E40 VE Tuning using MAF Numbers

[redhardsupra]

currently we dont have a reliably coming up with the BIAS numbers, thus we we doomed to rely on the stock numbers. so you can simplify that formula to:
GMVE=IFR*IPW*AFRwb*TEMP/MAP

however, if what you're really after is MAF<->VE conversion, there's a formula for that in the aforementioned writeup.

[GNXClone]

I'm trying this with an E40 using Tiffer's CFG from post #38. I've added a few more PIDs to the table to try and debug why the calculated GMVE is very far off from where it should be.

When running SD only, the car would idle nearest the command AFR with the GMVE values 698 to 700.

I performed a test where I forced the E40 into MAF only mode and completely zeroed out the SD tables, which were themselves almost spot on. Before I zeroed out the SD tables, the Mass Air Flow (SAE), Dynamic Airflow and VE Airflow all agreed. (can be seen in the from_121611 log file attached)

When I zeroed out the SD tables, VE Airflow went to zero, as expected. Also, the GM Volumetric Efficiency PID went to zero, meaning this is derived completely from the SD table.

I started the car and let it idle for a minute. Eventually it calmed down to where the Dynamic Airflow was in the .70 lb/min range (normal idle). And the RPM and MAP placed it in the area where it would have interpolated a GMVE near 700.

My issue is, the custom PID BlueCat defined and Tiffer put into his config, comes up with a very different GMVE result. It should be near 700 when idling, but I see values around 2000 for GMVE. The PIDs are all set to use metric and the formula is untouched from how I D/L'd it.

The numbers are off the same amounts even if the SD tables are not zeroed out, so zeroing them out isn't the cause.

I also added the individual PIDs used by the custom GMVE PID. These appear to all have legitimate values.

I'm not sure why this isn't working. My tune is scaled 50%, but that's already been factored into the MAF table, and the GMVE value is off by much more than 50%.

maf_to_ve_test,hpl shows the results of MAF only, zeroed out SD tables, and the calculated GMVE which is far too high.

from_121611.hpl shows the VE numbers from when the car was in SD only mode for tuning purposes. The idle areas are the ones of interest for comparing with the maf_to_ve_test log.

[redhardsupra]

what formulas did you use? I cant see them from the logs (is there a way and i'm just missing it?)

[GNXClone]

what formulas did you use? I cant see them from the logs (is there a way and i'm just missing it?)

This:
[PID.2126.MET] / [SENS.30.MET] * [PID.2321.MET] * 1000

Which I believe is GMVE = MAT / MAP * CYLAIR * 1000

I've attached the CFG. Taken from Post#38, then added some pids.

[DSteck]

Use this:

(([PID.2321]*[PID.2126])/([PID.11]))*1000

Log Dynamic Cylinder Air, Manifold Air Temp, and Manifold Absolute Pressure (SAE).

[GNXClone]

Use this:

(([PID.2321]*[PID.2126])/([PID.11]))*1000

Log Dynamic Cylinder Air, Manifold Air Temp, and Manifold Absolute Pressure (SAE).

This is MUCH closer, thank you DSteck! I was able to apply this to my existing test log and get plausible values. I can't wait to try it live.

By the way, I added .MET to the PIDs for:

(([PID.2321.MET]*[PID.2126.MET])/([PID.11.MET]))*1000

Prior to adding the .MET, I was getting values in the 5000 range rather than 700 range.

[DSteck]

Strange. I didn't have to do that even though I run my scanner in imperial mode.

[GNXClone]

Strange. I didn't have to do that even though I run my scanner in imperial mode.

I'm using Version 2.22. Perhaps that's the difference.

[DSteck]

I don't care enough to revert back and see. Haha. In my table, those items are set to metric units anyway. Maybe that's why.

[Pulse_GTO]

I apologize for bringing this back from the abyss, but I have searched and have not been able to find anything like this for a Gen III controller. Am I just not searching the correct terms?

[granatl]

Mind you, I only ever tune my own Gen IV computer, so forgive me if I'm way off here.
When I'm in VCM Scanner, there's already a "GM Volumetric Efficiency" PID under Engine->Airflow. I don't know if that comes from HP Tuners, or I have someone's custom config file. The numbers seem accurate for my car.

[Road]

Tech 2 shows ve efficiency too

[LSxpwrdZ]

This back calculation cannot currently be done on a Gen3 due to not having the manifold air temp PID. IAT will work if you have the bias table disabled and temp sensor in the manifold but most people do not have it setup this way.

[redhardsupra]

Sure you can, you solve for both GMVE and the Bias curve shape at the same time. It's not easy, it's not fast, hell, it's not even deterministic, but it does work.

[Pekka_Perkeles]

Matlab required?

[redhardsupra]

Oh yes, with multiple toolboxes too, probably about 10k$ worth

[DSteck]

MAT could be calculated in an Excel dump. If you can curve fit the bias function, you could calculate MAT within HPT using a custom PID. It wouldn't taken into account the filter, but it's all an approximation anyway.

[redhardsupra]

I'm not talking about merely calculating MAT, that's easy. I'm talking about optimizing it, simultaneously with GMVE in parametric form, for any number of dimensions too

[Pekka_Perkeles]

Oh yes, with multiple toolboxes too, probably about 10k$ worth

Expected that.

I've always enjoyed your posts and comments, although I am in the the different league. I'm playing in the The Readers league, as they say.

[LSxpwrdZ]

Be easier for HP to just add the MAT PID to the scanner. I know it exists because I can log it with other software...