93-94 truck VAF to MAF conversion project (1 Viewer)

[semlin]

VS and THA are the VAF related ECM terminals.

There are four terminals shown on the VAF

-THA is the ECM terminal for the IAT.
-VS is the ECM terminal for the VAF itself.
-E2 is an ECM terminal. Can't tell if it is a ground or power but you measure voltage on the other terminals across this one.

Finally, rhere is another terminal on the VAF labelled VC that the FSM does not explain. It connects to the VCC terminal on the ECM which is a TPS terminal.

the FSM gives the following voltage for the VAF to terminal E2 which I assume is a ground.

VS at idle = 1.2 - 2.4 v
VS at 3000 rpm = 0.8 to 1.3

Also,
VS with ignition switch on
-measuring plate closed = 3.5 to 4.5v
-measuring plate fully open = 0.2 to 0.5v

THA is measured at 20 degrees celsius at 0.5-3.4v

I just did some testing on my VAF.
Interesting results.

Mapping of the connector (not vaf side showed)
Engine off, Key turned to on.
pin 1,2,3,4,blank,5,6 (seen as you look at the connector)
Pin 1; 5V
Pin 2; floating, as shown above, varies with air flow when engine on.
Pin 3; ground
Pin 4; 5V
blank
pin 5; 12V
pin 6; ground



I measured voltage across the 2nd pin in from the right as you look at the VAF (2nd from the left on the connector).

Warm Engine, parked, no load.
Voltage with Engine off 3.93V
Idling at 700RPM; 2V
2K RPM; 1.31V
3K RPM; 1.04V
4K RPM; .84V

My initial thoughts of all this is that we may have to adjust the conversion of the MAF to between 0 an 4V.
I'll take a look at how I might do this and adjust the circuit design.

I don't think a MAF uses the full 5 volt range either so you may not need to compensate. but you might need to slide the range to have the same

hmm. comparing the manual specs to your results it looks like your pin two is the VS circuit and is more or less exactly in the middle of allowable spec.

i'm not sure which the THA/ IAT pin is as volts don't line up.

 

[landtank]

The range that is used is based on the air flow. A stock truck will flow up to basically 26 lbs of air while a Turbo truck has been recorded to flow up to 33lbs of air. So a stock truck will use less of the 5v range.

This is one of the reasons I was suggesting to get voltage readings at different rpms. Basically the same displacement motor should flow the same amount of air at any given rpm.

Now Lbs is derived from the value of the MAF calculated with the value from the IAT. But for our purposes we should be able to compare just the MAF signal and be close.

 

[EricE]

eric it looks like you have been very busy while i have been away. i wish i understood it. is that a linear voltage inverter? why does it need power?

do you think bench testing one vaf is best? judging by the fsm there is at least .5 volt of allowable fluctuation on an individual vaf at the same setting assuming it is working to spec. after 15 years i wonder if results may vary further. would it make more sense to take apart and figure out the circuit or to test multiple vafs and compare? not sure if the vaf can be taken apart and put back together.

anyway, send me a pm if you want me to repeat your tests on my vaf for verification.

The circuit needs power because that is how it works. hehe. Seriously, I can't answer it better than that.

You are over analysing it. We don't care about the internals of the VAF, only the external behavior. Meaning we really only care about the V on one pin. Pin #2.

 

[EricE]

I don't think a MAF uses the full 5 volt range either so you may not need to compensate. but you might need to slide the range to have the same

Yeah, I had already decided to try it as planned before trying to alter it for 4-0V instead of 5-0V.

 

[EricE]

One thing on the PCB design, you might want an in and an out connector, The factory harness is a 4 wire connector to the VAF but the MAF is a 5 wire connector.

So one side of the board gets the 4 wires from the factory harness and the other side of the board has the 5 wire connector already on it for the MAF sensor.

The MAF sensor uses two grounds which I believe can be a common one.

I am not sure how the connectors are going to work yet. On one side we need a connector to your MAF (you have a source for these connectors, right?).
Ideally we would also have a connector that duplicates the connector on the VAF such that our stock truck wire plug for the VAF could plug right into it. I doubt we can find these though. How I would typically handle that is to cut the stock plug off and wire a new plug in its place.

To help me get my head around how the connectors will work could you send me or post the pinout of the MAF you are using? I looked in the other thread but didn’t see it.

Also, I don’t think you want the MAF connector mounted right on the board. That would leave it (the PCB) stuck out in space and prone to breakage. Better would be to have the PCB mounted somewhere on your housing (maybe the front facing side) with the MAF plug connected to the PCB via wires. The VAF plug (or its replacement) could be right on the board or connected via wires, whichever works best.

 

[landtank]

If you look at my MAF you can mount the PCB on the bottom of it. It's a nice flat surface of enough size. If we get that far I'll incorporate the holes into the design.

As far as the plug to the MAF goes, Dan already has been able to get us the housing and wires with pig tails should be available from Toyota as well. How long these pig tails are I don't know, but we should be able to create a short harness that will run from the PCB up to the MAF and plug right in.

Getting the stock harness to interface with the PCB is going to be a little more tricky.

Here are the pinouts of both the stock MAF and the ones I'm using.

Stock pin out

P1 – 12v – Yellow/Red stripe
P2 – VG – Green
P3 – THA – Blue/Yellow stripe
P4 – E2 – Brown/Black stripe
P5 – E3 – Brown

22204-21010, 22204-07010

P1 – 12v – Yellow/Red stripe
P2 – E2G – Brown
P3 – VG – Green
P4 – THA – Blue/Yellow stripe
P5 – E2 – Brown/Black stripe

 

[semlin]

We don't care about the internals of the VAF, only the external behavior. Meaning we really only care about the V on one pin. Pin #2.

I agree, although we do need to confirm the IAT/THA works the same as on a MAF truck since that is the other input, and we also need to confirm the VCC terminal is just a ground or constant voltage.

Looking at your results, I also think we need to take more incremental measurements especially near idle to plot the curve here and see if we have any ledges.

0 to 700 rpm (700 rpm swing) = 1.93 v swing = (2.75 v swing per 1000 rpm)
700 rpm to 2000 rpm (1300 rpm swing) = .69 volt swing = (.53 v per 1000 rpm)
2000 to 3000 rpm (1000 rpm swing) = .27 volt swing
3000 to 4000 rpm (1000 rpm swing = .2 volt swing

The diminishing rpm to voltage curve this describes may be attributable to the fact that the VAF is measuring airflow and this does not translate into rpm on a linear basis. It would be great to get a range of a/f to rpm comparison for a stock truck (even with MAF since the a/f rates would be close to the same). Landtank do you have that data? If we had that we could approximately translate our rpm data into lbs of air flow and see if it is more linear than it appears right now.

I would also like to see if there is a sudden variation in voltage as it comes off idle. we know the VAF does not work well at idle and uses a vacuum in the FPR to compensate. We also know there is an anomaly on the resistance in the wiring diagrams that suggests two resistors may be used. I would not be surprised to see two voltage to airflow curves switching over at about 1000 rpm. This would make sense if the vanes are not precise enough to measure low airflow reliably. You would either compensate for this in the ECU or in the VAF signal. I hope landtank is right that the ECU is "pure". If so we will perhaps see the voltage curve switch which would be very encouraging for this project sicne it would mean the ECU is not tweaked and can accept an accurate signal from a MAF.

 

[landtank]

I agree, although we do need to confirm the IAT/THA works the same as on a MAF truck since that is the other input,

Comparing the trouble shooting guide between the two indicates that they are the same resistance at different temps.

Since this is part of the equation for calculating air flow I'm hoping it indicates use of the same calculations as used in the later ECMs.

 

[landtank]

I set up and got the rpm/VDC readings today.

Here is the raw data:

RPMs----VDC

650 - 1.35 vdc
1012 - 1.54 vdc
1533 - 1.753 vdc
2008 - 1.878 vdc
2562 - 2.057 vdc
3034 - 2.234 vdc
3516 - 2.347 vdc
3949 - 2.502 vdc
4529 - 2.655 vdc

Attached is a bar gragh of sorts.

 

[EricE]

I altered the PCB a bit to allow for a better interface to the plugs.

Having THA on the board is not nessasary but it is a little cleaner that way.
I envision a MAF plug hanging off the to-Maf side and a MAF plug receptical hanging off the to-ECU side. That requires one to cut the VAF plug off and wire in a MAF type plug. Thats pretty easy to do.

The MAF takes 5 wires but two of them are ground. I think we have all the wires accounted for here.

 

[EricE]

I set up and got the rpm/VDC readings today.

Here is the raw data:

RPMs----VDC

650 - 1.35 vdc
1012 - 1.54 vdc
1533 - 1.753 vdc
2008 - 1.878 vdc
2562 - 2.057 vdc
3034 - 2.234 vdc
3516 - 2.347 vdc
3949 - 2.502 vdc
4529 - 2.655 vdc

Attached is a bar gragh of sorts.

I keep wondering how much the computer can re-learn new values for the same RPM. I mean, it gets feedback from the O2 sensors on how to adjust things. So it may just learn a new curve over time such that it can work even if values between VAF and MAF are a bit different for a given RPM.

 

[semlin]

Rick you didn't by any chance get AF numbers for that test did you?

if I take rick's numbers and calculate a voltage per 1000 rpm rate for each interval we see a fair bit of internal variance in that rate, but it seems clear that the MAF voltage signal is non linear to rpm response, with an increased voltage generated "per rpm" at lower rpms.

RPM
650 - 1.35 vdc
1012 - 1.54 vdc = .19v swing for 362 rpm = .525v /1000rpm
1533 - 1.753 vdc = .213v for 521 rpm = .409v /1000rpm
2008 - 1.878 vdc = .125v for 475 rpm = .263v /1000rpm
2562 - 2.057 vdc = .179v for 554 rpm = .323v /1000rpm
3034 - 2.234 vdc = .177v for 472 rpm = .375v / 1000 rpm
3516 - 2.347 vdc = .113v for 482 rpm = .234v / 1000 rpm
3949 - 2.502 vdc = .155v for 433 rpm = .358v / 1000 rpm
4529 - 2.655 vdc = .153v for 580 rpm = .264v / 1000 rpm

compare this to the VAF signal

RPM
0 = 3.93v
700 = 2.0v = 1.93v for 700 rpm = 2.75 v / per 1000 rpm
2000 = 1.31v = .69 volt swing for 1300 rpm = .53 v / 1000 rpm)
3000 = 1.04v = .27 v/1000 rpm
4000 = .84v = .2 v/1000 rpm

actually not too far off at all. Both show decreasing voltage change per 1000rpm as RPMs increase and both are in the general ball park as to that rate of change.

 

[semlin]

I keep wondering how much the computer can re-learn new values for the same RPM. I mean, it gets feedback from the O2 sensors on how to adjust things. So it may just learn a new curve over time such that it can work even if values between VAF and MAF are a bit different for a given RPM.

it can relearn but there is a tolerance. If the o2 sensors require an adjustment of, I believe, +/- 10% to what the ECU signal would do from the raw MAF/VAF signal you will throw a CEL code. It may be +/- .5%. Looking at Rick's numbers though, I see what looks like random 50% fluctuations in voltage increase rates as the rpms increase, and there is between .5 and 1v fluctuation allowed in the FSM test for the VAF. So the ECU sensitivity to the MAF signal can't be that high in the first place!

 

[landtank]

The ECM adjusts for the variance through Fuel trim. According to my 1996 FSM it has a range of +/- 20%. Outside that it is considered a problem and I can attest that up into the 30% range the truck will still run properly and not throw a code.

I think the reason that the high RPMs drop off is because of that big honking sensor in the way of air getting through. But it still looks very linear considering the equipment I'm using.

 

[landtank]

Rick you didn't by any chance get AF numbers for that test did you?

rpm ---- lbs of air flow

650 - .73 lbs
1012 - 1.13 lbs
1533 - 1.71 lbs
2008 - 2.11 lbs
2562 - 2.95 lbs
3034 - 3.93 lbs
3516 - 4.54 lbs
3949 - 5.67 lbs
4529 - 7.03 lbs

 

[Lockd97]

rpm ---- lbs of air flow

650 - .73 lbs
1012 - 1.13 lbs
1533 - 1.71 lbs
2008 - 2.11 lbs
2562 - 2.95 lbs
3034 - 3.93 lbs
3516 - 4.54 lbs
3949 - 5.67 lbs
4529 - 7.03 lbs

Do those #'s seem low or am I missing something?

 

[landtank]

Do those #'s seem low or am I missing something?

there low because there is no load on the engine. So basically a little throttle produces a large rpm increase.

If I where to drive down the road with the truck wide open on a hill you would see air flow around 23 lbs

 

[Lockd97]

Makes sense now

Didn't snap you were taking values stationary.

 

[semlin]

thanks Rick, comparing volts to AF looks more promising. The rate of volts to lbs constantly decreases on the MAF and your numbers resolve themselves much better. We have what looks like a neater pattern than rpm/volts

650 - .73 lbs
1012 - 1.13 lbs = .4 lbs for .19 v = .475 v/lb
1533 - 1.71 lbs = .58lbs for .213 v = .3672 v/lb
2008 - 2.11 lbs = .4 lbs for .125v = .3125 v/lb
2562 - 2.95 lbs = .84 lbs for .179v = .2130 v/lb
3034 - 3.93 lbs = .98 lbs for .177v = .1806 v/lb
3516 - 4.54 lbs = .61 lbs for .113v = 0.1852 v/lb
3949 - 5.67 lbs = 1.13 lbs for .155v = 0.1371 v/lb
4529 - 7.03 lbs = 1.36 lbs for .153 = 0.1125 v/lb

And if I apply the MAF truck AF to RPM readings prorated to the nearest VAF truck readings I get these results

Estimated VAF AF to voltage
RPM
700 = .786lbs = 1.93v
2000 = 2.10 lbs = 1.31 v = 1.314 lbs for .69 v = .5250 v/lb
3000 = 3.88 lbs = 1.04v =1.78 lbs for .27 v = .1510 v/lb
4000 = 5.74 lbs = .84v = 1.86 lbs for .20 v = .1075 v/lb

Obviously these aren't as good since we don't know if a VAF truck flows the same air for same rpm at idle as a MAF truck but it helps and at least we see a similar linear pattern but the rates of voltage change to lbs differs. This seems to me to be key because this measures how the two systems use the available voltage range.

We need more rpm to voltage data points on the VAF. Then we need to prorate airflow from that RPM and then plot the v/lb ratios of the MAF and VAF trucks on the same graph. So far I think that just reversing the voltage won't work because it looks like the VAF voltage response to airflow is on a steeper curve than the MAF.

 

[EricE]

I am still proceeding under the assumption it will work. The electrical components are on order and should be here about the time Landtank sends his stock MAF. I'll test that breifly w/o an air cleaner hooked up as the '94 air cleaner cover has a smaller opening than needed. If that works I'll see if I can get my hands on a '95-'97 air cleaner cover for some real drive around testing.