CALC.PID for simultaneous MAF & VE Calibration - need sanity check

[WeathermanShawn]

Somewhere ( I can't find the post) 5.7ute (Mick) had a premise to develop a Calculated Pid that showed just the Airflow Correction that was being applied.

You have CYLAIR, DYNCYLAIR and a great pid would be 'Airflow Correction'. I remember Mick attempted to use IBPW, airmass, and other pertinent terms. Of course the math would be challenging. But ultimately we probably could compute it by looking at how the IBPW changes as DYNCYLAIR and CYLAIR change.

Now how do you that? I guess I am still skeptical until I see the process played out. You can get a valid end result, but it won't stand up in testing unless you can spell out how you got there...

[Gelf VXR]

where DAT = dynamic air temperature (blend of IAT and ECT)

.

Hi Jo, ther is no charge blend pid for the e40, not that I have come across :(


E40 was around 2005 for pontiac gto and Holden/vauxhall monaro's

[joecar]

Hi Gelf,

Post your .tun file... it may be possible to make a calc pid using the lookup() function... (this is what is being done with the LS1A 1997/1998 PCM since they don't have that pid either).

[Gelf VXR]

E40 stock tune file attached.

[joecar]

I'm going to look at the E40 file more tonight...

[Gelf VXR]

I think I'm ready to declare sucess, well better results than I've personally seen from seperate autove and MAF calibration.

Short log attached, I have a long trip to the coast planned to log this weekend for making further refinement.

To clarify, I am calibrating VE table and MAF from the single log file, I am tuning VE and MAF to the richer of either bank for a given cell (one bank is lean at lower cells while the other is lean at higher cells.

I am correcting the GM.AIRPERSEC pid using a CALC pid (attached) that uses fuel trims and BEN from wide band, if the trims are adding 4% and BEN is 0.99, then 3% is the correction factor. I then the divide the seperate values for left and right bank by the DYNAIR and MAF pids to obtain the correction factor for those pids to align with the AIRPERSEC pid, I enter these results into a spread sheet that takes the lower values from the left or right bank to create a single correction tables for the VE and MAF calibrations. I have applied these at multiply by 50% in the tune tool, which means subsequent logs and applications should close further to the desired results.

The results although not perfect, are very encouraging close to zero trims on which ever happens to be the rich bank at the time and positive on the lean bank.

If you play bank my log, you can observe that all three g/sec pids have converged.

If you check the LTFT attachments you can flick between the two and see that one or the other is close to zero, or equal distant neg or pos from zero.

This is now my preffered and simple method for tuning

I have attached my filters i used to obtain the results.

Update: Forgot to mention I have enabled dynamic fueling to the red line.

[Gelf VXR]

Condensed graph of the log attached showing the three g/sec pids and combined ST & LT fuel trims

[WeathermanShawn]

Mmm. Still not sure I follow.

If you are MAF-Enabled, how do you know that all the work you have done could simply have been done directly through the MAF. I.E., MAF controls Closed-Loop Trims.

I guess I will keep an open mind. When I tested the CALC. VE Table for LS1's, the only way I could prove it was to fail the MAF and then apply the equations/Trims.

Have you done the same, or am I just missing the premise?

In any case, good work. We need more methods for other OS's.

[Gelf VXR]

Mmm. Still not sure I follow.

If you are MAF-Enabled, how do you know that all the work you have done could simply have been done directly through the MAF. I.E., MAF controls Closed-Loop Trims.

I guess I will keep an open mind. When I tested the CALC. VE Table for LS1's, the only way I could prove it was to fail the MAF and then apply the equations/Trims.

Have you done the same, or am I just missing the premise?

In any case, good work. We need more methods for other OS's.

The main objective was to have the MAF, VE and any other g/sec agree on the value, which they now closely do, the fuel trims then apply equally to all three pids.

When I performed seperate autove and maf, they never agreed entirely which left the unanswered question, which pid do the fuel trims apply to.

[WeathermanShawn]

Gelf:

You see thats the conundrum. It looks correct, but you have to see if it is correct for all circumstances.

I.E., what if a tuner's MAF fails on the road. Will it still be valid? In my case, I did 3 tests.

1. Add 15% to all VE Table values and log Trims & Airflow.
2. Subtract 15% to all VE Table values and log Trims & Airflow.
3. +/- 15% to the MAF Table. Log Trims and Airflow values.

For the LS1, Trims followed the MAF Calibration to the greatest degree. But, you have to have it work if the MAF fails. It does happen, and only then can you really be sure.

Still, good work. I am just saying it may not be as easy as the results show. Perhaps I am wrong though. I will keep an open mind..