N13 Delco ECM Modification

[aconnoll]

Hi Guys, still are having no luck in getting the new rom mapped for a 2 bar sensor working. A while ago someone provided us with a 2 bar v8 bin and we have been extensively refering to this for some of the constant values as well as some other non VE and spark related tables. We have managed to make a rom that idles fine, but as soon as you hold the accelerator down it acts like its going to die.. and held for long enough it does. So thats our progress so far.. we would really appreciate some help if anyone has any time to check our rom attempt (compared to the stock AMXY). Kman, thanks for your help we have been learning from your pointers on where to look. Unfortunately I cant find anyone else that has attempted to make a stock rom but for a 2 bar map sensor on the internet.

[Kman]

Can you list the changes you've made so far as a matter of reference? Or re-post the current binary so I can see them?

If it can idle then some ALDL data will help diagnosis. My car is pretty well sorted at the moment, so I can make some time to have a look during this week.

Incidentally, I met a fella on the weekend who's built a Camira with a stock 1.6, a Toyota SC12 charger, no cooler, running 4psi on an untouched ECM. Apparently some simple electronics are used to fool the ECM into thinking it's running cold and the AFRs are reasonably good. Obviously a hack job, but an interesting excercise and probably quite appropriate given that it's intended purely for rally, not the road.

[aconnoll]

Ok, basically the rom called cjac1208_inj11.bin works great with the 1 bar tuning. Its pretty much an optimsed amalgamation of the main spark and VE from AKFL in the AMXY. Also included in the zip is the cjac20061028_2bar2.bin which is our latest attempt at modifying for 2 bar. We used the v8 2 bar rom we received a while ago from Elnino for reference in some areas.. Here is a list of changes we have done.. Its probably easier to look at the changes using the difference tool in tunerpro, but they are listed below anyway.

PARAMS AND TABLE CHANGED

---------------------- Constants
++ MAP PARAMS
KKACMAP

++ DIAG PARAMS
KK24MAP
KK33TPS
KK33MAP
KK33TIM
KK44TIM
KFANTCH
KFAN2TCL
KFAN2TCH
KFPTPSHI
KTCCTMPL

++ RUN A/F PARAM
KAFTCTH

++ IDLE A/F PARAM
KFILTCLT
KIDLETIM
KLEMGAIN
KLEMMAX
KLEMRPM
KFILTVE
KVEFLRPM

++ CRANK A/F PARAM
KAFCFTA

++ AE PARAMS
KAEDFTH
KAIESCN
KAETPS
KAEFUEL
KMANRPM

++ DE PARAMS
KDEDFTH
KDETPS

++ DFCO PARAMS
KDFCOSPH
KDFCOSPL
KDFCOMAP
KDFCOG
KDFCOSLL
KDFCOPW
KDFCOCLT

++ CLOSED LOOP PARAMS
KCLOXTH
KCLOXDB
KCLITMI
KCLITMX
KCLPROP
KCLTCLL

++ BLOCK LEARN PARAMS
KLCTIME1
KLCTIME
KBLMMAX
KBLMMIN
KVEMIN
KBLESRAD
KBLPMRAD

++ IAC PARAMS
KISSWNA
KACDLD
KPNDELAY
KETCDLTA
KIACCL
KIACFAN

++ WARMUP PARAMS
KF14TM1
KISALPC
KF14Tm1

++ CLOSED LOOP PARAMS
KDRDB
KKPNDB
KFILRPMH
KFILRPML

++ TRANSIENT IAC PARAMS
KTCDELTA
KMLGAIN
KMLMAX
KSSRPM
KSAGSTEP
KTCRATE
KSRRATE

++ BACKFIRE INHIBIT PARAMS
KBKFCRNK
KBKFRPM
KBFKTEMP

--------------------- Tables
F78 - TPS Default value vs RPM
F1 IDLE - Idle spark advance
F1 SAMAIN - Main spark advance
F59 - Baro adjustment offset
F24 - Max closed look integrator
F29H - Volumetric efficiency
F56 - Cold engine A/F
F60 - Open Loop A/F
F10 - Warmup IAC Motor pos
F12 - Powersteernig offset
F14 - IAC warmup decay
F16 - Throttle crack offset
F18 - Map Limiter

Let me know if you need any more info or if we can be of any assistance. Cheers, Adam.

[Kman]

Out of curiosity, give this one a go and let me know what happens.

EDIT: Sorry, wrong file, I'll update it shortly.

[aconnoll]

Sorry Kman, there is only a log file in the zip??

[Kman]

New bin below.

[aconnoll]

for kman..

[Kman]

I know this is getting extreme, but is there any chance you can capture some audio of the engine running with the 2bar sensor?

Looking at the logs I'm fairly certain the MAP reading for idle is too high. The IAC is open considerably further running the 2bar sensor which can only suggest either really bad fuelling or no enough ignition advance. If you can have a smell of the exhaust while it's idling and play with the fuel accordingly you might get a better result. The fact that you had an O2 ECM fault code at some point might be a give away (which one was it by the way?).

Also try fiddling with the static advance on the distributor to see if this changes the idle operation at all. Have a look at the advance with a timing light just to observe where it's idling and what happens when you give it some throttle. If you have the patience, try to observe the difference between the two setups with the timing light also.

Any more detailed explainations of the symptoms and observations might give us a further clue. Don't be a afraid to write an essay, the more detail the better. This would be a lot easier if I could be there in person

[cjorgens79]

hey kman, im Craig, Adams counterpart in this whole 2 bar mapping for our camira project.

I have captured a recording (engine.wav) of the engine idling and under acceleration.. sorry the quality is a bit poor, i can try do a better recording if this one is too hard to hear. At about the 6 second mark i give it a little rev (very quick throttle on/off), it pretty much cuts and then idles high for a couple of seconds after i release the throttle. At about the 11 second mark i hold it for a bit longer, you can hear it struggle, then back to idling high for a couple of seconds once i release throttle. At about the 16 second mark i rev it a few times, short thottle on, throttle off, repeated.

If there anything particular you'd like me to do during the recording, let me know.

-------------------------------------

Ive made another rom in which i pretty much started restoring constants back from the original amxy hybrid we did, to figure out what parts of the v8 rom made it idle ok. Basically, ive put everything back to the original settings except for the following

[ CONSTANTS ]
- Default map offsets (KKACMAP)
- Map related Diagnostic params
- Idle fuel constants
- DFCO param (KDFCOMAP)
- IAC Params
- IAC Adjustments
- Warmup params
- Transient IAC params

[ TABLES ]
- F78: TPS
- F1: IDLE
- F1: SAMAIN
- F59: Baro Adjustment
- F29H: VE
- F56: Cold Engine A/F
- F60: Open Loop A/F
- F10: Warmup IAC Motor Pos Offset
- F14: IAC Warmup Decay
- F16: Throttle crack offset (IAC steps)
- F18: MAP Limiter

So as you can see, pretty much the only things that are changed now are any MAP related constants, idle fuel and IAC related constants, and of course the above tables.

This rom seems to be the best of all so far. I have tried adding the Main Spark and VE tables that you sent in an earlier post, but they dont seem to work as well.

I have attached the latest bin (cjac20061113_2bar1.zip) and a winaldl log
(20061118_121443_LOG_2bar_061113_1.txt) of that rom at idle. The MAP and IAC values in this rom look a little better than the previous, but dont think there still quite right yet, although it does idle quite smoothly. There is also a section near the end of that log where i held it at about 30% throttle for as long as i could.. it basically spluttered and backfired, and pretty much wanted to die, but thought the log results might give some help.

---------------------------------------

Ok, ive also played around with the static advance of the distributor. It doesnt seem to make an aweful lot of difference to the idle no matter which way i move it. I was watching it with the timing light, and tried advancing it to the max and minimum. The main thing i noticed was that idle was always fine, but as soon as you give it any throttle %, it pretty much just dies. Even if the timing moved under throttle to a position that i could statically advance the dizzy to under idle, it still died, but was fine when statically advanced during idle. Hope that makes sense, basically as best i can figure, i dont think its an advance problem. I tend to feel that its more likely to be a fuel problem, its almost like once u give it throttle the A/F ratios just go out the window.
FYI, the timing light shows the base timing and the advance under throttle to be pretty much the same between the 1 bar and 2 bar.

----------------------------------------

Might just fire a couple of queries at ya about things i dont really understand, and im hoping you might know.

1. Any idea what the Air/Fuel options in the Flags/Switches section of tuner pro do?
2. Ive been trying to work out which A/F tables it uses when.. this is what im guessing so far.. During idle it uses the F57B:Idle A/F vs Coolant, even if the engine is cold, it uses that over the F56:Cold Engine A/F?.. im guessing it switches between F56:Cold Engine A/F and F60:Open Loop A/F when the cars not in an IDLE state depending on engine temp being less or greater than the KAFTCH: Max coolant temp for F56 A/F ratio?.
3. Is the stoich a/f ratio only used during the initial startup period of the engine, before it switches to the idle a/f table perhaps?
4. Can you explain to me how the IAC works..? and how this is different from the IFR? Im guessing IAC is somehow used to achieve expected map results?
5. What are the AE, DE and DFCO params?

Sorry about all the questions, only worry about them if u have time.. i just figure i might as well ask as much as i can to try and get a better understanding of whats going on.

Cheers

20061118_121443_LOG_2bar_061113_1.txt

cjac20061113_2bar1.zip

engine.wav

[Kman]

OK, I'll start with observations:

1) Injector Flow Rate (IFR) too low
When you give the throttle a stab the IFR should leap up, but it doesn't. Comparing it to my ALDL logs the IFR is *way* too low
2) Coolant temp too low
The ECM isn't activating Closed Loop operation to adjust fuel because the engine isn't getting hot enough. Sort out the thermostat, then you might see some fuel correction
3) O2 sensor *way....way* too lean.
You're dead right, it's fuel starved. A high O2 reading means it's not getting enough fuel **. The O2 should ready between 0.4V and 0.6V for stochiometric AFR
4) IAC jumps up when you apply throttle.
No idea what that means. The TPS is well into the run mode region so the IAC should remain static until it returns to idle. I'll think about this one.

If your observations regarding the ignition timing are correct, I think it's fair to assume the timing isn't the problem. Based on your log and the audio you've provided it sounds very much like a fuel starvation problem. It's possible the ECM does some sort of injector pulse width calculation that we're not aware of based on MAP. I know it does this for revs, but I expected that the fuel regulator would take care of the MAP/fuel balance, not the ECM. I could well be wrong, so you'll need to add a considerable amount of fuel to test this theory. Have a play with KINJFLOW and F60 just to test this theory.

Answers:

1) The flags relate to the engine configuration on which the ECM is being used. There's no need to play with these.
2) You would be correct, but the open loop table is only used when the open loop TPS threshold is breached. In normal cruise the ECM strives to reach stochiometric at all times.
3) Stochiometric is used during cruise operation, this is the basis of BLM closed loop fuel correction. During startup the F53 table is used.
4) Intake Air Correction (IAC) is a stepper motor controlled by the ECM which asjusts throttle bypass air to maintain idle and stop a stall condition. IFR stands for Injector Flow Rate, calculated by the ECM based on MAP and injector pulse width purely for diagnostic reasons. On other models it may be used for the trip computer.
5) Accelerator Enrichment (AE), Decceleration Enrichment (DE), Decceleration Fuel Cut-Off (DFCO).

AE: Typically not used on MPI engines, this is used to inject more fuel into a TBI manifold to compensate for a sudden throttle opening which will imediately lean out the AFR inside the manifold. This occurs because the injector(s) are so far away from the valve and the charge already inside the manifold can't be corrected. On MPI engines the injectors are much closer to the valves, so the ECM can correct fuel without any nasty transient AFR situations.

DE: Richens the intake charge such that the combustion rate slows and stops the exhaust valves from over heating during high-rev over run operation. Ever seen the flames from the exhaust of a V8 Supercar on over run? Same thing, just not as dramatic, often causes over run pop with a bigger exhaust.

DFCO: Completely cuts the fuel at closed throttle with very low MAP. More a fuel economy thing I think. Might rely on MAP rather than closed throttle, not sure, never bothered to look.

Anyway, that might keep you busy for a while, but feel free to ask more questions. I'll do my best to answer, or at least find the information if I can.

Good talking with you Craig!

Dave.

** Corrected in post 342

[cjorgens79]

hey dave, thanks for the info.

After reading what you said, regarding the really low IFR values, i looked at some of my original logs and noticed they appeared correct, but all the later ones are low. Turns out WinALDL was set to US instead of Metric conversion mode for the logging. I have re-recorded the 2 bar log in metric mode which makes it a bit more sensible. (20061118_181314_LOG_cold_2bar_metric.txt)

1. The IFR still looks to be a little low though when the throttle is applied. Have a look at the new log and see what you think.

2. Coolant temp is low because the engine is cold. I have been conducting these tests and logs with a cold engine. I know more stuff occurs with a hot engine in terms of fuel correction etc, but i figured it might be easier to get an idea when its cold and there isnt much variation to the settings.

3. Again, engine is cold so O2 sensor wont read anything useful until engine is warmed up.

4. I can see this now that you mentioned it. I notice this doesnt occur on the amxy hybrid rom, so i think your right in that it could be part of the problem.

Would it be beneficial to perform some more logging test with a hot engine so that the O2 sensor is functioning properly?

Thanks for the answers to my questions, thats made things a bit clearer for me. I had found some info on the web that said IFR was Idle Feed Restrictor, which was confusing me a bit. The other info you provided has also made some other settings make a bit more sense.

The latest 2 bar rom still has all the IAC settings from that 2bar v8 rom we've been using as a basis to try and get our 2 bar working. Im thinking some of those settings are perhaps causing that IAC to jump under acceleration when it shouldnt be.

I found a few places that refer to IAC as Idle Air Control, the camira manual and tuner pro and some websites. Is this a different thing to Intake Air Correction, or same thing different name?

If i lower the value of KINJFLOW, then in theory it would force more fuel in, because it would think it needs a longer pulse to get the desired amount of fuel, right? Would i be best trying just to change that? or change all the ratios on the F60 table instead?

Cheers again dave
20061118_181314_LOG_cold_2bar_metric.txt

[Kman]

IFR is definately Injector Flow Rate in this case, and that looks considerably better, but still very lean. Get some data with the engine up to temp and get at least a few minutes of logs, then we may have a better idea.

Even if the engine isn't up to temp yet, the O2 sensor doesn't take long to warm up at all, so the reading will still be meaningful. Either way some warm data will point us in the right direction.

IAC is more than just for idle in the case of the Delco, but yes, we're talking about the same thing. I'm still not sure why the IAC would be moving under throttle.

KINJFLOW should be increased to inject more fuel. It describes the relationship between engine capacity and injector flow capacity. By increasing KINJFLOW you're effectiely telling the ECM you have a bigger engine.... inject more fuel. Could do it in F60 also, but you'll soon run out of resolution

We'll get there.... more data!

Dave.

[cjorgens79]

Ive found that the O2 sensor seems to take a fair while to get up to normal operating temperature. We have an LED display in the car hooked up to the O2 sensor so we can see what its doing and when its working correctly.

Ah i was under the impression that the KINJFLOW param specified the throughput of the injectors, which is why i figured telling the ECM it had smaller injectors would cause it to hold the injectors open longer to get the same amount of fuel as it would from larger injectors, however in our case the injectors would still be the same so we would just effectively be pumping in more fuel.. but what your saying is that it specifies the required flow rate to the engine, so we need to increase it to get more fuel.

Hopefully ill get a chance to do some logging with a hot engine tommorrow. Ill post you the results as soon as i get the chance to log the data.

Cheers,
Craig

[cjorgens79]

Ok, got a chance to do some recordings. Ive done a recording of the amxy-hybrid 1 bar rom from when i started the car cold, drove it around, gave it some full throttle, bought it back in hot and let it idle.. basically covered most things it would go through. I figure it might be a useful reference to see how the car reacts in certain conditions. I then put the 2 bar rom and sensor in and did a recording with the engine nice and hot. I was able to hold the accelerator down for a bit longer before it wanted to die, if i gave it plenty of gas it splutter along for a little bit.

Ive had a quick look at the logs so far, and im still thinking the same thing, need more fuel. The IFR values still look too low under acceleration. Most other things look pretty good, the MAP seems to be responding how i would expect to throttle pos. The IAC is still jumping under acceleration.. any thoughts on that?

The O2 values from the 2 bar rom are quite interesting. When its idling it seems to go lean and stay there. Give it a bit of gas and it starts hunting around briefly then goes to a lean state and stays there. I could clearly see this occuring on the LED display. The engine was quite hot, so i would imagine the O2 would definately be hot enough to be giving accurate readings.

Have a look at the logs and see what you think. Im thinking we'll definately try changing the KINJFLOW to give it more fuel and see what happens. Any suggestions on settings? Its currently 0.11, im thinking maybe try 0.14?

Cheers
Craig

20061119_172306_LOG_cjacamxy_drive.txt

20061119_173115_LOG_2bar_hot.txt

[Kman]

Very strange Craig.... The O2 is reading lean, but the BLMs keep dropping. At 112 it's leaning it out by 12%

Also, the TPS threshold for idle is set too low, change it to 3.91. This should also improve idle quality a little.

Ok, Engineering troubleshooting 101; change one parameter at a time in order to determine which one is having an effect on the results. I think we need to start from scratch to figure out what's screwing up the results. The IAC jumping under throttle and the reverse BLM action are symptons of something being royally screwed. I can only imagine all of the parameter copying from the 2bar binary is to blame for this.

Sorry I don't have more time to look at this, but it might be wise to start from the base binary you've created for the 1bar sensor, insert the fuel and ignition maps, also adjust the F12 MAP limiter table based on the data you have then start making changes one at a time. Don't forget the TPS idle threshold either

I know this will be time consuming, but it's the only reliable way to get a good result. We need to be more methodical about this.

Catch you up later when I have more time.

Dave.

[cjorgens79]

hey dave,

Looking at the logs from the amxy-hybrid 1 bar rom, the BLM's are perfectly fine until the car is stopped and just idling, then it starts doing a similar thing to what you mentioned in the 2 bar logs. So im thinking either the BLMs dont give accurate readings while idle (probably not the case) or the 1 bar rom is screwed. The 2 bar rom we did was based of that amxy-hybrid rom, so perhaps our base rom is no good. I will have to do some logging with the original AKFL and AMXY roms and see what they show, to see if that BLM thing is being caused by a rom config problem, or perhaps something else.

I noticed there is a heap of params around the BLM stuff, concerning temperature ranges, VE ranges and RPM. I wonder if the values we see when outside those ranges are still valid?

I had also noticed that the idle TPS threshold was lower that the actual idle tps from the logs, but i forgot to ask you about it. So i will definately update that.

Ok, well i guess first step is to check out the original roms and see what they have to say.. wondering if it might also be worth trying to get another O2 sensor just incase its a bit dodgy.

Ill let ya know how we get on.

Craig

[cjorgens79]

oh, just another thought.. could the ECM be leaning out the fuel when idling as a fuel saving method? maybe thats why the O2 seems to sit on lean for a while, then a quick change to rich, then back to lean again.

Is there any evidence of this occuring during idle in your ALDL logs?

[cjorgens79]

We have finally managed to get it running well with the 2 bar. After playing with the injector rates and the AE settings we have a rom which seems to do exactly what we would expect based on the log results from the amxy. The IAC still sits up around the 30 mark, but the amxy does that anyway, so i dont think its a problem.

Log is attached if your interested.
20061210_112233_LOG_inj20 custom idling.txt

20061210_112405_LOG_inj20 custom 1st test drive.txt

[aconnoll]

Question to the moderators - Just curious but why is this thread not a sticky thread?? Its clearly an ongoing and important thread that is a must read for all ppl attempting to mod the Delco 808 themselves.

[Kman]

Sorry I haven't been more help recently fellas, I've been a little busy. I'm very interested to know what's changed, would you mind dropping me a copy of your latest binary?

Fantasitic work fellas! It seems I had the O2 voltage backwards, lower voltage is a leaner mixture as evidenced by your new logs. Now you can start getting the fuel and ignition right