foxsl

iTrader: (1)

2009 e38 ecu with an Ly6, low boost 5 psi, SD 80 lb injectors. At the most dangerous time, without any warning, the cyl airmass drops from 1.07 for example to 0.60 or way lower. The whole time, the throttle position stays wide open, commanded and actual afr doesn't change. Just the nasty reality of lots of spark advance as it's somehow being fooled into thinking it's at a much lower cylinder airmass. I thought I'd play it safer so I reduced the 0.60 cyl airmass timing in case it happened again, and the damn thing went to as low as 0.00 g/cyl which was around 40* timing while in boost.

At first I thought it was reduced power mode related but I have disabled p0068 tests and set the p101, 106 and 121 codes to a temperature I won't get up to ever.

Has anyone ever seen this before?

Longblue

I don't really know if this could be it, but in Efilive a while ago they found a hidden bit of the pcm that GM coded for a "desoot" mode. Apparently can cause a lean spot all of a sudden. It seems most people chased this by just adding more fuel at that spot. Worth a shot.

Darth_V8r

iTrader: (4)

Does your map sensor go low also when it happens?

foxsl

iTrader: (1)

Nope, stays at boost pressure

foxsl

iTrader: (1)

I already have desoot disabled but good call for anyone else reading this that may be tracking down 2 seconds or so of lean-ness when entering PE.
In my case the air/fuel ratio remains where it should be

foxsl

iTrader: (1)

A far out possibility was suggested to me but it may be the cause as my HP tuners support ticket didn't come up with any explanation. One person was saying they had a similar issue and it was related to the injector wires too close to the coil and related wires.
This only ever started happening since the 80 lb injector swap and the swap required spacers which put the connector/wiring closer the the coils. The owner mentioned today that he noticed an injector wire was partially wrapped around one of the coil wires, so this may be those rare crazy scenarios.

Can't wait to test it out now. To be super safe, I'll put all of the timing in the table to 10 degrees so if it references 0.00 g/cyl again for instance, no engine killing will occur 😁

foxsl

iTrader: (1)

It was the wiring, owner relocated the coils a little bit away from the injectors so now the coil and injector wiring isn't touching, problem solved.

G Atsma

That means the insulation is compromised. Get some new harnesses. Moving the coils is a band-aid.

foxsl

iTrader: (1)

He did the injector swap quickly a little while back and one of the coil wires was wrapped around the injector wire, that may have exaggerated the issue from day 1. Zero issues now.

ddnspider

iTrader: (26)

That is insane that EMI caused that issue. I thought for sure it would have been sensor related. Still not sure how injectors near the coil pack harness would cause that because IDC is calculated from PW which is an output, not an input to the ECU.

foxsl

iTrader: (1)

It's beyond me as well to find any explanation for it, it was the only change and did the trick. Super odd no doubt and dangerous as that's an engine killer. I've since been WOT many times and it hasn't ever happened again.

I would have never thought of this, someone on an hp tuner fbook group said that happened to them before and wiring too close was the issue.

foxsl

iTrader: (1)

Unfortunately it did it once again today. I was suggested to try setting Maf fail high and low to 0 hz, dynamic disable to 8192 and re-enable to 8191.
The old values were maf fail high and low =1 hz and dynamic disable 8000, re-enable 7900.

Since making those changes, it hasn't happened again and I've gone WOT from a few different speeds without issue.

As somewhat of a safety, I lowered the timing in the low cyl airmass part of 3000 rpm + as I'll never be reving much over 2500 rpm when cruising/driving normal.
The only time 3000 + rpm is seen is in boost so at least it follows the rpm axis and if it drops to 0.18 g/cyl airmass at 3500 rpm, the spark value there is 15 degrees. I'm not sure what effect it will have on deceleration but I'll try it out. Sure beats a 10 degree to 50 degree timing jump at 10 psi. It makes sense that the injectors will also follow the low cyl airmass and make it go lean...but at least the lack of spark will be better than tons of spark and a temporary lean condition.

ddnspider

iTrader: (26)

I almost wonder if you have a bad ECU. IAT and ECT sensors are all normal during the "ever"?

foxsl

iTrader: (1)

That seems like the only option I haven't visited, the ecu itself. Yes the IAT and ECT remain unchanged when it happens.
It's hard to tell in the log if the order of events are seen as the delay in channels but here is what changes in order (while WOT in boost):

-timing changes first from the 8 degrees it was at to 22 then 35 then 45 degrees, this is based on pressing play and pause on my video of the actual log played back so it's chopped up in increments

-then VE mg*k/kpa drops next from 2737 to 0 and cylinder airmass drops from 1.35 g/cyl to 0.71 as it plummets to 0.00
-air calc mode still says "hi speed" all this time, dynamic airflow still at 410 g/s

-ve airflow then drops from 54 lb/min to 0, dynamic airflow also drops to 0, air calc mode says "unknown"
-at this time the commanded afr and actual are still in the low 11s, then while still at full throttle, injector ms drops to 0.3 from 19 ms.

It all recovers again shortly after that back to normal.


-

G Atsma

I actually wonder if the wiring issue would have caused the ECU to be compromised.

foxsl

iTrader: (1)

It's so confusing but anything is possible as HP support had nothing to say.

ddnspider

iTrader: (26)

That VE airflow drops to zero is crazy....assuming you're SD only? You already said the map stays showing boost. Custom OS with HPT? Either they got a crazy bug or you got a bad ECU.

NicD

iTrader: (7)

Post up the actual tune file and a log, without it everybody is guessing. Also clarify that you are trying to run speed density or MAF? Desoot mode doesn't change your airmass.

foxsl

iTrader: (1)

Yup Sd only, tried the e38 airflow calc patch OS with no luck. Tried an old recommendation of adding 1 to the entire VVE, no luck.

Customer is trying to source another e38 ecu.

foxsl

iTrader: (1)

I'll post it in a bit, sd only. Desoot mode was disabled for boost purposes, not sure why it's being mentioned here.