Changes in VE at Zone Boundaries (MAFless Tune)

[Mileage]

Hoping to get a little help diagnosing why the "VE Sum Uncorrected" (SCALEDUNVE_DMA) pid is changing drastically at a rpm zone boundary (specifically going from zone 17 to 23). The corrected VE sum doesn't change, but the uncorrected does.

Does anyone know the formula for the SCALEDUNVE_DMA pid? IE, what tables are used in the calculation. Using only the coefficients from the VVE table, the two don't match up. At the point in question (139 kpa, 3200rpm) the VVE table has a value of ~1600. However the VE Sum Uncorrected pid has a value of ~2150. The MAP corrected VE pid is ~1800. (see attached log file from 1280 to1290)

The change in VE going from one boundary to another causes a change in the APC calculation and makes the engine go lean. I've also attached the tune file, which shows the VVE table is smooth between the two operating zones.

Background:
2.4L LE5 w/ E67 ecm, M62 s/c and max boost of around 160kpa, tuning MAFless at the moment.

TIA!
1129111292

[gmh308]

The output of the VVE table is ultimately the SD MAF number which drives fueling. MAFSD in BBL. Is this number moving around? Or stable like SCALCORVE_M?

Never gone into it real deep, but would expect that the SCALCORVE_M PID feeds the SD MAF number after all compensation/filtering is applied like MAP/TPS/RPM (etc) current and aged values. Whereas SCALUNCVE_M PID, is the value before smoothing. i.e. SCALUNCVE_M PID >> VE Compensation & Filter factors >> SCALCORVE_M PID >> SDMAF g/s number >> fueling number. Would expect then that it is pretty normal for SCALUNCVE_M to jump around as it is not yet smoothed/corrected.

With smoothing/correction taking care of in the 512 cells in VE Compensation B8003. GM tunes these tables by platform/powertrain to plug holes in the VVE output which you have come across. You could apply % +/- variations per column between zones 17 to 23 and see what happens. Your E67 table is simpler than many. Good luck. :)

It would be interesting to know how GM arrives at the B8003 values. Maybe a data logger and a software tool. :)

This is Blacky's formula for the VE number in a non VVE ECM. FYI. FWIW. :)

Parameter Min Max Res
VEpcm 0.0 3000.0 .2 "{SAE.MAF.gps}*({SAE.IAT.C}+273.15)/((displacement()*61.024)*{SAE.RPM}*{SAE.MAP.kPa})* 6155274.24"

[Mileage]

The MAFSD is moving around (at least we think it is). See the air/cyl in the log file. It changes drastically (20%) when the zone changes and the uncorrected VE changes.

We expected to see the air/cyl follow the corrected VE, but that does not seem to be the case in the log file. The corrected VE is must smoother and doesn't change as much (or any at all) between zones. The air/cyl follows the uncorrected VE. (they change the same % when the zone changes)

We tried making the B8003 corrections the same between zones 17 and 23, but that didn't change the air/cyl and the step was still present.

We are wondering if something is preventing the corrected VE from being used (fuel trims?). It would also be helpful to know (or at least how to figure out) what the actual equations are for calculating the VE sum (both corrected and uncorrected). We are seeing a large difference between what's being calculated in the VVE table and what is being reported by the ecm (VE sums). Perhaps knowing what other tables/parameters are influencing these sums would help lead us to the reason for the step at zone boundaries.

[gmh308]

Ok got it. Once you switch to zone 23 it all comes into line - but 23 is lean, so 23 needs to be increased? Air/Cyl is calculated from SD MAF, but it would be worthwhile logging SD MAF to ensure that the the VVE is the only contributor to the SD MAF calc.

[Mileage]

The VVE table is the same in zone 23 as it is in zone 17. Increasing the VE in zone 23 would result in a large step between zones in the VVE table, which isn't right.

I found the SD MAF pid, but it does not show it as supported. Do you know what other pids must be active to log it?

Thank you for the suggestions. This is getting frustrating!

[gmh308]

The VVE table is the same in zone 23 as it is in zone 17. Increasing the VE in zone 23 would result in a large step between zones in the VVE table, which isn't right.

I found the SD MAF pid, but it does not show it as supported. Do you know what other pids must be active to log it?

Thank you for the suggestions. This is getting frustrating!

Logging SDMAF on an E38 is straightforward on BBL. Maybe on E67 it is not supported. Are you using BBL?

Wouldnt worry too much about steps on the VVE table. GM doesnt. Havent see any data anywhere that indicates that it is a must. No doubt that is why they have the compensation factors as the generated surface is never perfect due to the polynomial math (3rd order multivariate polynomial ugghh). Take a look at a stock LS7 or CTS-V VVE.

17 is lower on the rising rpm side than 23, then 17 dips as it meets 23 - there is a valley there on rising rpm. If you are getting onto a torque peak 3200-4000 and the manifold and heads are really starting to work, then the airflow will go up and you will be lean as the 23 values are the same or lower than the 17 peak. What happens if you increase 23 by 10% or 15%?

[Mileage]

Thanks for the help gmh308. We were hesitant to just juice up the VVE tables since the ECM appears to rely on APC for ignition timing too. Also the jump was so sudden I questioned weather the the VVE table could generate coefficients to follow close enough. I guess if we did not smooth between zones it might have worked.

Regardless, we needed to get this project out the door (wife's car) so we gave up on the APC in SD and reconnected the MAF and tuned it. Thanks to a little previous help from SCDyne, this went fairly quickly. The AFR's are now consistent and the torque curve smooth. A little work on the timing should do it. Thanks again. K

[gmh308]

Go the MAF! :) May the MAF be with you, always!

[redhardsupra]

MAF is boring, polynomial hyper-surfaces on the other hand, are not ;)

[gmh308]

MAF is boring, polynomial hyper-surfaces on the other hand, are not ;)

LOL...knew Marcin would be there somewhere! Go the polynomial hyper-surfaces too. :)

[billion01]

Could you please post your tune so I can check it out? I desperately need help finishing this turbo solstice. I am going to read up some more and try a few more things but I would really appreciate it. I have a bunch of other projects I am trying to finish for customers so its hard for me to devote my whole day to figuring out this cars tune.

[gmh308]

Could you please post your tune so I can check it out? I desperately need help finishing this turbo solstice. I am going to read up some more and try a few more things but I would really appreciate it. I have a bunch of other projects I am trying to finish for customers so its hard for me to devote my whole day to figuring out this cars tune.

Post #2

[billion01]

oh sorry, I just figured the first tune you posted was the one you were having problems with. I was wondering if I could get the finished one so I could see what your parameters were set at. thanks alot for your help.

[Mileage]

We're getting some MIL's from the ETC due to the MAF not agreeing with the calculated air flow. I assume it is due to our VVE table being out of whack as we are getting pretty well demanded fueling up in the MAF range. We had to re-tune the MAF after we relocated it into a larger CIA. I'll try to post if we get all sorted. K