Bench Simulator - Triggering E38 ecu, or any ECU really...

[10_SS]

I got an Arduino running my 2010 Camaro E38 ECU... scanning with HPTuners... I'm using one pot from a Jimstim to create variable RPM... will use the other JimStim pot's to simulate other sensors... coolant, etc. This is WAY easier trying to figure out some things (like what some tables do) than trying to run it on the actual engine...

https://youtu.be/MZfEYJHkWoU

[10_SS]

I have a few DTC's coming up though...

[] P0030 - HO2S Heater Control Circuit (Bank 1, Sensor 1) (Pending) (Old) (Current) (Immature)
[] P0036 - HO2S Heater Control Circuit (Bank 1, Sensor 2) (Pending) (Old) (Current) (Immature)
[] P0050 - HO2S Heater Control Circuit (Bank 2, Sensor 1) (Pending) (Old) (Current) (Immature)
[] P0056 - HO2S Heater Control Circuit (Bank 2, Sensor 2) (Pending) (Old) (Current) (Immature)
[] P0102 - Mass or Volume Air Flow A Circuit Low (Pending) (Old) (Current) (Immature)
[] P0107 - Manifold Absolute Pressure/BARO Sensor Low (Pending) (Old) (Current) (Immature)
[] P0118 - Engine Coolant Temperature Sensor 1 Circuit High (Pending) (Old) (Current) (Immature)
[] P0122 - Throttle/Pedal Position Sensor A Circuit Low (SES) (Pending) (History) (Current) (Immature)
[] P0201 - Cylinder 1 Injector Circuit / Open (Pending) (Old) (Current) (Immature)
[] P0202 - Cylinder 2 Injector Circuit / Open (Pending) (Old) (Current) (Immature)
[] P0203 - Cylinder 3 Injector Circuit / Open (Pending) (Old) (Current) (Immature)
[] P0204 - Cylinder 4 Injector Circuit / Open (Pending) (Old) (Current) (Immature)
[] P0205 - Cylinder 5 Injector Circuit / Open (Pending) (Old) (Current) (Immature)
[] P0206 - Cylinder 6 Injector Circuit / Open (Pending) (Old) (Current) (Immature)
[] P0207 - Cylinder 7 Injector Circuit / Open (Pending) (Old) (Current) (Immature)
[] P0208 - Cylinder 8 Injector Circuit / Open (Pending) (Old) (Current) (Immature)
[] P0223 - Throttle/Pedal Position Sensor/Switch B Circuit High (SES) (Pending) (History) (Current) (Immature)
[] P0230 - Fuel Pump Primary Circuit (Pending) (Old) (Current) (Immature)
[] P0325 - Knock Sensor 1 Circuit (Bank 1) (Pending) (Current) (Immature)
[] P0330 - Knock Sensor 2 Circuit (Bank 2) (Pending) (Current) (Immature)
[] P0443 - Evaporative Emission System Purge Control Valve Circuit (Pending) (Old) (Current) (Immature)
[] P0449 - Evaporative Emission System Vent Control Circuit Intermittent (Pending) (Old) (Current) (Immature)
[] P0452 - Evaporative Emission System Pressure Sensor/Switch Low (Pending) (Old) (Current) (Immature)
[] P0463 - Fuel Level Sensor A Circuit High (Pending) (Old) (Current) (Immature)
[] P0480 - Fan 1 Control Circuit (Pending) (Old) (Current) (Immature)
[] P0481 - Fan 2 Control Circuit (Pending) (Old) (Current) (Immature)
[] P0522 - Engine Oil Pressure Sensor/Switch Circuit Low (Pending) (Old) (History) (Current) (Immature)
[] P0532 - A/C Refrigerant Pressure Sensor A Circuit Low (Pending) (Old) (History) (Current) (Immature)
[] P0615 - Starter Relay Circuit (Pending) (Old) (History) (Current) (Immature)
[] P0622 - Generator Field Terminal Circuit (Pending) (Old) (History) (Current) (Immature)
[] P0645 - A/C Clutch Relay Control Circuit (Pending) (Old) (History) (Current) (Immature)
[] P0807 - Clutch Position Sensor Circuit Low (SES) (Pending) (Old) (History) (Current) (Immature)
[] P080C - Upshift / Skip Shift Solenoid Control Circuit Low (Pending) (Old) (Current) (Immature)
[] P2068 - Fuel Level Sensor B Circuit High (Pending) (Old) (Current) (Immature)
[] P2122 - Throttle/Pedal Position Sensor/Switch D Circuit Low Input (SES) (Pending) (History) (Current) (Immature)
[] P2127 - Throttle/Pedal Position Sensor/Switch E Circuit Low Input (SES) (Pending) (History) (Current) (Immature)
[] P0223 - Throttle/Pedal Position Sensor/Switch B Circuit High
[] P0807 - Clutch Position Sensor Circuit Low
[] P2122 - Throttle/Pedal Position Sensor/Switch D Circuit Low Input
[] P2127 - Throttle/Pedal Position Sensor/Switch E Circuit Low Input
[] P0122 - Throttle/Pedal Position Sensor A Circuit Low

[schpenxel]

lulz

Which audrino model are you using?

[10_SS]

Uno, but most of them will work. UNO is pretty slow/basic.

For example... "Low Res Period"... increases the lower RPM is, and decreases as RPM increases... Since 8192 is the Max readable rpm... this shows the system keeps counting...

1000rpm: 14.8msec
3000rpm: 5.0msec
6000rpm: 2.5msec
8192rpm: 1.8msec
xxxxrpm: 0.9msec

And I've been in crank mode (350rpm) now for about 1 full min at -40f coolant temps and injectors are still at 90.6msec... that's a crap load of fuel... not sure I like that idea...

[tek1229]

I've been wondering about trying something like that, have a little side project going on and having a 'running' e38 ecu on the bench would be a handy thing to have, wondered about getting it out of cranking mode, if you get it to work I'd love to see the trick for that..

[10_SS]

Oh, no it works fine, fully functional. I was just staying in crank mode on purpose... no dropouts in crank/cam signal or anything.

CUrrently I have connected:
1) CRANK/CAm signals... this is all that's needed to make it "run"
2) MAP on a pot..
3) Coolant on a pot..
4) IAT on a pot

I had to disable "Oil Starve Protection Mode" to get the Injectors to continue to fire after the first couple seconds of startup as they would drop to 0.0 after startup... but yea, it works good. Just have to setup a PWM output on the Arduino to simulate the MAF... right now it's in Speed Density mode.

It's nice to finally see how fast some sensor inputs react... MAP, and the effect on timing, etc.

What's neat to watch is the "Intake Valve Temp" and it's effect on fueling... so example startup, hold 1500rpm, and hold coolant at 180F, IAT at 80F, etc, and just watch Intake Valve Temp increase slowly, and also watch the Injector pulse width increase slowly over time inline with IVT... that kinda thing is impossible to do on an engine (keeping everything else constant)... Pretty sure you could prove out the "Rich after flash" and if it is actually the IVT temps causing it... since you can easily cycle power and flash and see what resets it... even though your holding everything else constant... I'll have to play with that later... post some logs.

[schpenxel]

just watch Intake Valve Temp increase slowly, and also watch the Injector pulse width increase slowly over time inline with IVT... that kinda thing is impossible to do on an engine (keeping everything else constant)... Pretty sure you could prove out the "Rich after flash" and if it is actually the IVT temps causing it... since you can easily cycle power and flash and see what resets it... even though your holding everything else constant... I'll have to play with that later... post some logs.

Now that is interesting... I was going to ask if you could try to figure that out. Can you see if the injector PW stabilizes once the IVT temp stabilizes? Ideally hold everything else steady and just let the IVT work itself out. It should stabilize at some point I would think.. will be curious whether Inj PW does the same at the same time.... Hold commanded AFR/EQ ratio steady too

[10_SS]

Yes that would be simple, didn't have much time left last night after connecting everything... but running something like this is quick as it's not in the car... I'm using a spare E38 flashed with the same file (2bar enhanced OS) sitting on my bench, all wired up so I can still drive the car while I play with this setup.

[MikeOD]

Do you have a scope? I'd like to see a scope reading an injector driver compared to the reported pulsewidth after a flash...What I'm expecting is you should always see a difference...but the difference should be the offset and any short pulse adder. If there's more difference than that, that goes away over time after a flash, I believe you'll be witnessing the rich after flash...then maybe we could see if it correlates to IVT or something else...but as far as I can tell, the rich after flash problem is such a problem because it's something adding to the injector pulse, without being reported as doing so...the IVT stuff seems to report that correctly.

[schpenxel]

That is a good point.

I would LOVE to see some data between the "reported" injector PW and actual measured injector PW.

I built a sheet that will take the base PW and add on the offset/short pulse adders (and interpolate as needed to figure out those values since it's rarely right on one of the points in the tables), but I'm going on the assumption that the reported PW is basically the "base" pulsewidth before these two things are added. I do not know if this is truly accurate, but it would make sense if it is.

With this setup it seems we should be able to find out pretty quick whether that's actually the case or not.

I have been meaning to try to measure the actual PW with my oscilloscope so I can compare to what my calculator is spitting out, but haven't gotten around to it yet.

Is this something you could try for us if you have a 'scope? i.e. set it up so you're at relatively low injector PW's, and while keep everything constant, log MAP, RPM, Voltage and injector PW. At the same time, see what the actual delivered pulsewidth for the "injectors" is with a 'scope.

If I had the log + info on the real delivered PW I could check to make sure my math is right for the conversion.

It would also be interesting to see what happens after a "fresh" flash to see whether the actual inj PW's change in line with the reported ones or not.

[10_SS]

K... I'll see if I can get Injector PW scoped. In the meantime... here's a boring log of a fake start up warm... what I did was turn on logging without fake engine running... then varied the coolant temp all over to show the IVT following close, unless below 32F, then IVT spikes HIGH around 900+F... strange... then 32F and above it follows again. Then dialed in ~192F coolant temp, basically a MAF tune, with MAF unplugged... and started fake engine... injectors say they are pulsing, and AFR increases as IVT increases... and yes, if I disconnect power to the ECU for 2 seconds (reboot) then IVT starts from scratch which is always Coolant temp with engine not running, once running, IVT increases from coolant temp according to whatever formula it uses... if you start at -40f coolant, then IVT starts at -40c, and if you change coolant to 200F, IVT starts at 200F and increases...

This log attached...
196F fake startup, SD mode POWER reset.hpl
Dan_Camaro 4 Bench.cfg

If I let the IVT get up to ~300F idling as I did above, and instead of cycling full power (complete batt reset), but flash instead without even turning the key or power off... then the exact same thing happens... IVT gets reset to coolant temp. In the log... it takes only ~3mins for IVT to get back to ~295F...

[10_SS]

Well Im not getting much from the Injector pin outputs.. or I dont understand the signal... I can scope the Crank/cam signals no problem... are the injector pins the same type signal? I also tried a 4.6k pullup resistor assuming the inj pins ground when on... Wondering if one of the DTC codes are disabling injector output... ?

Ok I got injector pulse... disabled VATS but didnt "FULL FLASH" after. it cuts out over 1200rpm... now though...

[02statesman]

Ok I got injector pulse... disabled VATS but didnt "FULL FLASH" after. it cuts out over 1200rpm... now though...

That's probably your idle limit kicking in. You'll need to fit a fake tps and then use it to simulate a throttle movement to get you off idle fueling.

[schpenxel]

Well Im not getting much from the Injector pin outputs.. or I dont understand the signal... I can scope the Crank/cam signals no problem... are the injector pins the same type signal? I also tried a 4.6k pullup resistor assuming the inj pins ground when on... Wondering if one of the DTC codes are disabling injector output... ?

Ok I got injector pulse... disabled VATS but didnt "FULL FLASH" after. it cuts out over 1200rpm... now though...

I didn't have much luck getting a signal from mine either with a disconnected injector. I'm going to try tomorrow to see if I can get a signal with an injector that is connected and will let you know (needed to order some probing pins).

I know most things I've seen online use a 10:1 or 20:1 attenuator when reading injector signal.

[10_SS]

I got mine working fine. with nothing connected other than scope or a LED light.. I did disable the codes for injector tests though.. as 02Statesman said, now it's prob kicking out due to no TPS.

But.. moving on... at Idles, Injector PW reported in HPTuners seems to be about 0.9ms off (less) than what my Scope is seeing... not sure why... and I do see a little change in the scope Pulswidth while HPTuners is reading steady 2.21ms shortly after a "Full" Flash only, but not drastic... it went like this, all at 190F coolant, 900rpm, ~32kpa MAP... everything else constant.. (Open loop, SD mode)
1) 55 seconds at 900rpm: 14.66AFR, 290IVT, 32kpa MAP: Inj Pulse: 2.21ms HPTuners/3.37ms scope
2) 2min 50sec at 900rpm: 14.66AFR, 322IVT, 32kpa MAP: Inj Pulse: 2.21ms HPTuners/3.16ms scope
3) 11min 900rpm: 14.66AFR, 321IVT, 32kpa MAP: Inj Pulse: 2.21ms HPTuners/3.16ms scope

So it seems there is something going on with what HPTuners reports and the scope readings report at ~1min... since HP didnt change, but eh scope value did... also, theres around 0.9ms difference between the scope and HPTuners... ??

[schpenxel]

Gen 4's report "base" injector pw before short pulse adder and offset. Or, so we believe.

We need "battery" or ignition voltage and manifold vacuum to calculate offset values to confirm this theory..

Sounds like something else is going on though if only IVT is being changed.

Is commanded eq ratio or commanded afr staying constant? Open loop eq ratio changes with IVT if you haven't disabled that

[MikeOD]

Like I said...it's not reported as adding to the pulse width...but it's adding to the pulse width.

There's gotta be an additional injector offset table that we don't have access to. Offsets aren't reported as adding to the pulse, which is good, because the theoretical pulse is (with correct injector data) the direct way to calculate how much fuel it's injecting...the offset is dead time...so if you take the total pulse on your scope, remove the dead time, you should have what HPTuners is reporting...but as you just showed, at only 1 minute in, when you didn't vary anything (that I know of), you saw the pulse change from 3.37 on your scope to 3.16 on your scope.

Are you doing this with stock LS3 injector data in the test PCM, or do you have data for different injectors in there?

Can you post the file you're using in the test ECM? I'd like to back calculate the offsets given the target 2.21 ms pulse...I also need to know what voltage you're running the simulation at? 12 volts?

2.21 theoretical pulse, with stock LS3 data, the short pulse table would subtract 0.039, getting it to 2.171. Assuming it's reporting 58 PSI fuel pressure, then at 32 KPA, it's adding ~10.08 PSI to 58 to get the pressure delta, which is 68 PSI. 68 PSI and 12 volts in the offset table adds 0.9531...so 2.171+0.9531=3.1241 ms...pretty dam close to the 3.16 you're seeing on the scope...so where did the 3.37ms before that come from??? That's the rich after flash problem we're all fighting. That math all means that at 12 volts, and 32 kpa, with stock LS3 injectors and 58PSI static fuel pressure, if you need to deliver 2.21ms worth of fuel, then the PCM needs to hit it with a ~3.12ms pulse to give it time to "wake up" and move fuel.

[MikeOD]

I'm basing all that on the stock injector data from my personal 2010 Camaro LS3 file. Again though, if you have different injectors (you have a blower right?) then I'd like to see the file/data you've got in the test PCM.

Maybe there an offset multiplier that's relative to IVT or something? If there is, I'd LOVE to know what's in it so I can account for it. Drives me mental.

[MikeOD]

Gen 4's report "base" injector pw before short pulse adder and offset. Or, so we believe.

We need "battery" or ignition voltage and manifold vacuum to calculate offset values to confirm this theory..

Sounds like something else is going on though if only IVT is being changed.

Is commanded eq ratio or commanded afr staying constant? Open loop eq ratio changes with IVT if you haven't disabled that

Commanded EQ/AFR doesn't matter here...there's a commanded pulse of 2.21...and he told us the MAP pressure is set at a constant 32 KPA, so the only variable that matters for the tables we have access to, is voltage. I'm assuming he's holding it constant with the fake engine setup here...so that's why I just guessed at 12 volts.

Regardless though, commanded EQ/AFR doesn't matter, when you know target IPW, you're already past that in the equation.

[MikeOD]

Well...I just found something else that's a bit interesting. The makeup pulse minimum is 0.25ms. You said you had ~3.37ms for scoped pulse right after flash...and I calculated that it should be ~3.12ms. That's a 0.25ms difference. Now...are you positive in your 3.16ms measurement after, or is it possible that it was 3.12ms?

I might try to disable the makeup pulses just for a flash to see if I am no longer rich afterwords...maybe something is triggering the makeup pulse after a flash??? If I understand makeup pulses, it's a transient thing, when the injector pulse is already started and there's a change in airmass, the PCM is able to use the makeup pulse to try to get that fuel in??? Unfortunately I don't think it can remain disabled...I think it's necessary to run the engine properly.

Can you try disabling it on your test PCM and repeating the previous test with the scope?