Gen 3 Coyote Turbo MAF Tuning and FMEM Pedal and Torque Error

[61unicoyote]

Just got a Gen 3 Coyote turbo running and trying to log to get the MAF dialed in. Having trouble with torque error and FMEM pedal. Main thing I notice in the logs is the difference between Air Load and Desired Load and also the difference in Schedule Torque and Brake Torque.

The torque tables and driver demand tables have been scaled for boost. Do I need to add more to the driver demand tables or what is causing the torque error???

log2snip.hpl
log2snip.hpt

[Pistol_91]

Youre seeing 1.0 air load at basically 3k rpm at 70 % throttle. Your scaling may be ok but you may need to push out the load even further. But looks like a pretty normal tune file other than some minor changes with scaling and SD stuff. Desired and actual air load not matching is because its not matching in the inverse tables. You need to match air load that you see with rpm, mapped point, and torque. If the inverse tables arent right itll never run right. It looks like everything from 74-443 needs to be bumped up considering youre seeing high loads at such low rpm. Id def bump DD up a bit too. Just the parts where you start coming into boost. I never had a problem with a little too much DD. Always had problems when inverse tables arent right.

[engineermike]

If the torque and inverse models don't match, then I'd start there as Pistol said.

Beyond that, the Coyote logic doesn't allow for non-barometric TIP (Throttle Inlet Pressure) like the ecoboost does. This causes a problem because the PCM needs to establish an airflow vs throttle angle relationship, which it can not do if there is anywhere from 0 to 15 psi pushing air through the TB as opposed to 0, and the PCM is blind to it. I've contemplated ways to resolve this, such as somehow connecting intercooler pressure to the baro sensor or gross manipulation of the torque demand table. Ultimately, increasing max wheel torque error is the easiest fix. However, I can't condone this because the IPC is a safety system.

[61unicoyote]

Thanks for the thought-provoking replies. I disabled the IPC for now to tune the MAF.

The TT and Inverse TT do match, used the calculator.

The thing that I think Im observing is what engineermike is describing (ECU blind to TIP). My DD asks for 320ft lbs, the inverse TT says that will take .84 load, then the throttle opens (not knowing its FI) alot and the MAF reads 20 lbs/min at 3000 RPM resulting in an Air Load of 1.1 and indicated torque of 370 ft lbs.

The question is how do I tell the system to open the throttle less for a given Driver Demand? Do I do that in the ETC or the Inverse TT??

What if we change just the inverse TT to request less airflow from the throttle for a given DD request in the areas under boost but do not change the companion Torque Table? This would request less throttle but when the MAF reports back and the Air Load is calculated it will refer to the Torque Table and find the correct indicated torque. The indicated torque would come close to matching the Scheduled Torque.

Is there any other way to tell the throttle to open less for a given DD Torque Requests?

TT.jpg

[engineermike]

There’s not a simple answer for that but I’d start by logging desired air flow and maf.

[Pistol_91]

The inverse calculator is not a good tool to use. As far as I know, it's broken still. I could be wrong. If that is your before and after picture of the inverse tables then the before picture looks more accurate than the after picture.

[61unicoyote]

The first is with good TT and Inverse TT using the calculator. Didn't know the inverse calculator was broken. I also built an inverse interpolate function in excel....got the same results within significant figures. The second is my proposed gross manipulation to try to get the throttle to not open so much. I'm curious what the ECU will do when the two table don't match, inverse wise. Is the inverse table used for more than to calculate desired load from the demanded torque?

I also just learned there is an adaptive learn on the throttle blade in relation to scheduled torque vs brake torque (or desired vs Airload, not sure). I wonder how much this will adjust my indifference.

[engineermike]

There is a feedback control loop on airflow vs desired flow but it has limitations. This is why I recommended logging those two parameters.

[JLawson240]

Any chance it could have something to do with the Aircharge (Load at WOT) table under the speed density tab?? I noticed in the log about the time it hits FMEM desired load is around .84 at 2800 rpm which is curiously close to the .84 load clip at 2500 rpm in that table. Looking through all Roush Calibrations this table is calibrated accurately so I suppose it could be clipping you once you start to build the smallest amount of boost over what it used to see NA.

[veeefour]

Inverse calc is not broken(SD calc somehow is)....there are other issues in your tune that I see repeated over and over.

[Pistol_91]

Wouldn't it be considered broken when it populates 0.0 in a ton of cells? I've had it spit out good numbers and I've had it give me 0.0 alot. The SD is strange definitely. I found if you smooth and re calculate a ton of times it'll give you decent results.

[veeefour]

Well if you request it calc a 0 by 0 its still going to be 0.
Not claiming its perfect but it works guite good enough for me.

[Pistol_91]

Right. Which makes sense. I just don't ever find that on the higher end of the load and rpm it matches what the actual air load would be when doing a pull. I'll try it out again and test it. Thanks!

[61unicoyote]

Any chance it could have something to do with the Aircharge (Load at WOT) table under the speed density tab?? I noticed in the log about the time it hits FMEM desired load is around .84 at 2800 rpm which is curiously close to the .84 load clip at 2500 rpm in that table. Looking through all Roush Calibrations this table is calibrated accurately so I suppose it could be clipping you once you start to build the smallest amount of boost over what it used to see NA.

Thanks, I'll take a look.

[61unicoyote]

Inverse calc is not broken(SD calc somehow is)....there are other issues in your tune that I see repeated over and over.

any hints...don't like the way things are scaled?

[Pistol_91]

any hints...don't like the way things are scaled?

I don't think Load at WOT is your problem, but it could be changed... I think veefour is referring to stuff like SD. Looks like it's never been touched. SD had fixed ALOT of my drivability issues I ever had. Nobody talks about it much, but it seemed to fix literally all the crazy random jolt of throttle I would get mid rpm. Alot of stuff hasn't been messed with but is ok, like OP cam angles. I think this preignition table will limit you down the road once you get into high loads at WOT. I took your same exact torque values you had inputted and put it through the inverse calculator, and it gave me different values. Even though the values everywhere are very very close it's still not matching on the high end. The torque values seem awfully low to me considering you're a turbo car. I think your scaling isn't enough either. You only changed the bottom 2 rows it looks like. Also, do you still have stock IMRCs?
SD.png
preIGN.png
inverse.png
TQtables.png

[Pistol_91]

Something in this direction may help you a little bit. But maybe veefour or mike can chime in on it.
TQ TQinv.png

[veeefour]

Correct, some here would create "books" about how to "tweak" your torque model and IPC while it is not even necessary...

SD is more important, even with Anticipation turned OFF - you can't really turn off MAP compensation in Ford MAF operated PCMs.

[61unicoyote]

Something in this direction may help you a little bit. But maybe veefour or mike can chime in on it.
TQ TQinv.png

Thanks Pistol_91! I had capped to load at 1.6 similar to the stock tune capping at 1.1 even though the inverse calculator would calculate higher loads for the stock tune. I'm just trying to tune the MAF with a light wastegate spring. I'll take a look at this and thanks again.

The preing table is my catch all ign limiter, just in case adders get out of control. Plan to bump that up once MAF is tuned and I get an idea how it response to the turbo installed (size of turbo, spool speed, etc).

I'm making changes and plan to log some more.

Thanks!

[61unicoyote]

Correct, some here would create "books" about how to "tweak" your torque model and IPC while it is not even necessary...

SD is more important, even with Anticipation turned OFF - you can't really turn off MAP compensation in Ford MAF operated PCMs.

Thanks Veeefour.....I didn't realize I still had Anticipation turned OFF. I actually meant to have it on and the Filter OFF. I'll make some changes and re-Log.

Thanks again!