E90 woes, random WOT

[edcmat-l1]

It's been a minute since I've posted here. I have a couple vehicles, couple PCMs having some issues. I consider myself a decent diagnostic guy but these are getting the best of me. Let's jump right in.

1st vehicle won't even crank on the HPtuners modified PCM. The vehicle will start on a stock unmodified PCM. I'm doing the OE programming myself, I have my own MDI2 and TDS subscription. I'm an ASE master tech for 30 years, trained GM tech, etc. Been doing this a long time.

I can go through the IMMO procedure with one PCM in the truck and it won't even click. Put the stock PCM in it, go through the same IMMO procedure and it runs. I've done it multiple times to verify.

Second vehicle is much more interesting. This one does start and run. It's a '19 1500 6.2 with a Whipple and LTs, HF cats and exhaust. Loud as hell but that's not pertinent. It runs really good 99% of the time, but it's schizophrenic (ain't gonna lie had to google it for correct spelling). It will run perfect for weeks and then all of a sudden lose its mind. The throttle will do some wonky shit, sometimes going WOT with the pedal angle less than 20%. I've attached a screen shot and data log. I'm not posting the tune file so don't ask. If some of you guys that have experience with these want to reply I'll share the file with you. It does it on my file and it does it on the Whipple file. There is no repeatability to it. You can't make it do it. The truck will run perfectly normal for weeks at a time, flip out for anywhere from 10 seconds to a few minutes and then go back to normal.

I've worked the DD tables in the area where this thing does its thing. I've made it better, but it will still act up at total random, then go back to normal without cycling the key, flashing it, whatever. It'll go back to normal on the same drive. If you go through the data log you'll see that.

I've done some NN tuning on the Dodge cars, but not these. I can't believe it's a NN tuning issue just by the randomness. I could be wrong and I'm more than willing to hear argument that I am.

Back to the PCM that won't start. We bought a brand new PCM just for a test because obviously we were stuck with a vehicle that won't start and I can't scream at HPtuners because I don't know if it's a PCM issue. Even after it starting on the new PCM I'm not sure I can blame it on the PCM. So, on to the next chapter.

I tried putting the PCM that wouldn't start one truck, into the other (Whipple) truck that does run but has throttle issues, hoping I could flash over it and start fresh. Well, guess what that's a big fat no. Install the PCM, the MDI2 recognizes the VIN of that PCM which now does not match the vehicle. Manually enter the vehicle VIN in the MDI2/SPS program and it won't flash over it. so apparently these PCMs once flashed initially, and literally locked to that vehicle. There's no more swapping in a new/used pcm and flashing over it.

Sidebar to this saga, if anyone has extensive knowledge of the auto start/stop systems on these things please reply here or PM I have some issues no one has been able to solve.

[tunerpro]

On the E90 that won't crank, it sounds like it may be damaged. Contact support about that.

On the WOT randomness, that's part NN, and its part that the ECM doesn't have an accurate post-TB / pre-SC pressure signal. Try moving the MAP to the inlet side of the supercharger, post throttle body, and see if that helps. Good luck, these new ECUs are a PIA!

[edcmat-l1]

On the E90 that won't crank, it sounds like it may be damaged. Contact support about that.

On the WOT randomness, that's part NN, and its part that the ECM doesn't have an accurate post-TB / pre-SC pressure signal. Try moving the MAP to the inlet side of the supercharger, post throttle body, and see if that helps. Good luck, these new ECUs are a PIA!

Yeah I was thinking it's possibly at least in part due to lack of TIAP. But just the shear randomness of it doesn't make sense to me.

I've done some NN training on the Dodge stuff but not these. I'm about to though.

On the other, I have been in touch with support. Been around a min know how the process goes. But I needed to do some additional diag. work, not at their request but just to verify in my mind.

[limited360]

Yeah I was thinking it's possibly at least in part due to lack of TIAP. But just the shear randomness of it doesn't make sense to me.

I've done some NN training on the Dodge stuff but not these. I'm about to though.

On the other, I have been in touch with support. Been around a min know how the process goes. But I needed to do some additional diag. work, not at their request but just to verify in my mind.

Definitely seen the random WOT spike on an stock E90 L87, with just basic tuning. Only way I reduced it from happening was working on the Driver Demand tables and a little min spark.

What I noticed was the Predictive torque would spike and at the same time the immediate torque had a sag (was inverted to predictive reaction). I was able to majorly reduce it but it seems to just randomly happen.

It seems to happen when I have DAFM disabled and I am at a 25-35% pedal region, shortly after a shift typically. To me it seems like an inaccuracy in the NN zone transition, makes a jump and thinks it needs a significant correction to the predictive torque and immediate drops due to predictive jump. Which since it happens in a region DAFM would normally be active, I just think its a poor NN model in this region. NN path seems to the correct long term solution, just need enough seat time to sort it in my case!