TBSS MAF Over Reporting Airflow

[LastCall]

I did a MAF only tune on an E67-based LS2 a few weeks back, and all was good. Used a wideband and got the AFRs within a few percent of commanded. It ran pretty good. The ECM was recently removed to install in another TBSS, and then back into the original. That triggered a crank sensor relearn DTC. Shortly after, two codes came up. P0172 and P0175, basically both banks rich. I logged it and noticed the MAF was reporting 23-34 grams/sec at idle (850 rpm). I looked at the tune and it was at the 4400hz mark, which did indeed reflect 24 grams of air. Going through the logs a few weeks back, it was idling around 3300hz and 11 grams/sec. So I assume the long term negative fuel trims (pegged at -25%) are from the excessive and incorrect airflow being reported to the ECM, and the O2's trimming it to the max. I reset the fuel trims in the scanner, and they were still -25%. I'll post a before and after log as well as the tune.

So the question is: What would cause this? I would think bad MAF, dirty MAF, or MAF wiring? I don't see how the ECM removal and reinstall could do it.

Thoughts anyone?

[joecar]

Some $0.02 thoughts:

When it was previously tuned, were the LTFT's cleared...?

When it was being borrowed, was it edited/flashed...?

Try cleaning MAF with CRC MAF cleaner... MAF could be bad and still appear to be working. MAF wiring problems will cause DTC's.

Please do post logs.

[LastCall]

Good news, there was a restriction (electrical tape-lol) on the maf screen. Yes I cleared LTFT when I logged it, but ran it in open loop with wideband when I tuned the maf curve. Swapping the ecm's required a crank relearn, so that. Will report back when I hook back up and log.

Thanks for quick response JoeCar!!

[joecar]

good job :cheers: lol, electrical tape, where did it originate from...?

( this shows how important visual eyeball inspection can be )

[LastCall]

The maf was to small for the intake plumbing, so there was electrical tape on the mounting surface to make it fit snug.

Yes it does highlight the importance of a visual inspection.

[LastCall]

So raising this one back up - same truck.

So I updated the injector data from stock to Fast 36#. Since the original MAF curve was calibrated with 36% injectors, but stock data, it had to be reworked. I did that and got some very interesting results. To get LTFT part throttle fuel trims within a few percent, I had to add about 40% to the MAF curve. These injectors were 30% larger, but not sure where the other 10% comes from. Difference characteristics at idle? Weird.

One issue I see now is that idle ariflow as measured by the MAF is 15-16 g/sec at 850 rpm. The LS2 is heavily cammed, but that seems like way too much idle air.

What should I check?

[joecar]

Are the FAST injectors 36 lb/hr at 3-bar...?

Are the other injectors 36 lb/hr at 4-bar...?

[LastCall]

Yes, 36 at 3 bar. I converted the values with a spreadsheet based on 4 bar pressure. I'll post the tune later.

One note, I could not find injector voltage offsets, only in gen 3 format. And I read you couldn't copy, paste, and interpolate between 3 and 4 voltages because there like values don't correspond to one another. Redhardsupra posted about it. That still valid?

[joecar]

...

One note, I could not find injector voltage offsets, only in gen 3 format. And I read you couldn't copy, paste, and interpolate between 3 and 4 voltages because there like values don't correspond to one another. Redhardsupra posted about it. That still valid?I don't remember... do you have a link to RHS's post...?

[LastCall]

Yes, see post 34.

Correct IFR tables for LS3/LS7 injectors in Gen III
https://forum.efilive.com/showthread.php?t=11498

[joecar]

Yes, see post 34.

Correct IFR tables for LS3/LS7 injectors in Gen III
https://forum.efilive.com/showthread.php?t=11498


Yes, 36 at 3 bar. I converted the values with a spreadsheet based on 4 bar pressure. I'll post the tune later.

One note, I could not find injector voltage offsets, only in gen 3 format. And I read you couldn't copy, paste, and interpolate between 3 and 4 voltages because there like values don't correspond to one another. Redhardsupra posted about it. That still valid?Yes, correct, the interpolation is not linear, but is as square root...

linear interpolation between cells is an approximation, I haven't checked how close it is, but it's not far off as long as the interpolation is only between adjacent cells.

[LastCall]

Thanks for the input Joecar.

I was referring to the voltage table and interpolating there. So it sounds like I should I take known values and interpolate between them to fill in the table?

[joecar]

Thanks for the input Joecar.

I was referring to the voltage table and interpolating there. So it sounds like I should I take known values and interpolate between them to fill in the table?Yes, same applies to voltage table.

[LastCall]

Yes, same applies to voltage table.

Going to try that, thank you sir.

Do you think its possible the engine could want that much airflow at idle though?

Hard to get things dialed in when there is not good injector data out there!!

[joecar]

...

Do you think its possible the engine could want that much airflow at idle though?

...Not physically, but mathematically maybe... this is what happens when the modelling is off.

[LastCall]

Here is a current copy of the tune as well as a link to the injectors. I did not adjust the injector voltage offsets yet, they are stock still.

http://www.fuelairspark.com/fas/8-36-lb-hr-378cc-min-injectorshtml/

Also included is a short log at idle. I did not adjust for the -12% error with the tune I attached. So the numbers at idle htz would be about 12% lower than reported.

Could the voltage offsets be causing the over-reporting of the airflow?

[joecar]

Here is a current copy of the tune as well as a link to the injectors. I did not adjust the injector voltage offsets yet, they are stock still.

http://www.fuelairspark.com/fas/8-36-lb-hr-378cc-min-injectorshtml/

Also included is a short log at idle. I did not adjust for the -12% error with the tune I attached. So the numbers at idle htz would be about 12% lower than reported.

Could the voltage offsets be causing the over-reporting of the airflow?Only if they are sufficiently far out (I don't know how far they have to be), and if they are baked into the MAF and VE tables.

[joecar]

What is the IFR of those injectors at 58 psi (4-bar)...?

I calculated 5.24 g/s (41.6 lb/hr).... this would be about 5% different than your B4001 400 kPa cell.
I calculated that the 480 kPa cell should be 5.24 * sqrt(480/400) = 5.74 g/s.

Your B4001 400 kPa cell says 5.0547 g/s...
based on this I calculated that your 480 kPa cell should say 5.0547 * sqrt(480/400) = 5.537 g/s
your 480 kPa cell says 5.3391 g/s which is about 3% off.

So overall your 480 kPa cell is about 12% off.

So the in-between cells are off by some amount between 5% to 12%.

[joecar]

Since your IFR is lower than what the injectors flow, then if the MAF were corrected like this it would be under-reporting...

so, from looking at Idle after LTFT MAF adjustment.efi (https://forum.efilive.com/attachment.php?attachmentid=19963&d=1467233482) I think try this: bring your MAF table down by 10% and see what the LTFT's say.