011, 12:29 AM
Here's an option I used on an MR2 about 10 years ago... we were going to market it, but my business partner at the time and I had a falling out.

Mustang Cobra hotwire MAF (Very little restriction)
Split second programable signal controller
5S-FE temp sensor mounted somewhere in the intake piping before the turbo.

Our cost on everything was around $300.00, the signal from the MAF was not as dirty as the AFM, and had far less restriction, which caused a bump in turbo efficiency and response. We were seeing SIGNIFICANT power increases, and smoother dyno curves. Also as the signal converter was programable for more than 20 points, you could tweak your injector pulse width through the converter.


Need moAr info on this. Me and an inactive board member NuclearHapinesS back in the day were working on an AFM Removal Tool (we called it ART), but it never really worked. However we were trying to convert it to MAP.

How did you get the signal to read backwards? The controller? (The 3S Afm starts at 5v and reads down to 0)

temperacerguy
10-12-2011, 07:43 PM
Need moAr info on this. Me and an inactive board member NuclearHapinesS back in the day were working on an AFM Removal Tool (we called it ART), but it never really worked. However we were trying to convert it to MAP.
How did you get the signal to read backwards? The controller? (The 3S Afm starts at 5v and reads down to 0)

The split second is a simple signal converter... just set input and output, so we just made a cross over table. Also, if I remember correctly it also had a BAR correction on it, but this was 10+ years ago. Strange, we called ours the AFM-R. (Air Flow Meter-Removal) It worked EXTREMELY well, and there were MANY MANY benefits with it. Turbo spooled faster, there was an increase in turbo efficiency thus the boost pressures rose. What I was most impressed with was that the TVIS dip was eliminated. The turbo's new found efficiency compensated for the change in dynamics of the intake manifold. It was really a great product, and we could have made a fortune on it... but various things occured and the company disolved. My business partner and I both looked at each other and thought that we were taking advantage of each other, and it just didn't work out, so this went by the wayside...

ok, so I went back to split second's website. www.splitsec.com and found what we used. It's called the PSC1-004 (AFM to MAF conversion). It DOES have a BAR correction to it. It's a laptop programable signal converter. It was VERY easy to program, and we did our programing on a dyno, however the plans were to create an absolute cross-over table on a flow bench for the production units. It can be made to work with any MAF. However we found that the aftermarket MAFs made for Mustang Cobras were the least restrictive at like 4.5" in diameter with a very small hotwire sensor in the flow. Then just wire in any Toyota temperature sensor (They pretty much all had the same temp/voltage ranges back then), and the ECU has everything it needs

retail on the PSC1 is like $239, jobber was nearly half of that. The MAF was like $100 jobber, and a quality K&N air filter was like $40. So, we were looking at marketing this as a $400-$500 package, and nearly doubling our money.

Luni
10-12-2011, 07:45 PM
Bruce. Do this.

temperacerguy
10-12-2011, 07:55 PM
To be honest, there were many innovations that my partner and I did 12 or so years ago that were "cutting edge" or first to do so back then which have become common place in the MR2 community. (Such as shimless buckets, Turbo'ed Spyder and so on), I am really suprised no one has done this yet.

Luni
10-12-2011, 09:31 PM
Its called sharing of information fucker. If you have information, share it! :p

temperacerguy
10-12-2011, 10:52 PM
Um... didn't I do just that numb-nuts?

andy
10-13-2011, 12:23 AM
arduino

Hookecho
10-13-2011, 12:39 AM
Ok, this is interesting to me.

I've looked at the the PSC1-004. That I understand. Now, which Mustang Cobra MAF do I need. I see a 70mm and 80mm. They are all over ebay.

I found this 70mm one for $25

http://cgi.ebay.com/ebaymotors/70mm-mass-air-flow-94-95-Mustang-GT-Cobra-19-24lb-MAF-/220868822785?pt=Motors_Car_Truck_Parts_Accessories&hash=item336ccec701#ht_500wt_1182

And this 80mm one for $40

http://cgi.ebay.com/ebaymotors/70mm-mass-air-flow-94-95-Mustang-GT-Cobra-19-24lb-MAF-/220868822785?pt=Motors_Car_Truck_Parts_Accessories&hash=item336ccec701#ht_500wt_1182

All I should need to do this conversion is the MAF, PSC1-004, and the data cable. Am I right?

belanger33
10-13-2011, 02:09 AM
best bet moneywise is upgrading to a gen 3 ecu, injectors, MAP sensor, ignitor(a camry 4 cyl. 1992-2001 will work with that setup and cost 15 $ at the scrapyard). With that setup and your powergoals you should be perfect for a long time!

Hookecho
10-13-2011, 02:24 AM
^ Dude what? I can't get that stuff from a Camry.

Have you read this thread at all.

temperacerguy
10-13-2011, 04:08 AM
the beauty of the controller is that you can use just about any MAF you want... You will also need a temperature sensor

Hookecho
10-13-2011, 04:11 AM
Do I wire the temp sensor into the controller?

temperacerguy
10-13-2011, 04:22 AM
No, you wire it into the factory harness on the intake temp sensor circuit. The AFM has an internal air temp sensor.

Luni
10-13-2011, 06:48 AM
Bruce, you can make yourself a PnP harness by going to a junkyard and pulling the AFM connector off a bad AFM (or just using one off one you got if you got a bad one). Then you can simply wire direct to VS, VC, THA, and E2 without cutting your harness. Like I do for my Gen3 EFI swaps.

Hookecho
10-13-2011, 01:11 PM
I'm confused now. I thought the MAF sensor had a hot wire temp sensor in it. Why do I need a separate temp sensor?

temperacerguy
10-13-2011, 10:03 PM
The MAF uses a hotwire to calculate the volume of air passing through it. It cannot judge temperature, just the volume of the air passing through it.

The Temp sensor is for the ECU to do temperature compensation.

Luni
10-13-2011, 11:52 PM
Some MAFs can. The stock MAF off my Celica can. It has a thermistor and the hotwire in it.

Bruce, Just use your stock 5S temp sensor. And do what I said about the AFM plug make your own harness thingy.

pokemeintheeye
10-14-2011, 04:42 AM
Wow! Sure got my attention. When you get it all together Hookecho, please share. I'll have to go that route too just because.

alltracman78
10-14-2011, 05:14 AM
Some MAFs can. The stock MAF off my Celica can. It has a thermistor and the hotwire in it.

Bruce, Just use your stock 5S temp sensor. And do what I said about the AFM plug make your own harness thingy.

The MAF sensor and IAT sensor in your Celica are 2 different circuits.
Same basic setup as the AFM [circuit wise], it's 2 sensor combined into 1 housing.

As far as the AFM connector, you could also remove the wires from it and use a different connector so it wouldn't be as awkward.
You should only need a 4 pin connector.

joe's gt
10-14-2011, 06:40 AM
As far as a goal for just eliminating the AFM and mild modifications, I think the whole MAF conversion sounds like the best solution so far in terms of the ability to improve throttle response, acceleration, and fuel tuning in general. However, tuning timing is where real power is achieved and that is where this method falls short. Depends on what your long term goals are Bruce. If you plan on putting cams in this, you could see more significant gains from the ability to tune both fuel and timing.

Luni
10-14-2011, 06:43 AM
The MAF sensor and IAT sensor in your Celica are 2 different circuits.
Same basic setup as the AFM [circuit wise], it's 2 sensor combined into 1 housing.

As far as the AFM connector, you could also remove the wires from it and use a different connector so it wouldn't be as awkward.
You should only need a 4 pin connector.

My method is the best IMO because you dont have to hack up your stock harness for any part of it. Thats why I do it that way.

Hookecho
10-14-2011, 07:01 AM
Well, I replaced the the plug for my AFM months ago with a new one. When I did this I used quick connects on the wires. So it would be simple to pull it apart and wire on a new plug using the same method.

I'm going to give this MAF conversion a try. This could prove very beneficial to us all. First I want to buy a wideband kit. I may order a new one next week. If anyone has one they want to sell me just let me know.

alltracman78
10-14-2011, 05:06 PM
You need to learn to remove the individual pins so you don't have to cut your harness any more. :)


My method is the best IMO because you dont have to hack up your stock harness for any part of it. Thats why I do it that way.

No hacking required man :)
I'm even more against that than you are.


Get both sides of a 4 pin connector, remove the 4 pins from the AFM connector and install them in the "new" connector. Wires will hold better in the other side of a normal connector than the AFM part.

I hope that makes sense, I know it's a little clumsy.

Luni
10-14-2011, 05:34 PM
Oh, I understand what youre talking about. Id still rather just do it my way. But primarily cause thats the way Im used to. Ive never been all that great at unpinning. I dont think I have the right tool. MOST of the time I can get it out and be fine. Sometimes I just get the one thats a total PITA and it wont come out.

Funkycheeze
10-14-2011, 06:03 PM
There is also a product called the MAFT-PRO that you could use. I have seen some of the supra guys use it.

I think the gen 3 swap would be the best idea.

Hookecho
10-14-2011, 08:47 PM
You need to learn to remove the individual pins so you don't have to cut your harness any more. :)





I know how to do that. I've repined a few harnesses. The problem I had with the old AFM plug was that the connectors had corroded. So rather than finding the new connectors I cut and spliced. Does Toyota even sell those little connectors?

klapa
10-14-2011, 11:42 PM
I know how to do that. I've repined a few harnesses. The problem I had with the old AFM plug was that the connectors had corroded. So rather than finding the new connectors I cut and spliced. Does Toyota even sell those little connectors?

If you are talking about the wiring end (harness) connector - you can get that from Dr. Tweak. I bought one with pins a while back as mine has broken locks and comes loose on occasion.

BTW - a very educational thread!

l0ch0w
10-14-2011, 11:54 PM
I smell setting up a test rig!!!

Really all it boils down to is whatever voltage value is fed to the ecu at any given air flow value. If you set up say a butchered shop vacuum with a potentiometer to vary the air speed, you could in theory graph the output of the MAF, and whatever hotwire setup you were running. Then with circuitry you could get the hotwire to communicate the same way the stock MAF does to the ECU. Then you could in theory create a plug and play sensor.

klapa
10-15-2011, 12:08 AM
I smell setting up a test rig!!!

Really all it boils down to is whatever voltage value is fed to the ecu at any given air flow value. If you set up say a butchered shop vacuum with a potentiometer to vary the air speed, you could in theory graph the output of the MAF, and whatever hotwire setup you were running. Then with circuitry you could get the hotwire to communicate the same way the stock MAF does to the ECU. Then you could in theory create a plug and play sensor.

Exactly the way a piggyback works - and AFIK the AFM output is 0-5V - the problem with allot of piggybacks as far as I know from reading around is that they have a linear output to the ECU - and the actual transducer (AFM in this case) is not linear.

This is especially true when using a piggyback to interface between a wideband O2 with a 0-5V output and an ECU that expects a narrow band O2 with 0-2V output.

alltracman78
10-15-2011, 01:18 AM
AFM is 5 - 0V.
And the signal is linear [voltage, I've measured it while driving numerous times :)]

You have the O2 signals mixed up too.
O2 sensor [narrow band] is 0 - 1V.
A:F sensor [wideband] is actually a current signal, not a voltage signal. It gets converted to a voltage signal [2.5 - 4.5 IIRC] for viewing.


I know how to do that. I've repined a few harnesses. The problem I had with the old AFM plug was that the connectors had corroded. So rather than finding the new connectors I cut and spliced. Does Toyota even sell those little connectors?
I've never checked, but I would assume they do.
The pin and a short section of wire anyways.

klapa
10-15-2011, 03:28 AM
AFM is 5 - 0V.
And the signal is linear [voltage, I've measured it while driving numerous times :)]

5V-0V or 0V 5V is simply a matter of "semantics" here - what it IS is simply a potentiometer driven by an spring loaded air vane - and whether it is 0V closed or open can easily be taken into account by any data acquisition system. So - you know for a fact that the signal - the movement of the air vane and thus the output DC signal voltage - IS linear with the CFM air flow of the engine? The reason I ask is to know whether or not the ECU actually interprets the signal linearly or has some table lookup function instead.


You have the O2 signals mixed up too.
O2 sensor [narrow band] is 0 - 1V.
A:F sensor [wideband] is actually a current signal, not a voltage signal. It gets converted to a voltage signal [2.5 - 4.5 IIRC] for viewing.


Perhaps I do - I thought most narrow bands were 0-2V - yet a current mode signal is simply converted to a voltage mode signal by a resistor - and undoubtedly this is what is done in any ECU which uses a micro-controller with an A/D converter that measures voltage. I'll check the datasheet again on the Bosch wideband hat I have.

MrWOT
10-15-2011, 06:40 PM
maft pro will do toyota vam too, even has fuel pump switch so 165s can use it :)

alltracman78
10-16-2011, 02:27 AM
5V-0V or 0V 5V is simply a matter of "semantics" here - what it IS is simply a potentiometer driven by an spring loaded air vane - and whether it is 0V closed or open can easily be taken into account by any data acquisition system. So - you know for a fact that the signal - the movement of the air vane and thus the output DC signal voltage - IS linear with the CFM air flow of the engine? The reason I ask is to know whether or not the ECU actually interprets the signal linearly or has some table lookup function instead.
It's not semantics, the ECU seeing 0V or 5V from the AFM are very different circumstances. :)
It's important because most airflow sensors have a 0 - 5V signal while the AFM has a 5 - 0V signal. With most airflow sensors no airflow is 0V and max airflow is 5V with the measured airflow in between [I'm sure you understand all this, I'm trying to spell it out so others that may read this thread understand].
Toyotas AFMs use a backwards signal, 5V is no airflow and 0V is max airflow. This makes it harder to just "swap in" another type of airflow sensor.

Yes, the signal going into the ECU is an analog voltage signal, dropping voltage as airflow increases.
Inside the ECU I'm sure it is changed to a digital signal, and there probably is some type of table lookup.

l0ch0w
10-16-2011, 04:13 PM
AFM is 5 - 0V.
And the signal is linear [voltage, I've measured it while driving numerous times :)]

You have the O2 signals mixed up too.
O2 sensor [narrow band] is 0 - 1V.
A:F sensor [wideband] is actually a current signal, not a voltage signal. It gets converted to a voltage signal [2.5 - 4.5 IIRC] for viewing.

I've never checked, but I would assume they do.
The pin and a short section of wire anyways.

Yes, when you move the flapper door with your finger the signal is linear but only along its physical movement, the problem is its unlikely that if airflow increases linearly that the flapper door opens directly proportional to that. Here is a vauge example of what you might see on the test rig...

This is very vauge and numbers have just been randomly chosen for simplicity's sake...

If 10 CFM of air reads 4V
And 100CFM reads 1V
But 50CFM reads 1.7V
The graph for the flapper door is most definitely not linear, its more likely parabolic.

By graphing a large number of points at varying flow rates, you would be able to determine the actual readout of the sensor at a given CFM value.

Then by testing a hotwire MAF on the same setup at the same airflow points, you can then create a simple digital converter box that converts the voltage at any given CFM value to the desired voltage. Its less of a piggy back, and more of a signal converter for the sensor.

Its not as simple as just moving the flapper door...

alltracman78
10-16-2011, 05:59 PM
I'm sorry, I think I caused the confusion. We're talking about different things.
I was using linear [and read it as] a smooth line; an analog vs digital signal.
I wasn't using it in a math graphing sense.
In that sense I don't know if it's a linear signal. I havn't checked that.

I also wasn't giving any advice about building the actual circuit, that's something I don't know a lot about.
I just wanted the sensor signal info to be correct.

klapa
10-18-2011, 02:01 AM
I'm sorry, I think I caused the confusion. We're talking about different things.
I was using linear [and read it as] a smooth line; an analog vs digital signal.
I wasn't using it in a math graphing sense.
In that sense I don't know if it's a linear signal. I havn't checked that.

I also wasn't giving any advice about building the actual circuit, that's something I don't know a lot about.
I just wanted the sensor signal info to be correct.

Yea - I understood what you said - and why you said it - that it is not just "semantics" for many here - it is important to understand that you cannot just swap in any old sensor.

I'm just trying to learn here.

Though I will say - at least for me - converting a 5V - 0V analog signal to a 0V - 5V analog signal is a piece of cake that would cost maybe $1, and I could rig it up in about two hours - including the box!

klapa
10-18-2011, 02:07 AM
I'm sorry, I think I caused the confusion. We're talking about different things.
I was using linear [and read it as] a smooth line; an analog vs digital signal.
I wasn't using it in a math graphing sense.
In that sense I don't know if it's a linear signal. I havn't checked that.

I also wasn't giving any advice about building the actual circuit, that's something I don't know a lot about.
I just wanted the sensor signal info to be correct.

Yea - I understood what you said - and why you said it - that it is not just "semantics" for many here - it is important to understand that you cannot just swap in any old sensor. I had heard about soe of these AFM's with the reverse signal - for whatever reason I thought the had a 0V - -5V (negative polarity) signal.

I'm just trying to learn here.

Though I will say - at least for me - converting a 5V - 0V analog signal to a 0V - 5V analog signal is a piece of cake that would cost maybe $5, and I could rig it up in about two hours - including the box!

klapa
10-18-2011, 03:25 AM
Yes, when you move the flapper door with your finger the signal is linear but only along its physical movement, the problem is its unlikely that if airflow increases linearly that the flapper door opens directly proportional to that. Here is a vauge example of what you might see on the test rig...

This is very vauge and numbers have just been randomly chosen for simplicity's sake...

If 10 CFM of air reads 4V
And 100CFM reads 1V
But 50CFM reads 1.7V
The graph for the flapper door is most definitely not linear, its more likely parabolic.

By graphing a large number of points at varying flow rates, you would be able to determine the actual readout of the sensor at a given CFM value.

Then by testing a hotwire MAF on the same setup at the same airflow points, you can then create a simple digital converter box that converts the voltage at any given CFM value to the desired voltage. Its less of a piggy back, and more of a signal converter for the sensor.

Its not as simple as just moving the flapper door...

Yea - that was exactly my point - that the fact that the AFM is a spring loaded air vane driving (moving) a potentiometer it is very doubtful that it is linear as far as the voltage vs. airflow curve. The spring prolly does not have anywhere near a linear curve of deflection vs. pressure. The tighter the spring compression - the less movement vs. pressure (force of the air) - this is a non-linear transfer function.

This might have been at least partially compensated for with a "taper" on the potentiometer - much like potentiometers for audio level control have a logarithmic taper.

Yet it is true that perhaps this is not the thread for thoughts such as these.

temperacerguy
10-19-2011, 03:20 AM
Wow....

Or you could have just asked someone that's done it... (Me)

No the airflow vs voltage is not linear.
No, it's not just a direct conversion, not just voltage inverted between the two sensors, they also have different voltage vs flow curves.

Yes, setting up a blower motor with a pentiometer is an easy way of setting up a flow bench.... However running them one at a time will cause variences that will effect the outcome. (as with the leaf blower X voltage does not mean X airflow all the time. I would recomend testing them in-line with each other and plotting the voltage outputs simultaniously. I would also not use a shop vac as most non-commercial units don't have the CFM to max out the sensors. However most leaf blowers do.

So an ideal test bench would be Leaf Blower->at least 1 meter of feed pipe to minimize turbulance (MAF or AFM)->1 meter of pipe->(The other MAF or AFM)-> 1 meter of pipe to exit.

Run the leaf blower from minimum flow to obtain any movement from the AFM, and measure the every .2V and the coresponding voltage on the MAF.
Then swap the positions of the MAF and AFM, while theoretically they should give the same results, there will be small variances. This will give you a good range for a starting point.

rizin
10-19-2011, 04:02 AM
Been following this I like the different road. Do you think gluing straws in a pipe would help with the turbulence better then just the distance of pipe? I know the process of getting the straws straight might be a pain but could be a way to help with variances with the sensors.

temperacerguy
10-19-2011, 04:05 AM
I would not glue straws in there as the glue and the straws themselves would cause a restriction.

Hookecho
10-19-2011, 03:11 PM
Just use something honeycomb shaped to keep a laminar flow.

klapa
10-20-2011, 12:09 AM
Wow....

Or you could have just asked someone that's done it... (Me)

No the airflow vs voltage is not linear.
No, it's not just a direct conversion, not just voltage inverted between the two sensors, they also have different voltage vs flow curves.



Yea - I kinda figured that - if the air vane movement vs. airflow is NOT linear - then simply inverting the signal would place the "flattening" at the wrong end of the curve. The ECU would not be getting the proper feedback as one would think it uses a lookup table function to would compensate for the less linear deltaV-vs.-deltaCFM at high flow as opposed to low flow.

As an example - I would imagine that a simple spring loaded air vane would produce less deflection for a proportional increase in airflow as it neared maximum deflection - unless is had some sort of "calibrated" variable rate spring. Thus if the potentiometer has a linear taper, then the change in voltage would be in less proportion to the change in airflow as the flow increased. Thus - simply inverting the signal to the ECU transfers the flattening of the curve from the high flow end to the low flow side.

Good idea with the leaf blower as a "flow bench" - ghetto but functional.

temperacerguy
10-20-2011, 12:47 AM
Just use something honeycomb shaped to keep a laminar flow.

Getting 10 feet of PVC sewer pipe is both cheaper, and easier to obtain than 3" thick expanded honeycomb material of any type.

I don't see why people make things so complicated


Good idea with the leaf blower as a "flow bench" - ghetto but functional.

Why spend excess money and time on something that will be used for only a couple hours then thrown away.

If I were marketing these, making a test bench is a simple, yet expensive thing to do. a large, high flow fan with built in potentiometer, a short section of piping with bi-directional cross vains to reduce turbulence, a callibrated flow meter, another short section of piping with bi-directional vains, then the meter you are measuring. This rig would cost at least $1000 to make, and would take a few weeks to assemble the components and to assemble.

You don't need to know what volume of air is going through each meter to give X voltage... You just need to know what voltage is given for the same volume of air. Y(airflow) = X1(voltage of AFM) = X2 (voltage of MAF) once you plot the X1=X2 table, you punch that into your signal converter and there you go.

As for how the stock computer works... There is a X voltage from AFM = Y injector pulse width. Once that base calculation is made, the computer then does correction off of it from the other various parameters... RPM, air intake, O2, MAP, and so on.

Hookecho
10-20-2011, 01:53 AM
Wow....

Or you could have just asked someone that's done it... (Me)


Getting 10 feet of PVC sewer pipe is both cheaper, and easier to obtain than 3" thick expanded honeycomb material of any type.

I don't see why people make things so complicated



I don't understand why you're being pompous.

We get the fact that you have some experience with the research and development on this matter. I myself have no clue of what you are talking about. It's all over my head. So I did some googling to better understand it. Please don't mistake my suggestion for my being versed on the matter. I'm just trying to learn enough on my own to be able to contribute to the conversation.

klapa
10-20-2011, 03:58 AM
You don't need to know what volume of air is going through each meter to give X voltage... You just need to know what voltage is given for the same volume of air. Y(airflow) = X1(voltage of AFM) = X2 (voltage of MAF) once you plot the X1=X2 table, you punch that into your signal converter and there you go.

Well - that IS a very good point - we don't really need to know the ABSOLUTE VALUE of airflow - just to compare the readings between the transducers and to determine the transfer function - and thus finally to build the table - or analog transfer function - such that we need to convert the signal in such a way that the ECU performs as we want it to.

To do such - we would only need to compare RELATIVE measurements between differing sensors- as opposed to ABSOLUTE measurements. Yet to attain usable results with such relative measurements between specimens would entail REPEATABILITY of the measurement apparatus - and this would include REPEATABILITY of being able to generate the same flow for any "set-point" - (however you would determine such) - you choose to use with the leaf blower.

temperacerguy
10-20-2011, 04:00 AM
Not being pompous, this is just a very simple thing that seems people are trying to over complicate.

You can get the final voltages just by inverting the signal voltages, and then calibrating on a dyno... But people want to flow bench it, and over think the issue. with complicated test rigs and the like. The theory is simple, the application is simple. The Theory has been proven to be sound, so now all it takes is building it.

klapa
10-20-2011, 04:12 AM
I want to try your leaf blower and pipe idea - I have a spare AFM that is I think a 4th GEN - it would be a good exercise to try this idea because it is good - while it might not be totally "quantitative" - it might be "qualitative" enough to do the job.

It fits into my plans because my old leaf blower bit the dust last year - I need a new one.

Did I mention that it is "cheap" (cost effective) :)

I need to re-read your post - how do you determine your flow "set-points" with the leaf blower?

temperacerguy
10-20-2011, 04:16 AM
Well - that IS a very good point - we don't really need to know the ABSOLUTE VALUE of airflow - just to compare the readings between the transducers and to determine the transfer function - and thus finally to build the table - or analog transfer function - such that we need to convert the signal in such a way that the ECU performs as we want it to.

To do such - we would only need to compare RELATIVE measurements between differing sensors- as opposed to ABSOLUTE measurements. Yet to attain usable results with such relative measurements between specimens would entail REPEATABILITY of the measurement apparatus - and this would include REPEATABILITY of being able to generate the same flow for any "set-point" - (however you would determine such) - you choose to use with the leaf blower.

All you need is relativity. What you are using to make your set point is not the voltage of the leaf blower, or anything with the leafblower itself, you are measuring the MAF output relative to the AFM output. So here's all you need to do...

hook up the rig. adjust the voltage to the leafblower till you read 4.8v from the AFM. Record voltage of the MAF
adjust the leafblower till you read 4.6 from the AFM, record the MAF voltage.

Do this same thing every .2 volts till you have maxed out the AFM

Once you have the cross over table, the only reason you would need to repeat this test is if you changed the MAF to a different size/style.

temperacerguy
10-20-2011, 04:37 AM
I want to try your leaf blower and pipe idea - I have a spare AFM that is I think a 4th GEN - it would be a good exercise to try this idea because it is good - while it might not be totally "quantitative" - it might be "qualitative" enough to do the job.

It fits into my plans because my old leaf blower bit the dust last year - I need a new one.

Did I mention that it is "cheap" (cost effective) :)

I need to re-read your post - how do you determine your flow "set-points" with the leaf blower?

Cost effective is highly key...

The set points of the leaf blower are "Whatever it takes" to match "x" voltage on the AFM. No real need to destroy an old leaf blower... Hell, you can use a leaf blower plugged into an outlet that's using a rotary dimmer switch that has been wired to an outlet...

If I were making the flow bench, here's what I would do...

Take your leaf blower, and mount it to a base of some sort that keeps the outlet of the leaf blower approximately 1" off the bench.
Mount the outlet of the leaf blower to a PVC pipe with a diameter similar in size to the AFM inlet
couple the leafblower to the PVC pipe, couple the PVC pipe to the AFM.

Now for wiring it all up...

Buy a standard 120 outlet
Buy a dimmer switch
buy a dual module surface mountelectrical box
buy 2+ground plug
buy 1' 12-2+ground romex

Mount dual module electrical box to the leafblower base
Run plug wire into electrical box
Wire plug to dimmer switch inlet (dont forget ground)
wire output of dimmer switch to the electrical outlet (don't forget ground)
Mount dimmer and outlet to the electrical box

Test rig by plugging in a cheap lamp into the outlet, then plugging the plug into your standard household outlet.

Adjust voltage with the dimmer switch and see that your lamp increases and decreases intensity.
unplug lamp, plug in leaf blower. Turn leafblower to "on" position,

Now, not sure how good the motors in leafblowers are, but you might need to crank the voltage UP to get the leafblower to start, then reduce voltage to obtain low flows... also many times electrical motors don't like to run at low speeds, so make sure not to run it at too low of a speed for too long or you may overheat and burn out the motor.

temperacerguy
10-20-2011, 05:17 AM
ok... Did some research...

Most leaf blowers flow in the 450 cfm range.. At 7000 RPM at 1.5 BAR (approximately 7.7psi) and the efficiency of our engines, the Turbo should be ingesting about 480cfm. So you should be able to get most of the way with a standard leaf blower, and you can extrapolate the remainder of the graph from what you recorded earlier.

MrWOT
10-20-2011, 08:05 AM
You could also put an MAF with a known transfer function inline, and record the voltage vs. the vane meter.

joe's gt
10-23-2011, 07:01 PM
Its critical that you get the pipe lengths long enough to have fully developed laminar flow. When fluid enters a pipe it sticks to the side walls and only develops a uniform velocity profile after a certain length. This length can easily be determined by the equation EL (entrance length) = .06 x Reynolds Number x ID of pipe.

The dimensionless Reynolds number for laminar flow is around 2100. You could more accurately figure out the Reynolds number particular to the specific airflow device you are using by the equation: Reynolds Number = Density of Air x Velocity of air x ID of pipe / viscosity of air. The velocity of air can be determined by the CFM of whatever device you are using divided by the exit area. I would also convert CFM to CFS (cubic feet per second) because most of the air flow calculations involving velocity are done in terms of seconds.

So if you are setting up the flow meters in series, after the VAM, I would do the minimum entrance length of pipe again so the air can develop a full uniform velocity profile after going through that restriction.

Luni
10-23-2011, 07:56 PM
I love you engineers on my site. :). Good stuff Joe.

l0ch0w
10-23-2011, 09:54 PM
o jeez... look what i started...


I think I would look at replacing the leaf blower motor with a comperable powered DC one (ebay ~$50) and a controller board would even things out much better.

The most important areas to graph would most certainly be the ones at very low (Idle) air speeds... And these might be very difficult to achieve on the leaf blower's AC motor.

I dont see why flow turbulence is an issue at all, you should be mounting the leaf blower with the suction end sucking air through the maf... just as the engine does it... You could even install your choice of air intake on it for consistency....

I dont see it being necessary to mount them serially... Only if you determine that you have poor repeatabilty with your controller would it be necessary...

For example, if you find that the stock MAF reports two different voltages when say 10V is applied to the motor in two separate tests, then you could assume that the motor or the blower, or the voltimeter has a repeatability problem (which I find extremely unlikely...)

EDIT:

I did some math, if you happened to own an anemometer or if you bought one, if it could measure up to about 65MPH you would be set. Given a 4 inch pipe, 480cfm through a 4" pipe would come out to be about 61.36mph... I guess you could increase the pipe diameter too, a 6 inch pipe would yield lower air speeds...

klapa
10-25-2011, 04:42 AM
Its critical that you get the pipe lengths long enough to have fully developed laminar flow. When fluid enters a pipe it sticks to the side walls and only develops a uniform velocity profile after a certain length. This length can easily be determined by the equation EL (entrance length) = .06 x Reynolds Number x ID of pipe.

The dimensionless Reynolds number for laminar flow is around 2100. You could more accurately figure out the Reynolds number particular to the specific airflow device you are using by the equation: Reynolds Number = Density of Air x Velocity of air x ID of pipe / viscosity of air. The velocity of air can be determined by the CFM of whatever device you are using divided by the exit area. I would also convert CFM to CFS (cubic feet per second) because most of the air flow calculations involving velocity are done in terms of seconds.

So if you are setting up the flow meters in series, after the VAM, I would do the minimum entrance length of pipe again so the air can develop a full uniform velocity profile after going through that restriction.

Of Course - we are all a little "short on the bucks" here!

I would like to set something up - I have data acquisition equipment such that I can make computerized measurements at a high speed sample rate and save the data for analysis.

Now - I don't have any way to actually MEASURE raw flow of air - so just for a moment I would like to combine some of the above posts here - towards a goal - and wonder if the effort would be worthwhile.




You could also put an MAF with a known transfer function inline, and record the voltage vs. the vane meter.



So we have some sensor that is accurate enough and has a "known transfer function" that we could rob off a junkyard car? Keeping in mind - we are not trying to be NIST here - at least to start off - just to get some sensor that is ~5X more accurate than an AFM - and some sensor that we can equate a voltage (or current) to ABSOLUTE airflow.




ok... Did some research...

Most leaf blowers flow in the 450 cfm range.. At 7000 RPM at 1.5 BAR (approximately 7.7psi) and the efficiency of our engines, the Turbo should be ingesting about 480cfm. So you should be able to get most of the way with a standard leaf blower, and you can extrapolate the remainder of the graph from what you recorded earlier.



Whatever - my "setpoint" at least to start - would be the RPM of a gasoline engine leaf blower - I would use a pickup coil on the sparkplug lead to measure that. The "reference sensor" would tell us the actual flow.

In any event - it is true that the "reference chamber" (the flow pipe used for measurement) must be designed correctly - and I think JoeGT seems to have a good a grip on that.

Aside from all that - I don't know the difference between a MAP sensor and a MAF sensor - I think the terms are used interchangeably - or are these different sensors?

Anybody have links to datasheets?

temperacerguy
10-25-2011, 06:36 AM
MAP means "Manifold Absolute Pressure" and measures the pressure of a manifold
MAF is "Mass Air Flow" and measures the flow of air via different means, but most typically via a hotwire.

klapa
10-25-2011, 06:54 AM
MAP means "Manifold Absolute Pressure" and measures the pressure of a manifold
MAF is "Mass Air Flow" and measures the flow of air via different means, but most typically via a hotwire.

Well - I knew that - at least insofar as what the acronyms mean - but what does that mean - really - in terms of what transducer to use for replacing that big ole' clunky primitive AFM?

Again - can anybody point me to a datasheet for any (either) of these transducers?

I realy would like to set up a test - just for shits and grins. I KNOW about electrical measurements - and I KNOW about REPEATABILITY - as I have quite a bit of experience in the "measurement field".

So - anybody game to help?

I have access to multitudes of measurement apparatus at my work - I have a desire - and I am not "shy" to take a chance.

I NEED information, to wit:

Let us first define if this is an exercise worthy of our time.

If so - let us next define the transducer we would use as a comparison - or a reference.

Finally - let us define the "pipe" - the test chamber.

I will build it all - and measure it all.

rizin
10-25-2011, 07:20 AM
Don't forget the VAF which is what the 2nd gen 3sgte has.

klapa
10-25-2011, 07:38 AM
Don't forget the VAF which is what the 2nd gen 3sgte has.

Well - gee thanks - that explains allot, lol......

temperacerguy
10-26-2011, 04:24 AM
OK, you can't compare AFM/MAF based fueling strategy vs Speed Density (MAP) fuel injection.... So, no MAP and MAF are not comparable.

Basically with an AFM/MAF system, the basis for fueling is an actual measurement of how much air is entering the engine.

with a Speed Density system, the fueling is based off of the load (MAP reading) and RPM, calculating the air being injested knowing the volume/size of the engine...

Now, you will not be able to use the input voltage of your flow source as a basis for flow. Because the flow is highly variable based off of restriction.

Look at a vaccum cleaner. if you cup the end of the cleaner, the motor will spin out of control, and voltage will spike, yet flow be nothing... Similar principal with the fan/restriction/voltage issue.

joe's gt
10-27-2011, 06:03 AM
Of Course - we are all a little "short on the bucks" here!

I would like to set something up - I have data acquisition equipment such that I can make computerized measurements at a high speed sample rate and save the data for analysis.

Now - I don't have any way to actually MEASURE raw flow of air - so just for a moment I would like to combine some of the above posts here - towards a goal - and wonder if the effort would be worthwhile.

No need to measure the flow of air. Just use the maximum flow rate of whatever device you are using, or if you don't know that, just use the generalized Reynolds number of 2100 and ignore flow rate completely to calculate the minimum entrance length of pipe needed. As long as your length of pipe is greater than the entrance length before and after each sensor you will have a uniform velocity profile and a much more reliable reading on your sensors.

temperacerguy
10-27-2011, 06:18 AM
No need to measure the flow of air. Just use the maximum flow rate of whatever device you are using, or if you don't know that, just use the generalized Reynolds number of 2100 and ignore flow rate completely to calculate the minimum entrance length of pipe needed. As long as your length of pipe is greater than the entrance length before and after each sensor you will have a uniform velocity profile and a much more reliable reading on your sensors.

Joe, What I believe Klapa is trying to do is make a repeatable test rig that he can set for various flow rates then measure the sensor voltage output.

Personally I believe this is overkill and making things more complex than they need to be. All that needs to be done is to simultaneously measure the output voltages of the AFM and MAF for various different flow points.

klapa
10-29-2011, 02:37 AM
Joe, What I believe Klapa is trying to do is make a repeatable test rig that he can set for various flow rates then measure the sensor voltage output.

Personally I believe this is overkill and making things more complex than they need to be. All that needs to be done is to simultaneously measure the output voltages of the AFM and MAF for various different flow points.

Well - time will tell - cuz I am going to set something up.

Maybe that would be overkill to measure sensor output at various flow rates - yet another thing to consider here is that we are trying to replace one sensor with another while still using an engine computer that expects the original transducer.

I don't think this is exactly "simple" - as one transducer will have a different transfer function than the other - for example maybe the AFM might "flatten out" the curve of flow vs. output voltage at high flow rates while the MAP/MAF might not do such on the same curve. We might assume here that the ECU has a "lookup table" function that is tailored to the AFM - and thus we would need to shape the MAP/MAF output to mimic the AFM output if we would wish to use the same ECU.

Aside from such complications - which admittedly might not be necessary - what needs to be done is to TEST the the darn thing and post some comparison data.

I will start by buying some piping such as I have on my FMIC - I would think this would be fairly inexpensive. I have a "guinea pig" AFM that came with the 3SGE engine I bought - I think it is off a ST165. I will look around to find some car makes that have MAF sensors and MAP sensors - and rob some of these from junk yards.

I don't know much about "tuning" - but I DO know about measurement - and I have access to some pretty sophisticated calibrated instrumentation for electrical measurements.

temperacerguy
10-29-2011, 07:23 AM
Again. I really think you are not seeing the forest for the trees. Ive had too little sleep to explain what I mean here. Tomorrow I will post a lengthy explanation. I don't doubt you have tons of experience with measurent. But we don't need to know the absolute functions of each airflow device, we just need to know the relative outputs of each device to each other.

temperacerguy
10-29-2011, 04:44 PM
OK, first let's talk about the fuel injection system of the 3S-GTE (and all toyota vehicles with an AFM). They use a version of Bosch LU-Jettronic style of fuel Injection licenced by Bosch. This calculates fueling (in a vastly simplified explaination) by the following.

It looks at the voltage of the AFM, and based on a look-up table, calculates the airflow coming in. and then creates a base fueling calculation.
It then looks at the temperature, and adjusts the fueling
it looks at the manifold pressure, and adjusts the fueling
it looks at the RPM and adjusts the fueling
it looks at the O2 sensor, and fine tunes the fueling.

So, yes there is a look-up table that it uses to calculate airflow based on a matched calculation to the AFM the system was designed for.

So, now we know how the system works...

Now lets look at the AFM There are two circuits the first is the flow circuit. The 2nd is the temperature.
All the flow circuit is, is a petentiometer connected to a flapper door that is spring loaded. As air is sucked through the AFM, the airflow pulls the flapper door open further. the more airflow, the more the door is pulled opened adjusting the potentiometer, sending a lower voltage to the ECU. The airflow can be calculated by the drag curve of the vane. Each AFM is calibrated on a flow bench to meet a standard flow. This calibration is stamped on each AFM.
There is a temperature sensor integrated into the AFM (and protrudes into the airflow),

Now, is the flow vs signal linear? I highly doubt it. I am guessing that it's more progressive. The big downfall of the AFM is that there is tons of drag and resistance with this design, which is what we are trying to get rid of.

Now, the way a MAF works is by suspending a heated wire in a tube. As the air passes through, it cools the wire, which lowers it's resistance, which allows more current to pass through and thus airflow can be calculated by the voltage passed back through the wire.

temperacerguy
10-29-2011, 05:23 PM
OK...

Now. you can test the voltage Vs flow of both of these meters by using a calibrated flow meter where the transfer function of the meter is known. in-line with the meter you are testing.

You use a airflow source, flow through the calibrated flow meter, then through the meter you are testing. You increase the airflow till the calibrated flow meter matches the velocity you are testing, then measure the output of the meter you are testing. You repeat this process for each flow set point that you wish to record. This will then graph out to a absolute flow vs voltage curve for this sensor.

You repeat this process for both meters. Thus you will have an absolute flow vs output voltage for both meters... you then can create a cross over table, which can be input into the signal converter.

This would be the absolute, BEST way of testing the sensors and building the crossover table. The problem is that it's HUGELY expensive to purchase a calibrated flow meter for the volume of air we are talking about.

Now, it sounds like the way that Klapa wants to test things is by hooking up the flow source(leaf blower) to a ammeter to read the amp draw of the flow source, and to create a amp (of flow source) to voltage output (of the two sensors) table. Then match amp draw to each other and thus have a comparison table between the two sensors. This would work IF amp draw of the flow source was equal to airflow. Unfortunately it is not. For the same amp draw of the flow source, the actual airflow is highly variable depending upon the restriction. This would not be a problem if the sensors we are measuring have the same or highly similar restrictions. But the whole point we are changing the meters out are because the airflow restrictions are vastly different.

Now the reality of the situation is that we do not need to know the actual airflow vs voltage for each sensor. We just have to know the voltages of each sensor for the same airflow.

By hooking up the sensors IN LINE with each other. the airflow going through one sensor will match the airflow going through the other sensor. If we measure the simultanious voltages of both sensors for various airflows, we can create a cross over table that will match the airflow vs voltage outputs of both sensors.

Here is a graph showing theoretical voltage vs airflow curves for both sensors:
http://i236.photobucket.com/albums/ff73/temperacerguy/airflowMAF.jpg

If we set the airflow to F1 (by adjusting the amps to the flow source) we can then see that the Voltage of the AFM is 4.9, and the voltage for the MAF is .7
http://i236.photobucket.com/albums/ff73/temperacerguy/airflowF1.jpg

We then set the airflow to F2 (by adjusting the amps to the flow source), we measure the voltage of the AFM as 4.7 and the MAF is 1
http://i236.photobucket.com/albums/ff73/temperacerguy/airflowF2.jpg

We then set the airflow to F3 (by adjusting the amps to the flow source), we measure the voltage of the AFM as 4.3 and the MAF is 1.7
http://i236.photobucket.com/albums/ff73/temperacerguy/airflowF3.jpg

We continue adjusting the airflow, and measuring simultanious voltage data points for the AFM and the MAF. we will then build a conversion table.

AFM MAF
5V = 0V
4.9V = .7V
4.7V = 1V
4.3V = 1.7V
3.5V = 2.7V
and so on and so on.

Once you have this conversion table. you input the conversion table into the split second signal converter, and you are all done.... The computer will see the correct voltage for the airflow that's being measured.

temperacerguy
10-29-2011, 05:26 PM
Now, Klapa... If you have measurement equipment that can simultaniously measure voltage of multiple sources (and cross refference them based on time)... Then you can build this lookup table in under 3 minutes...

Set your equipment to measure the voltage output of both MAF and AFM sensors. start your airflow source at the lowest setting.... then slowly increase the amps till you have maxed out your source. That will create FAR more data points than you will need to set up your signal converter.

OK, so I pulled my old split second software... it turns out that it has inputs for ever .1v so between 0 and 5V, you have 50 data points :-) that is TONS of resolution, which should allow a perfect 1:1 replacement, and also allow for some fine tuning beyond what the AFM is capable of.

Carolina91GT-S
10-29-2011, 08:47 PM
I like the idea of setting up this test bench for a number of reasons. I am willing to help. Klapa, we'll discuss it when I see you next week.

All that being said....I bought a MAP ECU 1 used on the supraforums for $150. Eliminates the crappy MAF and even allow you to change the maps and get better performance. Once the MAPECU 3 came out the used MAPECU 1 started to get cheap. It is sitting on a shelf in my office right now, i haven't had a chance to use it. My project car is in the paint shop.

~Chris

l0ch0w
10-29-2011, 11:16 PM
Now, you will not be able to use the input voltage of your flow source as a basis for flow. Because the flow is highly variable based off of restriction.

Look at a vaccum cleaner. if you cup the end of the cleaner, the motor will spin out of control, and voltage will spike, yet flow be nothing... Similar principal with the fan/restriction/voltage issue.

Ahh.... forgot about that...

What about a simple RPM meter on the motor then????

Im only trying to think of simpler ways to avoid having to use a pitot tube or an anemometer, or air flow meter at all...


I like the idea of setting up this test bench for a number of reasons. I am willing to help. Klapa, we'll discuss it when I see you next week.

All that being said....I bought a MAP ECU 1 used on the supraforums for $150. Eliminates the crappy MAF and even allow you to change the maps and get better performance. Once the MAPECU 3 came out the used MAPECU 1 started to get cheap. It is sitting on a shelf in my office right now, i haven't had a chance to use it. My project car is in the paint shop.

~Chris

I just dont like the idea of straight up piggy backs... especially ones that might have a difficult time replicating the stock voltage values, which we have proven are non-linear. The trick here is creating some sort of analog converter (split-pi topology?) that can follow a graphical input and output based on flow.

l0ch0w
10-29-2011, 11:37 PM
OK...

Now. you can test the voltage Vs flow of both of these meters by using a calibrated flow meter where the transfer function of the meter is known. in-line with the meter you are testing.

You use a airflow source, flow through the calibrated flow meter, then through the meter you are testing. You increase the airflow till the calibrated flow meter matches the velocity you are testing, then measure the output of the meter you are testing. You repeat this process for each flow set point that you wish to record. This will then graph out to a absolute flow vs voltage curve for this sensor.

You repeat this process for both meters. Thus you will have an absolute flow vs output voltage for both meters... you then can create a cross over table, which can be input into the signal converter.

This would be the absolute, BEST way of testing the sensors and building the crossover table. The problem is that it's HUGELY expensive to purchase a calibrated flow meter for the volume of air we are talking about.

Now, it sounds like the way that Klapa wants to test things is by hooking up the flow source(leaf blower) to a ammeter to read the amp draw of the flow source, and to create a amp (of flow source) to voltage output (of the two sensors) table. Then match amp draw to each other and thus have a comparison table between the two sensors. This would work IF amp draw of the flow source was equal to airflow. Unfortunately it is not. For the same amp draw of the flow source, the actual airflow is highly variable depending upon the restriction. This would not be a problem if the sensors we are measuring have the same or highly similar restrictions. But the whole point we are changing the meters out are because the airflow restrictions are vastly different.

Now the reality of the situation is that we do not need to know the actual airflow vs voltage for each sensor. We just have to know the voltages of each sensor for the same airflow.

By hooking up the sensors IN LINE with each other. the airflow going through one sensor will match the airflow going through the other sensor. If we measure the simultanious voltages of both sensors for various airflows, we can create a cross over table that will match the airflow vs voltage outputs of both sensors.

Here is a graph showing theoretical voltage vs airflow curves for both sensors:
http://i236.photobucket.com/albums/ff73/temperacerguy/airflowMAF.jpg

If we set the airflow to F1 (by adjusting the amps to the flow source) we can then see that the Voltage of the AFM is 4.9, and the voltage for the MAF is .7
http://i236.photobucket.com/albums/ff73/temperacerguy/airflowF1.jpg

We then set the airflow to F2 (by adjusting the amps to the flow source), we measure the voltage of the AFM as 4.7 and the MAF is 1
http://i236.photobucket.com/albums/ff73/temperacerguy/airflowF2.jpg

We then set the airflow to F3 (by adjusting the amps to the flow source), we measure the voltage of the AFM as 4.3 and the MAF is 1.7
http://i236.photobucket.com/albums/ff73/temperacerguy/airflowF3.jpg

We continue adjusting the airflow, and measuring simultanious voltage data points for the AFM and the MAF. we will then build a conversion table.

AFM MAF
5V = 0V
4.9V = .7V
4.7V = 1V
4.3V = 1.7V
3.5V = 2.7V
and so on and so on.

Once you have this conversion table. you input the conversion table into the split second signal converter, and you are all done.... The computer will see the correct voltage for the airflow that's being measured.

That thing is $400.... I think it might be possible to make your own...

The point is to be able to do this for substantially cheaper than a gen 3 EFI...

joe's gt
10-29-2011, 11:40 PM
I happen to agree with Temperacerguy that airflow is irrelevant, in fact it is MORE accurate to completely ignore it because it is highly variable and less repeatable in two different tests than exposing the flow meters to the same air flow at the same time. Klapa, by hooking them up in series, you are in effect accounting for the differences in transfer functions immediately rather than setting up two different tests and increasing variables by adding back in both the different transfer functions of the two devices and variability of the airflow. However, due to the restriction of the VAM and the much lower restriction of the MAF, I would put the MAF first in the series of hooking them up so as to allow the maximum original airflow to reach the VAM.

l0ch0w
10-29-2011, 11:54 PM
I never said that it was important... i just dont like the unrepeatability of the whole thing, and the different potential airflow characteristics of the test rig if set up back to back.... You need some sort of stepped value to follow in order to create a line of best fit.

temperacerguy
10-30-2011, 12:14 AM
What about a simple RPM meter on the motor then????

Nope, due the the flow dynamics of a fan, RPM does not equal flow (Again, think of cupping your hand over a vacuum) with higher restriction, RPMs increase while flow decreases due to cavitation.



Im only trying to think of simpler ways to avoid having to use a pitot tube or an anemometer, or air flow meter at all...

Already done..... see my idea of in-line MAF/AFMs




I just dont like the idea of straight up piggy backs... especially ones that might have a difficult time replicating the stock voltage values, which we have proven are non-linear.

This converter has ZERO problem replicating stock voltage values... It's not a linear look-up table. It will output ANY voltage for any given voltage... The damn things was designed to do EXACTLY what we are trying to do... It's been used on BMWs for many many years now.

temperacerguy
10-30-2011, 12:19 AM
That thing is $400.... I think it might be possible to make your own...

The point is to be able to do this for substantially cheaper than a gen 3 EFI...

List price on it is $322. However I don't know any idiot that pays list price. I am sure you can find a dealer willing to sell it for $250ish.

alltracman78
10-30-2011, 12:36 AM
just to get some sensor that is ~5X more accurate than an AFM
I have a beef with this, in a general sense.
An AFM is actually more accurate both while stock and with mods [especially with mods] than either a MAP sensor or a MAF sensor.
The problems come when the AFM is maxed out [ you would have the exact same problem with a MAP or MAF sensor] or when it gets old [MAP/MAF don't have this problem so much].
I'm not trying to start an argument or sidetrack this [it's good stuff, I'm learning too :)], I just hate it when I see that and wanted to correct it. :)



Aside from all that - I don't know the difference between a MAP sensor and a MAF sensor - I think the terms are used interchangeably - or are these different sensors?

Dunno if this was covered yet, it was in the quote so I figured I'd address it.
A MAP sensor reads air pressure inside the intake manifold. It's not a very adaptable system, for complete accuracy you need to reprogram the fueling once you change pretty much anything in regards to airflow in the engine.
A MAF sensor reads air FLOW through the intake. It's a more accurate system, especially once you start modding. Because it measures flow it can adapt to changes better than the MAP system. However, if you change the intake tract the MAF is mounted in it screws up the readings because the MAF doesn't actually measure all the airflow through the intake, it only measures a calculated portion of it, and assumes the rest. Once you change the flow properties around the MAF it changes the calculated portion, and unless the ECU knows this it can't adjust to the changed flow. This is why OBDII cars with aftermarket intakes and MAFs throw codes sometimes.
IIRC technically an AFM is a MAF, so it would be the exception, because it actually measures the complete airflow coming into the engine, even if the intake is changed.

temperacerguy
10-30-2011, 12:39 AM
I happen to agree with Temperacerguy that airflow is irrelevant, in fact it is MORE accurate to completely ignore it because it is highly variable and less repeatable in two different tests than exposing the flow meters to the same air flow at the same time. Klapa, by hooking them up in series, you are in effect accounting for the differences in transfer functions immediately rather than setting up two different tests and increasing variables by adding back in both the different transfer functions of the two devices and variability of the airflow. However, due to the restriction of the VAM and the much lower restriction of the MAF, I would put the MAF first in the series of hooking them up so as to allow the maximum original airflow to reach the VAM.

Thanks for the agreement... It means that I explained my reasoning well.

Theoretically it shouldn't mater which meter goes first. the pressure (or pressure drop) of air going through one meter will in no way affect the other meter. These sensors measure FLOW and FLOW alone. If there is any density of air change by pressure drop through one meter, the other will not be affected. As long as the rig is sealed between the two sensors, X volume going through the test rig would be flow matched at any point in the test rig regardless of pressure.

As long as the turbulence of the restriction is accounted for, the other meter should not be affected. (This is why I recomended the length of pipe between the sensors) in fact the MAF is so sensitive to turbulance (due to it's small sampling tube) that the manufacturers recomend placing them either behind a filter box, or 9" of tubing to reduce the turbulence to the sampling tube.

Speaking of the test rig... the diameter of the tubing should be equal to the input/output of the sensors you are using to reduce restriction in the test rig itself. Thus you should be using approximately 4" tubing. 2" intercooler tubing will cause too much restriction, and too much turbulance in the transitions from tubing to sensor to create a reliable flow signal. This is why I recomended PVC tubing as it's cheap and easy to get.

temperacerguy
10-30-2011, 12:50 AM
I have a beef with this, in a general sense.
An AFM is actually more accurate both while stock and with mods [especially with mods] than either a MAP sensor or a MAF sensor.
The problems come when the AFM is maxed out [ you would have the exact same problem with a MAP or MAF sensor] or when it gets old [MAP/MAF don't have this problem so much].
I'm not trying to start an argument or sidetrack this [it's good stuff, I'm learning too :)], I just hate it when I see that and wanted to correct it. :)


Dunno if this was covered yet, it was in the quote so I figured I'd address it.
A MAP sensor reads air pressure inside the intake manifold. It's not a very adaptable system, for complete accuracy you need to reprogram the fueling once you change pretty much anything in regards to airflow in the engine.
A MAF sensor reads air FLOW through the intake. It's a more accurate system, especially once you start modding. Because it measures flow it can adapt to changes better than the MAP system. However, if you change the intake tract the MAF is mounted in it screws up the readings because the MAF doesn't actually measure all the airflow through the intake, it only measures a calculated portion of it, and assumes the rest. Once you change the flow properties around the MAF it changes the calculated portion, and unless the ECU knows this it can't adjust to the changed flow. This is why OBDII cars with aftermarket intakes and MAFs throw codes sometimes.
IIRC technically an AFM is a MAF, so it would be the exception, because it actually measures the complete airflow coming into the engine, even if the intake is changed.

Thank you for bringing up this point:

I have been mentioning at times (and in my last post) about how turbulance will affect the MAF reading. It's why when we first designed this product years ago, we included a cone style filter which had a large filter area, and a bell shaped base with approximately 8"s from the filter to the entrance of the MAF. We also used the mustang MAF because the Mustang's MAF is designed by the factory to be extremely close to a cone style factory filter. Thus creates a very stable airflow pattern by the design of the MAF itself.

Import MAF sensors for the past 12 yrs or so, are typically just a small hotwire sensor that is dropped into a factory molded tube When you replace this factory molded tube with an aftermarket piping the airflow around the sensor is often affected (and thus the check engine lights). The Mustang MAF is incorporated into it's own tube and is less likely to have this problem of unstable signal. The other reason that there are problems with aftermarket intakes is because of the metering tube size. Typical import molded tubes are obscure sizes like 63mm. a size that is optimal for the flow and design of the engine, but has no real good counterpart in "standard" sizes of alluminum tubing in which aftermarket intakes are made from. As such, you get a slightly larger or slightly smaller tube than the factory system was using... The computer doesn't know that the tube is a different size, it just sees voltage from the hot wire that's suspended in the tube.. If the tube is larger than factory, then the velocity of the air moving past the hotwire is slower, and the computer (basing airflow off of the factory size) thinks there is less air entering the engine. It's internal diagnostics arn't seeing the volume of air it expects based off of RPM and load and it tosses a check engine light. If the tube is smaller, then the velocity of air is faster, and the computer thinks there is more air than actual, and again tosses a check engine light

However you are also overlooking one of the serious problems with an AFM... Airflow isn't the only thing the AFM measures. unfortunately the AFM flapper door has mass, thus any time the meter moves, the movement will affect the flapper door position. so, if you accelerate/decelerate/corner hard, the forces applied to the flapper door will work against the spring, and cause the meter to be off. I see people who mount their AFMs on it's side (for space considerations), and know that every time they go over a bump that their fuel mixtures are off.

l0ch0w
10-30-2011, 03:19 AM
ok... now for simple sake of arguement, what if you were to put a map sensor after the AFM/MAF inline and a flow restrictor to help increase pressure???


Nope, due the the flow dynamics of a fan, RPM does not equal flow (Again, think of cupping your hand over a vacuum) with higher restriction, RPMs increase while flow decreases due to cavitation.



Already done..... see my idea of in-line MAF/AFMs





This converter has ZERO problem replicating stock voltage values... It's not a linear look-up table. It will output ANY voltage for any given voltage... The damn things was designed to do EXACTLY what we are trying to do... It's been used on BMWs for many many years now.

temperacerguy
10-30-2011, 03:23 AM
ok... now for simple sake of arguement, what if you were to put a map sensor after the AFM/MAF inline and a flow restrictor to help increase pressure???

HUH? Sorry to say, but this makes absolutely zero sense at all. Why would you want to increase pressure? Even in a flow test rig?

l0ch0w
10-30-2011, 03:34 AM
Like this...

http://img256.imageshack.us/img256/6262/testrig.png (http://imageshack.us/photo/my-images/256/testrig.png/)

Uploaded with ImageShack.us (http://imageshack.us)

temperacerguy
10-30-2011, 03:35 AM
Like this...

http://img256.imageshack.us/img256/6262/testrig.png (http://imageshack.us/photo/my-images/256/testrig.png/)

Uploaded with ImageShack.us (http://imageshack.us)

I know what you are saying... but "Why"?

l0ch0w
10-30-2011, 03:37 AM
because running them back to back will potentially cause unrepeatable turbulence issues with the second in line maf...

temperacerguy
10-30-2011, 03:40 AM
OK..... Not really sure you are seeing the real problem here. All adding a restriction to the system will do, is reduce the efficiency of the blower motor, with zero effect on the pressure differential ahead of the blower caused by the different restriction values of the AFM/MAF.

l0ch0w
10-30-2011, 03:44 AM
i was thinking ahead of myself, I know it wont increase flow at the AFM. MAF, but what it would do is increase the available pressure that the MAP sensor would be able to see right?

how about we move this to chat real quik, i think i must be missing something here...

temperacerguy
10-30-2011, 03:47 AM
we're not dealing whatsoever with a MAP pressure here.... there is no pressure sensor in this setup. We are measuring flow sensors.

l0ch0w
10-30-2011, 03:49 AM
sigh... the map pressure would only provide us with an x-value... its unrelated to an actual measurement of airflow... You still arent addressing my issue with the fact that the flapper style AFM will adversely affect the airflow across the hotwire MAF, unless you made the rig really big with lots of bends that also reduce the efficiency of the blower...

temperacerguy
10-30-2011, 04:00 AM
actually I did address it.... by adding a length of pipe between the sensors. I recomended a 1 meter pipe in my first design, however I believe that you could get away with as little as 18"s between sensors.

joe's gt
11-01-2011, 03:08 AM
I personally believe it is just more accurate to put the MAF first in the series. Mainly because we are dealing with compressible flows (air) here. With incompressible flows the following statement would be true:


Theoretically it shouldn't mater which meter goes first. the pressure (or pressure drop) of air going through one meter will in no way affect the other meter. These sensors measure FLOW and FLOW alone. If there is any density of air change by pressure drop through one meter, the other will not be affected. As long as the rig is sealed between the two sensors, X volume going through the test rig would be flow matched at any point in the test rig regardless of pressure.

The reason is in a steady state system, mass flow rate is conserved (fundamental law of fluid dynamics), in other words, mass flow rate IN equals mass flow rate OUT. Mass flow rate equals density x volume flow rate. For incompressible flows, density remains constant through restrictions, so therefore, flow rate will also remain constant. This is not a correct assumption for air (however, later on I will explain that in our case we can safely assume it as incompressible). Air is considered a compressible flow, which means that when it flows through the restriction of the VAM it DOES change density and in order for mass flow rate to be conserved, volume flow rate MUST change which would mean the next meter in the series would not see the same flow rate.

However, we are lucky and we are using low Mach number flows at constant temperature here so honestly, we can actually safely consider air here incompressible and assume the density does not change considerably through the restrictions. I am just saying, its just as easy to put the MAF in front of the VAM, so why not take out the negative compressibility effects of the air through the VAM by making it the last meter in line. Due to the air being constant density in this situation, it most likely will be negligible whether it is in front or back, so why not just put it where it will have the least negative effect from the start. The MAF will past the most air with the least compressibility and therefore least negative flow rate effects.

But I definitely agree its critical to have that minimum length of pipe before and after whatever meter is put first so the flow can develop a uniform velocity profile before and after the restriction.

temperacerguy
11-01-2011, 07:22 AM
True. However the density of air is compensated for as the higher density air has more specific heat and therefore will cool the hotwire in an appropriate manor. This is theory. In practice its just easier to put the maf first, and remove variables

MrWOT
11-01-2011, 09:10 PM
Look up the factory ecu code to find out at which voltage steps it makes changes? I remember someone on mr2oc dumped it long ago and dissassembled most of it. I think the steps are fairly broad if I remember right?

temperacerguy
11-02-2011, 04:45 AM
Look up the factory ecu code to find out at which voltage steps it makes changes? I remember someone on mr2oc dumped it long ago and dissassembled most of it. I think the steps are fairly broad if I remember right?

The question is moot as to what the computer does with the information, mimic the output of the AFM and it will work as factory.

Carolina91GT-S
11-15-2011, 04:07 AM
I don't think you can just mimic the AFM output with a MAP signal converter and have it work as factory. Not directly anyway. the MAP ECU guys state specifically that to eliminate the MAF and change to MAP based that they also take RPM into account.

~Chris

joe's gt
11-15-2011, 06:25 AM
I don't think you can just mimic the AFM output with a MAP signal converter and have it work as factory. Not directly anyway. the MAP ECU guys state specifically that to eliminate the MAF and change to MAP based that they also take RPM into account.

~Chris

True, but this thread is mainly about converting the VAM (vane airflow meter) signal to work with a MAF (Mass airflow meter).

l0ch0w
11-15-2011, 08:20 AM
Remember, were trying to convert the Flappy door to a hotwire :P not MAP sensor...

I just thought of something though... A hotwire is going to be affected by external temperature quite substantially... I would think it would be necessary to couple in an ambient temperature sensor to the converter so that if its freezing cold outside, the converter could make the adjustments to the output signal accordingly.

Carolina91GT-S
11-15-2011, 09:04 PM
Is there really any benefit in going from MAF to AFM? Considering the work involved? They are both inherrently large and restrictive when compared to the MAP based system and for the amount of of effort you guys are talking about you could switch to MAP.

Not arguing....asking

~Chris

Hookecho
11-16-2011, 12:20 AM
For 99% of us the AFM will never be a restriction.

Luni
11-16-2011, 01:01 AM
The AFM itself isnt much of a restriction as far as the housing goes, however the way it reads and resolves once its all the way open is. Its a huge problem IMO.

Youd really have to drive a Gen3 powered car that runs on MAP then go back to the AFM. Its night and day. I know I say this way too much, but it is no question, no contest. This is the reason why I just went with Gen3 EFI system. Driveability is SOOO much better.

alltracman78
11-16-2011, 02:08 AM
Too bad I never had a chance to test that with a stock turbo, I'm curious as to the difference.
It's definitely better with a turbo that flows way more than the AFM can read.

MrWOT
11-16-2011, 03:11 AM
It's very much worth the move to hotwire, and yes l0chOw, it needs a temp sensor. Most have those integrated now. If you're staying in the limits of the factory ecu, I'd say it's a must do, I doubt any of the original vane meters are still in spec...

The savings in gas alone would probably cover the cost in a year or two :p

Carolina91GT-S
11-16-2011, 03:54 AM
True AFMs are not near as restrictive. Some of the Supra guys are hitting 600HP using the stock AFM and see little to no gain from removing the AFM as a restriction up to that point. AFM is better than MAF IMO. So, I'll rephrase, Is the effort and expense worth converting to AFM when you could convert to MAP for the same effort and expense?

For me, I vote MAP all the way.

alltracman78
11-16-2011, 04:31 PM
If it's the same cost/expense to convert to either MAP or MAF, there's no question.
MAF is better.

alltracman78
11-16-2011, 04:32 PM
That is, assuming both have stock reliability/driveability.
That's important [to me at least].

Mafix
11-16-2011, 05:12 PM
i honestly saw little gain from eliminating the afm. i had the hks vpc unit.
i saw huge gains from shit canning the entire electrical system. the ONLY stock part left is the distributor and i'm using that for the cam signal and nothing more. in all reality, and i'm being dead honest, if you are keeping the stock ecu; you are wasting time. i know that's blunt and all but it's true.

and from a tuner standpoint, i prefer map. much easier to work with. and if the tune is correct, there is no issues.

Carolina91GT-S
11-16-2011, 05:33 PM
If it's the same cost/expense to convert to either MAP or MAF, there's no question.
MAF is better.

No....I'm asking if you are currently using a MAF system and are going through the effort and expense to convert it either AFM or MAP, then why would you pick AFM instead of MAP?


But to your comment... Why is MAF better than MAP?

temperacerguy
11-17-2011, 02:22 AM
switching from AFM to MAF was a HUGE, HUGE HUGE HUGE difference... I wish I still had the dyno curves, but this was back 10 + years ago.

temperacerguy
11-17-2011, 02:24 AM
No....I'm asking if you are currently using a MAF system and are going through the effort and expense to convert it either AFM or MAP, then why would you pick AFM instead of MAP?


But to your comment... Why is MAF better than MAP?

there are a number of reasons... intake manifold pressure does not equal to airflow... If you change the displacement or VE (such as changing cams, porting the intake or so on) of the engine, then the MAP based system will be fueling incorrectly. with a MAF based system, the fuel requirements are made based off of actual airflow entering the engine, and thus will change the pulsewidth accordingly

temperacerguy
11-17-2011, 02:32 AM
if the tune is correct, there is no issues.

This is exactly the point... getting the tune right. If you change ANYTHING on a MAP based system, you need to re-program the engine, while with an airflow based fuel calculation, as long as you are within the max parameters of the system, the computer will automatically calculate proper fueling.

and using a VPC is trying to fool the computer into a speed density system (MAP BASED) and guess at the airflow entering the engine... Not the best way of doing things, and huge limitations, but it was the cheapest and easiest way of getting rid of the AFM back 20 years ago...

I do agree that a fully programable (NOT factory 3rd GEN) MAP based system is best... If you have access to dyno tune the engine and dyno tune every time you change a parameter... but unfortunately most people dont have the ability, or the drive to properly dyno tune an engine..

MrWOT
11-17-2011, 02:53 AM
I'll be retrofitting a complete, software tunable oem odb2 setup. Should pass cali smog too.

joe's gt
11-17-2011, 06:46 AM
I'll be retrofitting a complete, software tunable oem odb2 setup. Should pass cali smog too.

Write up will be required with such a modification, otherwise insta-ban.

Mafix
11-17-2011, 02:10 PM
This is exactly the point... getting the tune right. If you change ANYTHING on a MAP based system, you need to re-program the engine, while with an airflow based fuel calculation, as long as you are within the max parameters of the system, the computer will automatically calculate proper fueling.

and using a VPC is trying to fool the computer into a speed density system (MAP BASED) and guess at the airflow entering the engine... Not the best way of doing things, and huge limitations, but it was the cheapest and easiest way of getting rid of the AFM back 20 years ago...

I do agree that a fully programable (NOT factory 3rd GEN) MAP based system is best... If you have access to dyno tune the engine and dyno tune every time you change a parameter... but unfortunately most people dont have the ability, or the drive to properly dyno tune an engine..

very true. but i have access and knowledge and a dyno and such. so for me that's no issue. and luckily i've never had a tune issue. though sometimes working out the littlest things are annoying as hell.
although i honestly stand by what i said earlier. the toyota ecu is the problem, not all the crap under the hood.

Hookecho
11-17-2011, 02:38 PM
Werd

Luckynumber5
11-18-2011, 04:05 PM
From a tuning standpoint converting from a volume air flow based system (ie maf/afm) to a map sensor is a tuning nightmare. If you want to go map go with an ECU designed for it, the 3rd gen efi sounds pretty solid. I can tell you from personal experience switching to speed density based metering from volume based metering is uhh challenging to say the least. Switching to a hot wire setup from a flap style setup may result in an improvement probably from the extra resolution the hot wire can produce but it will be up to you to calibrate the piggyback to "translate" it for the ecu.

Mafix
11-22-2011, 04:38 PM
From a tuning standpoint converting from a volume air flow based system (ie maf/afm) to a map sensor is a tuning nightmare. If you want to go map go with an ECU designed for it, the 3rd gen efi sounds pretty solid. I can tell you from personal experience switching to speed density based metering from volume based metering is uhh challenging to say the least. Switching to a hot wire setup from a flap style setup may result in an improvement probably from the extra resolution the hot wire can produce but it will be up to you to calibrate the piggyback to "translate" it for the ecu.

and what have you been tuning with?

Luckynumber5
11-28-2011, 06:12 PM
DSMlink, HPtuners, Megasquirt, TunerproRT. Just to name a few ;) Or 99% LoL

Luni
11-29-2011, 01:07 AM
very true. but i have access and knowledge and a dyno and such. so for me that's no issue. and luckily i've never had a tune issue. though sometimes working out the littlest things are annoying as hell.
although i honestly stand by what i said earlier. the toyota ecu is the problem, not all the crap under the hood.

I agree with that to an extent. The ECU/AFM is the problem in this scenario. So I ditched em both :)

Mafix
11-29-2011, 09:14 PM
DSMlink, HPtuners, Megasquirt, TunerproRT. Just to name a few ;) Or 99% LoL

while all of those are good, none of those are quite what i've worked with. and not to sound rude but i don't believe for one second that is 99%, lol.
at any rate, any ability to tune is great! good times.

Luckynumber5
11-30-2011, 03:20 PM
Thats just what I have worked with, like I said converting from the flapper style afm to a hot wire shouldn't be hard, but changing it over to a map sensor using a piggyback no thanks.

Mafix
12-01-2011, 06:43 PM
yeah not a fan. agreed there.