Transient problems with $EE 2bar SD

[Catmaigne]

Hi, I've been having some trouble with nailing down a 2bar SD tune for my Camaro. The car has a 383 LT1 with a Procharger D1 and is running 2bar SD in open loop. The biggest problem I'm trying to solve are lean transients on tip-out, although tip-in seems decent. The AFR consistently spikes to full lean on my wideband between shifts (stick car), even after adding a ton of fuel to the corresponding VE cells. Those areas of the VE table are super rich during steady state conditions despite the spikes. This seems to happen even without the car going into DFCO. The driveability isn't too bad with the current tune, but I'm afraid of lean spikes potentially causing knock after high RPM shifts.



Things I've tried playing with:

-126EF Min Transient Fuel Mass -- I had issues with this in an earlier thread where I couldn't get the logged bpw under 1.24ms with these Siemens Deka 80 injectors. I've tried several values from 2mg to 3.75mg with little change to the spikes since initially lowering it from stock. At 4mg, the spikes were certainly less pronounced and shorter in duration, although the throttle response was somewhat sluggish.

-SD airflow patch -- currently enabled
-124B4 Derivative Gain -- Have tried stock Z28, stock Y-body, and currently am using Y-body +20%
-124C3 Cyl Correction vs TPS -- Have slightly boosted all values at 12% TPS and up by +20%

-12021 Delta TPS threshold -- pushing this up higher than the delta TPS between shifts really killed the tip-in, so I set it back to 3%
-12413 Ramp out if lower than 12021 -- haven't changed this from the stock setting of 2, but should I? Figured my issue was a ramp-in related one
-12414 Ramp in if higher than 12021 -- don't remember this affecting much, I think I tried 0.40, 3.5, and currently have it set at 1.1

-126F5 Fuel boiling time -- rescaled for 2bar
-12848 Cyl wall impact factor -- rescaled for 2bar

-126AF High Baro -- Was set to all 1's in an early version of the tune in an attempt to sidestep baro update. Only goes down to 65kpa, so probably not of much use.

-Low injector pulse adders -- Was suspicious of Siemens' data so I recently tried doubling them after seeing logged bpw drop to as little as 0.82ms, but the lean spike still pegs the wideband



Any ideas on what I should try next or how to work around the spikes? I thought about pulling loads of timing in the bottom column of the spark table to create a pseudo DFCO, but the table only goes down to 25kpa and unfortunately the MAP dips on negative TPS transients are all around 20kpa.

Here are some pics of me testing throttle response at idle while changing a few parameters: https://imgur.com/a/VqYp50I

Here's a pic from the most recent tune revision where I really fattened up the low MAP VE cells. You can see it get very fat at low TPS %, but there are still lean spikes. There is a pulse width difference in the logs, usually 20% or less between the spikes and steady state which doesn't seem to account for the wideband activity.
Copy of the EEhack log can be found here: https://drive.google.com/file/d/1sUM...ew?usp=sharing

[kur4o]

Looks like the map is capped at 20kpa. And you are loosing resolution. ALso when the throttle is off at some points the dfco is engaged.
There are spikes on a tip in also.

I will look if we can rescale the map output alittle. What map are you running. We need to make sure its output is linear throughout the range.

Do you have a second map that can be hooked on the engine. I guess that readings are approximate only.

I will look at some code to see how it can be fixed.

[Catmaigne]

Looks like the map is capped at 20kpa. And you are loosing resolution. ALso when the throttle is off at some points the dfco is engaged.
There are spikes on a tip in also.

I will look if we can rescale the map output alittle. What map are you running. We need to make sure its output is linear throughout the range.

Do you have a second map that can be hooked on the engine. I guess that readings are approximate only.

I will look at some code to see how it can be fixed.

I've seen the MAP dip to 13-14kpa in some older logs with WOT pulls, but nothing below that. And DFCO would turn on occasionally because I zero'd the delay timer, but there seemed to be spikes even with it off (see the last pic in the imgur link, rev50). I did notice some tip in spikes too and assumed the derivative gain and pumpshot stuff needs a little work, but I've been afraid to dial them up but because they might exacerbate the tip out behavior.

I am running a Delphi PS10082 which I believe is for a Syclone/Typhoon and Sunbird. I have 2 MAPs sitting around, one is a stock 1bar and the other is a 3bar LS9 MAP. I could patch the tune and log one on D27.

The MAP scaling came from here to account for offset, is this correct? http://www.ltxtech.com/forums/showth...-speed-density

[steveo]

it kinda looks like you're hitting DFCO. i see a hard throttle drop into heavy vacuum and then the injectors cut, and after a momentary delay the exhaust is full of air. that's not just lean, it's simply not firing at all. i would try disabling DFCO completely to see if the problem persists.

[Catmaigne]

it kinda looks like you're hitting DFCO. i see a hard throttle drop into heavy vacuum and then the injectors cut, and after a momentary delay the exhaust is full of air. that's not just lean, it's simply not firing at all. i would try disabling DFCO completely to see if the problem persists.

I'll make the DFCO timer something absurdly long and get another log with that tomorrow. The only log I have with a delay set over 0 is this @ 0.5s. It's just revs in my driveway, car was on for a few mins before recording started.
https://drive.google.com/file/d/113a...ew?usp=sharing

I just noticed that there's consistent activity with the EGR duty cycle, maybe this is the problem?
Just checked and the settings are min MAP 103kpa, min RPM 6375, but for some reason the max RPM field is blank. I'll try setting it to 0.

Edit: Actually just checked the tune with EEX instead of the EEXtra XDF I'm using and the max RPM shows up as 0 there. I checked some other tunes with EEXtra and they generally showed EGR RPM min/max values for most tunes, but a few parameters were missing on others. Weird.

Also checked some logs and it looks like the EGR thing just showed last week when I applied some patches for EOIT. The lean tip out is present in logs before patching, so I don't think that's my problem.
Those patches were:
-Tside comm for delta ve and EndOfInj
-End of injection

[NomakeWan]

Are you sure "Tside comm for delta ve and EndOfInj" didn't hijack the EGR Duty Cycle output data to be used as one of those parameters? Basically, are you sure that you're actually seeing EGR duty cycle?

[kur4o]

Those patches were:
-Tside comm for delta ve and EndOfInj
-End of injection

I suggest you remove those patches and use the final version of modded eehack that contains all the recent patches for pcm control.
http://gearhead-efi.com/Fuel-Injecti...6&d=1598039824
From this thread
http://gearhead-efi.com/Fuel-Injecti...ll=1#post84164

Extract the exe to your eehack installation folder and run from there.

[Catmaigne]

Patches removed and I flashed with the modded eehack.

I took an earlier revision of the tune before all the excess patching and modified a few things to be more like the later versions. Unfortunately the EGR DC problem is back despite the only patch being the SD Airflow one and showing no weird activity in the logs. I'm going to flash unedited earlier version of my tune to see if it goes away.

Anyway, here's a log with DFCO delay upped to 2s. Still have lean spikes despite being really rich at low TPS steady state.
https://drive.google.com/file/d/1gq5...ew?usp=sharing

Edit: I flashed an unmodified version of the earlier tune I referenced onto the car. I have prior logs of this tune with minimal EGR DC activity, but for some reason the EGR DC is acting strange now with a constant DC of 30%+. Could this be a hardware problem with the pcm?

[Catmaigne]

Any ideas on what I should try?