Still problems with B3647 in OS3

[ringram]

Just wondering if anyone else has come across this.
Generally when cruising at a contant speed the commanded values change and bear no resemblence to the table B3647. If I come to a stop or slow right down it locks back onto the table.
I have a number of logs showing this issue, Ross suggested using stft and commanding 14.63, this works fine, but means I cant command other values instead. Also Im not 100% confident of the NB vs WB OEM NB Sim etc they differ by 3% so Id like to fuel both banks the same.

For example commanded might be 14.7:1 but shows as 15.21:1 or even 13.43:1 none of these values are in the table anywhere, certainly not at the map points shown anyway. Let me know if anyone wants to take a look at the log or .tun file.

Ive done a full reflash with 1290003 in case it was revised, still no go. It faults almost every time. Base system is 2000 Holden R8 Clubsport M6.

Any comments or suggestions welcome.

[Tordne]

Mate,

I've had a good look through your tune and I can't for the life of me find anything that should cause this!

As you know I am running the same OS (01290003) and don't have any such issues.

Assuming that this can be reliable reproduced.. One thing I'd be interested in you trying is going to the 01290002 OS and seeing if it does the same thing. This should be quite easy (I've done it a few times) just full flash the 01290002 OS then use your 01290003 as the cal. Just need to check the custom tables and then test it out.

Other than that.. I have no clue :(

Cheers,

[dfe1]

I don't know if this is relevant to your situation, but when you run in semi-open loop, the accuracy of the commanded air fuel ratio is dependent on the accuracy of the injector flow rate table. As an example, if actual injector flow rate is 32 pounds per hour and the IFR table is set for 30 pound injectors, the PCM will compute pulse width based on 30 pounds per hour. That will make the mixture richer than commanded (to achieve a given AFR, pulse width will obviously be shorter as injector flow rate increases). The accuracy of the VE table is also a concern because the PCM has to know how much air the engine is processing in order to detemine the amount of fuel required. If you're commanding one AFR and the wide band is telling you that it's seeing something different, programmed values aren't the same as actual values, or you have a sensor issue.

[Tordne]

What 'ringram' is seeing is the logged Commanded AFR GM.AFR PID actually reporting different AFR's to those defined in the fuelling, PE and/or associated multiplier tables. This is not the normal difference between commanded and actual. This is really weird.

[dfe1]

Sorry, I misunderstood, but consider--"sunny London, UK". You know there are going to be problems.

[Tordne]

Sorry, I misunderstood, but consider--"sunny London, UK". You know there are going to be problems.

Totally agree :exactly:

Yeah, it is really weird. He was telling me about it and I was like "send me the tune and the log". I had a look at it and was like WTF, you need Ross for this one!

[dfe1]

I have the solution! It's called a carburetor. It's made of metal and it leaks. Perfect for the UK.

But I do have to agree with you. It sounds like something for Ross/Paul. Either an unseen modifier is sneaking around somewhere altering things or there's a computational issue causing what you see not to be what you get. I'm sure they'll get it figured out-- they always do.

[GMPX]

I had a look at it and was like WTF, you need Ross for this one!

I've seen the log........:help2: :wtf1:

Cheers,
Ross

[ringram]

Just so that everyone knows what Im talking about check out the pics.
I set the tune and scanner so the cell is highlighted.
You can see from the Commanded vs Actual that VE is pretty close as they track fairly well.
The problem is that Im commanding 14.7:1 AFR in table 3647 and getting commanded 14:08:1 and actual 12:75:1 (ignore the commanded vs actual difference, just concentrate on the mismatched commanded values) the throttle transitions are me trying to shake the blighter off.
Funny thing is as mentioned slowing down to a near stop seems to reset it for a while and things will be ok. So the commanded value is being changed somewhere and not matching the table commanded.

Free dancing banana smiley for anyone who can fix this.

As mentioned I have done a full PCM reflash with OS3 from latest 7.3 release.

[Delco]

I've seen the log........:help2: :wtf1:

Cheers,
Ross

Ross remember we had that in the early days when we first did that in the custom OS , I thought we had solved that a lone time ago as I havent seen that on a log for a very long time.

It was holding PE mode as you can see from the log above .

[ringram]

Can I log anything else for you guys to help smash the bug?
.. Maybe OS4 might bring me more luck when its out

8-)

[BowlingSS]

Can I log anything else for you guys to help smash the bug?
.. Maybe OS4 might bring me more luck when its out

8-)

When is 0S4 coming out?:coool:

Bill

[ringram]

After Delco mentioned this could be based on PE not disabling I have rechecked my PE trigger settings and maxed out MAP as well as setting PE delay to max. It shouldnt have triggered before, but certainly now shouldnt. Ill report back if it fixes the problem now.
Basically Im not using PE and just doing enrichment from B3647 based on MAP only.

[Extinct]

What does it do when you command 14.7 using the Bi-directional controls ?

[ringram]

Im pretty sure it would, problem is it might take an hour or so to show. Then Im usually driving at speed on the motorway, so messing with bidi controls isnt the best idea. Also it resets once its gone into idle mode.

Im messing with PE triggers and so on at present, seems ok at the moment, but wont know for a while.

[bK]

I had a very similar problem, strange Commanded AFR readings at different MAP and RPM settings, not reflecting the "Commanded Fuel vs. RPM" table B3647

What I found was the Lean Cruise was enabling and causing problems between the 2 tables of "Lean Cruise Target AFR" B3639 & "Commanded Fuel vs. RPM" B3647.

I noticed that the Commanded AFR would occasionally get stuck at 1 particular value and not move until I slowed down. The big clue was under heavy acceleration the commanded AFR never went under 14.63:1 when it should be nearing 12.5:1.

Since increasing the "Vehicle Speed Lean Cruise Enable" & "Vehicle Speed Lean Cruise Disable" commanded AFR has been exactly where it should be.

[Garry]

After changing to OS5, I noticed today that the commanded AFR was kind of off ... low 10's, rising with more RPM ... going through the tables, I noticed that the basetune I used to copy my old settings to had a pretty crazy AFR in B3647 ... after changing this to something like Richard posted above, everything looks a lot better ... so, I figure, I have to go back and do AutoVE again ... ?!

[ringram]

I still get some weird events myself. Ive found if I set PE MAP to say 85 and PE vs RPM to say 13.5:1 then it helps kick it back inline somewhat. OS5 seems much the same and I started from scratch. Well a HSV GTS scratch anyway.

[TAQuickness]

richard - would you mind sending me the tune and log as as well? TAQuickness at Houston dot RR dot com.

[ringram]

Will do. Ill just need to do some more logging. Ive not logged for a few weeks. (Sacrilage I know). Trying to fish out an old log and matching tune wont be possible as I dont track them nicely.
I was hoping for some v2 blackbox action to assist.

[TuneMaster]

Send me the tune and log as well, i'll take a look and see if i can help. steve304@hotmail.com

[ringram]

Ok fast forward 5 years and this problem still exists!
This time in COS5 for a Camaro 02.
Maybe its to do with Lean Cruise? Who knows. Still same issues as the logs above.
Looks like its a bug thats never been fixed.
Im using all the latest software etc.

Anyone got any fixes? I guess Ill try disabling lean cruise and maybe not use b3647. Seems a bit sad though.

-- Ok, I reread what Id posted before. Im going to set PE trigger to 105kpa on the basis that b3647 will take over and PE mode was the issue. In effect disabling PE and running 100% off the custom table. Will test and confirm results.

[joecar]

Hi Richard,

Just note that in OL, the PCM commands the richer of B3647 and B3618 (at the current operating point).

Yes, I would disable LC... you can achieve equal or better results using B3647.