Help with log file please.

[jorge]

Hi all!!
I have been looking at this log for some time know and I can't tell what is going on. Can someone look at it and help me with it.
1) I know I have knock but don't know if its false or detonation.
2) It looks like the O2 switching is not right at points throughout the log it flat lines and the wo2 goes rich (frames 3400 to 3500) and then lean (frame 5914 to 5945). Am I looking a bad O2s?
9938
Help… Thanks
Jorge

[swingtan]

I can't check the log right now, but the only way to tell if the knock is real or false is to listen to the engine with knock phones.

Simon

Sent from my HTC Desire

[WeathermanShawn]

Jorge:

I assume your are running SD Closed-Loop. I say this as I see a current DTC for a low MAF performance.

It looks like you are having a very poor control of your fueling. I am quite surprised as normally closed-loop would not have AFR's as erratic as yours. In blunt terns I have never seen AFR's going from lean to rich as wildly as yours. If it was an open-loop Tune or COS, I might think your VE Table needs some work.

The frames you mentioned indicate your O2's responding to your erratic AFR. So they are trying to Trim because your fueling is very rich in the frames you mentioned, then very lean in the second series of frames you mentioned.

The KR you mentioned may or may not be real, but it looks like it is at mid-throttle on a warm day. It's not 'that much', but quite frankly is the least of your worries.

I would suspect bad injectors, bad O2 sensor..this log looks like more like an open-loop Tune that needs some work.

Post up your Tune and Mods..

Thanks..

[joecar]

Hi Jorge,

I see:

- as Shawn said, the MAF DTC means airmass is coming from the VE table.

- the O2 sensor cycle/switch time is about 0.7s which is slow, compare this with 0.2s-0.3s for new sensors; they still allow the PCM to compute trims, but new sensors will allow the PCM to compute better trims.

- the steady state average HO2 voltage for bank2 is about 100-200mV below the bank1 voltage (e.g. 300mV vs 500mV), this might be indicating old O2 sensors.

- lean spikes on deceleration, DFCO is enabled; it looks like you have electronic throttle (it's a 1999 Vette).

- an instance of high load without enriched commanded fuel (either PE and commanded fuel tables are not sufficiently rich); this is where you see a single knock instance, this might be related to insufficient enrichment.

- looks like you might be running SOL (semi-open loop) where STFT is trimming tightly to stoich, this produces the tight ripples you see in wideband AFR... Shawn: is this what you're referring to...?

- when trimming, the average wideband BEN (during steady throttle) is about 0.98 which may indicate offset errors in the wideband voltage; or at least this is what I think on seeing this in your log; see next point.

- what AFR is your B3601 set to...? what AFR is your LC-1 programmed to...? what fuel are you running (E10)...?

The O2 "flatlining": it's not really flatlining, but it is indicating rich or lean conditions (compared to stoich);
- frame 3438: looks like its transitioning from driving to idle, it might be performing the cat efficiency test (need to log EQIVRATIO see this); the HO2 voltage swings in frames 3885-3905 support this.
- frame 5913: looks like closed throttle deceleration, wideband jumps lean shortly after this (after dynamic/wall fuel has evaporated);

If you log EQIVRATIO instead of AFR, we can change your calc_pids.txt to use this and to use Lambda from the LC-1 (you can still have an AFR "viewing" pid).

[swingtan]

Sorry, I got caught up with cooking dinner last night and forgot about this....

I'm going to go out on a limb here and say that I think the O2's are OK. While there are times that the switching slows down, these periods coincide with low engine airflow, so you would expect the switch rates to drop. Also the slight difference in overall O2 voltage may be just the difference in airflow for each bank. The specific log point referred to in post 1 (frames 3400 -3500) is on a deceleration event, so we know that in this case, dynamics will probably push the mixture a bit rich. In the whole scheme of things, it's not a huge issue. Other than that, the LTFT's indicate that fueling is generally a bit too rich in the non cruising areas, especially idle.

The knock looks real from the log ( frames 2900-3030) and is probably due to too lean a mixture. The event that started the knock was the sharp throttle opening which pushed the MAP close to atmospheric pressure. As with joecar, I'd have set the tune up to drop in to PE mode. It does also look like you are running a little too much timing in the low RPM range as well though. The steps I'd take to work on this would be....

[list=1]
Fix any DTC errors that are not expected. ( If you're running a SD tune, then you need the MAF DTC).
Drop the high spark table by 3' across the board.
Spend time on getting the fueling correct. Always start with fueling. dial in the VE table and then if you want to run a MAF, dial that in as well.
Restore the high spark table and work on timing.
Finally, work on the finer points of the tune to get it right.
[/link]

Simon

[WeathermanShawn]

Jorge:

Joecar & Simon did a better job evaluating your log than I did. The electronic throttle vs cable always throws me when evaluating DFCO..I.E., I always expect 0.0 TPS%.

Joe, good call on the cat efficiency test.

Jorge, always best to attach your Tune and mods..luckily your issues got addressed by two of the best.

Let us know if you have any further issues..

[jorge]

WOW !! Thanks guys for all your help!
I had no idea you could tell so much from this log. It just goes to show how little I know when it comes to tuning.
Yes it is a SDCL tune 1999 Ls1 vette with ETC.
Joe,
The Mods are bolt on - LT No cats, Ls6 intake, Cold Air Intake, No MAFF, 160* Therm, 1999 tune migrated to 2002.
The LC1 is setup per the tutorial here.
I checked for voltage offset and I din't have any. 2.0Vlts on LM Programmer same on the scan tool.
I am running E10. B3601 is set to 14.63.
I also thought of logging lambda and EQIVRATIO but dint know how to set up the LC1 or the scan tool, If you can point me in the right direction I would be glad to do so.
Simon,
I will drop the timing table by 3* and replace the O2s. I just don't feel good about them.
And continue to tune. I am using Mr Prick wright up on CL tuning http://forum.efilive.com/showthread.php?11423-Long-Term-fuel-Trims..&highlight=mask+and+shield
Or should I go to Calc VE Tuning Tutorial?
Here is the tune file
9941
Keep the comments coming.
Thank you,
Jorge

[jorge]

Shawn,
In regard to the injector is there a way to do a balance test with the scan tool?

[WeathermanShawn]

jorge:

Not sure on the injector test. I misinterpreted your lean frames..it was DFCO. Your injectors are probably fine.

If you have no MAF the CALC.VE won't be applicable (you need MAF g/s). SDCL is a good Tune to go with it under the circumstances.

I'll let the others comment on your wideband and Scan Tool set-up.

Looks like a nice car. Good luck.

[joecar]

Hi Jorge,

- you checked analog voltage, ok good.

- is your calc_pids.txt file edited...? If so then please post it here.

- if you're running E10, then set B3601 to 14.2...
this won't matter when trimming (each fuelmass calculation will be wrong, but trims fix it up),
but it matters when in OL (calculates the correct fuelmass).

- with E10, you can program your LC-1 (using LM Programmer software) to 14.2;
or you can leave your LC-1 at 14.7 and instead use a Lambda calc pid instead;
a trick: program the LC-1 to AFR 1.000, then the AFR_LC11 pid is Lambda;
[ the LC-1 analog voltage output is Lambda multiplied by the programmed stoich AFR (and scaled to 0-5V) ]
then the BEN correction becomes EQIVRATIO * Lambda.

- if you log both GM.EQIVRATIO and GM.AFR you will see that GM.AFR appears filtered (is missing some events/transitions/squiggles);
you may not want to waste a pid on GM.AFR, so if you really must view commanded AFR, a "viewing only" calc pid can be made for this purpose: AFR_VIEW = stoich / EQIVRATIO, where stoich would be 14.2 for E10.

[jorge]

Hi Joe,
I have changed a few things in the calc_pids.txt file
Here it is

9945

Ok Joe, I don't quite understand your terminology so please bear with me a bit.
B3601 set it to 14.20 or 14.12
And when you say "calculates the correct fuelmass" you mean the VE table? It will be wrong? I just need to retune the VE table with the new stoich value in B3601?
Also,
"program the LC-1 to AFR 1.000, then the AFR_LC11 pid is Lambda"
I don't know what these values should be. Do you have these values?

"[ the LC-1 analog voltage output is Lambda multiplied by the programmed stoich AFR (and scaled to 0-5V) ]
then the BEN correction becomes EQIVRATIO * Lambda"
I don't understand this statement.

"- if you log both GM.EQIVRATIO and GM.AFR you will see that GM.AFR appears filtered (is missing some events/transitions/squiggles);
you may not want to waste a pid on GM.AFR, so if you really must view commanded AFR, a "viewing only" calc pid can be made for this purpose: AFR_VIEW = stoich / EQIVRATIO, where stoich would be 14.2 for E10."

Ok I selected GM.EQIVRATIO. where do I go from here. I don't realy need to view GM.AFR so it's not an issue.

Thanks for sticking with me!
Jorge

[joecar]

B3601 set it to 14.20 or 14.12
And when you say "calculates the correct fuelmass" you mean the VE table? It will be wrong? I just need to retune the VE table with the new stoich value in B3601?
Yes, set the correct stoich in B3601,

you could then multiply the VE table by 14.7/14.2 = 1.035,

or you could do another AutoVE pass, this is probably best (but see below first).




"program the LC-1 to AFR 1.000, then the AFR_LC11 pid is Lambda"
I don't know what these values should be. Do you have these values?
I was thinking outside the box and I may have injured myself, please ignore that comment for now.
:doh2:

The LC-1 has a programmable stoich AFR value, using the LM Programmer software you can see that it is set to 14.7 (pre-programmed by Innovate); if it is not 14.7 you can set it to 14.7 (which is what CALC.AFR_LC11 requires).

The LC-1 internally obtains Lambda from the O2 sensor viewing the exhaust gas;
the LC-1 internally multiplies Lambda by stoich (14.7) to produce AFR (which is what CALC.AFR_LC11 reads).

So we can obtain the LC1 Lambda by doing AFR_LC11 / 14.7.




"then the BEN correction becomes EQIVRATIO * Lambda"
I don't understand this statement.

BEN is defined as {CALC.AFR_LC11} / {GM.AFR} which is wideband_AFR / commanded_AFR.

But, wideband_AFR / commanded_AFR = commanded_EQR / wideband_EQR (since AFR = stoich / EQR).

And, commanded_EQR / wideband_EQR = commanded_EQR * wideband_Lambda (since EQR = 1/Lambda).

But, wideband_Lambda is AFR_LC11 / 14.7.

So, BEN can be defined as {GM.EQIVRATIO} * {CALC.AFR_LC11} / 14.7 (since Lambda = AFR / stoich)

[ Note this underlying relationship: AFR = stoich / EQR = stoich * Lambda ]
[ This relationship applies separately to commanded AFR/EQR/Lambda and to wideband AFR/EQR/Lambda ]




Ok I selected GM.EQIVRATIO. where do I go from here. I don't realy need to view GM.AFR so it's not an issue.
See attached calc_pid.txt file, replace your file with the attached one;

- remove GM.AFR, CALC.BEN_LC11.
- select GM.EQIVRATIO, CALC.WO2BEN1, CALC.AFR_LC11.
- change the BEN_LC11 map to use CALC.WO2BEN1 instead.

The rest of the AutoVE procedure remains the same.




Thanks for sticking with me!
Jorge, no worries
:cheers:

[jorge]

Joe,
Thanks again!
I caught a cold so I am in bed with a fever. When I get back on my feet I will replace the O2s and report back.
Jorge