15 Supercharged Chevy SS - MAF is fine, VVE is fine, together they are not fine

[rao]

15 Chevy SS, manual transmission,, LSA supercharger with 2.55" pulley, BTR PD supercharger cam, 60lb injectors, stock MAF. I have dual LC2 widebands and created a calculated PID that compares the average of both banks to the commanded AFR.

I have a fair amount of tuning experience, but none with the e38 controller and none with efiLive.

Installed the mechanicals and everything is working well. I know I have the correct fuel injector data.

The tune is far from complete, but I have set about dialing in the VVE and MAF and both are to the point where the error is within a fairly small range.

With the car in VVE only mode it runs very well and goes to wot nicely.

With the car in MAF only mode (B8024 and B8025 set to 0) it runs very well (although it tends to run very rich at idle on initial startup) and goes to wot nicely.

With the car in "normal" mode - VVE and MAF working together factory style it runs very well until you go WOT and then the AFRs spike very lean - very lean.

Attached are the 3 latest versions of the tune (I disabled LTFT on all of them):

Here is a scan showing the AFRs going sky high:

19139


I know I am missing something, but I am not sure what it is.

I appreciate any advice. Thanks.

[joecar]

When you tuned VE/VVE, did you make sure a MAF DTC showed up (immediately)...?

Is FPR un-referenced or is it manifold-referenced...?

[rao]

When you tuned VE/VVE, did you make sure a MAF DTC showed up (immediately)...?

Is FPR un-referenced or is it manifold-referenced...?


I did make sure that the MAF fails when tuning VVE

The FPR is set up as it was stock, un-referenced (the IFR changes with fuel pressure).

I have no fueling issues when running MAF open or closed loop or SD open or closed loop.

It seems to be happening when going into boost from relatively low RPM.

[joecar]

Do you have any log files of VE+MAF showing this...

[rao]

Attached is the log that I took the above snapshot from, along with 2 others that I did immediately afterwords , one MAF only and the other mafless.

[rao]

I now understand a little better what is happening, although I still don't understand why it is happening.

What seems to be going on is that whenever it is in blended mode, going to WOT causes fueling to be way off in the lean direction.

When the PCM is in MAF only or VVE only this does not happen.

I set up several tunes which set the transition to maf only B8024 and B8025 at various points (I put B8025 a few hundred RPM under B8024), from the stock setting of 4000 and then 3500, 3000, 2500, 2000 and 1500) The 1500 rpm tune is attached.

I then ran a quick log with each tune and any time the PCM goes into WOT or PE BEFORE the transition point it goes very lean - the commanded AFR is fine, but the actual fuel delivered obviously isn't, any time that the PCM goes into WOT or PE AFTER the transition point it is fine.

I would prefer to use the stock-like blended system, but I can't figure out why it doesn't like WOT.

[joecar]

I'm looking at your 01-17 files...

[rao]

I'm looking at your 01-17 files...

Thank you; I appreciate any ideas you might have.

[joecar]

Hmmm, your wideband waveforms are hard to read, even in the logs from post #5...

I thought I could read them, but when I saw your 01-17 logs I'm having trouble...

and both widebands agree with each other.

Try this: use LM Programmer to set both widebands to 1/6th averaging, let's see if this improves their output.

[rao]

Interesting, the widebands are crystal clear to me and they are independent of one another. The AFR Error is a custom PID.

Post a screenshot of what you are seeing.

When the problem arises you can feel it quite distinctively.

[statesman]

I don't do much with boosted engines, but I thought I'd take a look at your logs anyways.

Your cylinder air seems to be reporting very low values when you go into boost in MAF/SD hybrid mode. The values being reported make it look like your VVE is being pegged at 1 bar. I really don't know why it's doing this.

[rao]

I don't do much with boosted engines, but I thought I'd take a look at your logs anyways.

Your cylinder air seems to be reporting very low values when you go into boost in MAF/SD hybrid mode. The values being reported make it look like your VVE is being pegged at 1 bar. I really don't know why it's doing this.

I just figured it out - the "predictive coefficients" ("VE Components" in EFILive) were the culprit. The funny thing is I had tried them stock and tried the with the ZL1 values. What I hadn't tried until today was zeroing them all out - PROBLEM SOLVED :D

It took me 5 weeks of experimenting, searching and reading but apparently they are known to cause issues with boosted applications.

Thanks to everyone for their help.

[joecar]

I just figured it out - the "predictive coefficients" ("VE Components" in EFILive) were the culprit. The funny thing is I had tried them stock and tried the with the ZL1 values. What I hadn't tried until today was zeroing them all out - PROBLEM SOLVED :D

It took me 5 weeks of experimenting, searching and reading but apparently they are known to cause issues with boosted applications.

Thanks to everyone for their help.Are you talking about table B8003 VE Components under VE Compensation...?

Good deal :cheers:

Post your new tune file.

[joecar]

Interesting, the widebands are crystal clear to me and they are independent of one another. The AFR Error is a custom PID.

Post a screenshot of what you are seeing.

When the problem arises you can feel it quite distinctively.
Well, look at the waveforms for AFR_LC11 and AFR_LC12 wrt to AFRATIO_DMA in this pic from one of your logs...

the wideband AFR's sort or follow AFRATIO, they dip when it does, but they generally don't follow it (to me), I would expect them to follow much closer.

19172

[rao]

Are you talking about table B8003 VE Components under VE Compensation...?

Good deal :cheers:

Post your new tune file.

It is most of B8003 except for column "AirPerCylCrctn" which stays at 1.00

and B8053 (I zeroed them as out as well even though I don't have AFM

Tune attached



Well, look at the waveforms for AFR_LC11 and AFR_LC12 wrt to AFRATIO_DMA in this pic from one of your logs...

the wideband AFR's sort or follow AFRATIO, they dip when it does, but they generally don't follow it (to me), I would expect them to follow much closer.

19172

They will follow it more closely when I am finished, which will take a little while :) I wanted to sort this out before I went down the MAFless or MAF only path.

[joecar]

Try getting VVE and MAF working together (now that you solved the problem by zeroing B8003 (except for AirPerCylCrctn column).