Commanded AFR changing by itself - table set correctly

FlashPro questions & answers specific to the 2006-2009 (US, UK, Asia) S2000
Post Reply
futureal33
Posts: 43
Joined: Wed Apr 21, 2021 11:28 pm

Commanded AFR changing by itself - table set correctly

Post by futureal33 »

Hi All,

I'm having a strange issue where my AFR CMD is changing by itself.
I've attached the log and map

Notice how the WOT Lambda Adjustment table is set to all 12.46 in the full throttle cells, yet if you look at the log at 11:06 the AFR CMD drops from 12.46 downwards (richer).

I can't work out why its doing this?

I've confirmed the correct map is on the car, as I know it looks like the log and the map dont match up - but the log is definitely from the same map as attached here.

One thing I did wonder.. I have quite a high K.Control here (I'm in the middle of tuning my ignition and knock sensitivity tables), is there any way that can influence the AFR?
Thanks for any help!
Attachments
Rev 3.8 log1.fpdl
(1.8 MiB) Downloaded 145 times
EnglishN.SD. Rev03.8 (13.2-13.0 Taper)(25Ign)(OEMSense)(FP 3.7.5.0) DONT USE.fpcal
(9.26 KiB) Downloaded 139 times
EFICU
Posts: 3295
Joined: Mon Feb 25, 2019 10:14 pm

Re: Commanded AFR changing by itself - table set correctly

Post by EFICU »

After roughly 20 revisions on this S2000 with no issues, the WOT lambda adjustment value is straying from the 12.46 commanded in the table and dragging the AFR to very rich along with it. After this, we went back to the last known good revision that didn't have this issue, suddenly it had the issue too. We both updated our FlashPro Manager software and attempted to reinspect by going back one more revision and it worked fine with no issues. So, we proceeded as we thought it was all fixed, the next revision had the issue all over again.

Oddly enough, I am having the same issue on my 07 TSX now after uploading a new revision today. Hopefully the Hondata guys can give us some insight here. Seems like corrupted files, but should we both delete our FP Manager software and re-upload maybe? Or is it my FP Manager causing the issue since all the files are coming from my computer? Not sure what else it could be, all the sudden the issue happens on two cars with no changes to WOT tables or anything seems really strange.

Here are a couple photo references from the OP's datalog to hopefully get some help on. You can see the WOT tables are the same in both calibrations, yet in the green circle we were getting the 12.46 value now the WOT adjustment is going down to 11:56 and richer.
Nick-S2000.jpg
Nick-S2000.jpg (2.19 MiB) Viewed 4312 times
Nick-s2000-2.jpg
Nick-s2000-2.jpg (2.34 MiB) Viewed 4312 times
User avatar
Hondata
Site Admin
Posts: 10435
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Commanded AFR changing by itself - table set correctly

Post by Hondata »

It's not some random change with a software release - we have not changed anything with the S2000 logic for some time and regardless we do not change the logic Honda uses for the WOT compensation.

It's not a corrupt installation file. Everything that is loaded into the ECU is deterministic so it will behave exactly the same if you uninstall and re-install the software.

It's not the individual FlashPro. Again everything is deterministic so a different FlashPro will upload exactly the same binary into the ECU.

The WOT compensation table is the base WOT compensation. There's probably another 25 tables and 100 parameters that ECU uses to adjust the AF target, which are used under varying conditions.
Hondata
EFICU
Posts: 3295
Joined: Mon Feb 25, 2019 10:14 pm

Re: Commanded AFR changing by itself - table set correctly

Post by EFICU »

Hondata wrote: Sun May 30, 2021 6:53 am It's not some random change with a software release - we have not changed anything with the S2000 logic for some time and regardless we do not change the logic Honda uses for the WOT compensation.

It's not a corrupt installation file. Everything that is loaded into the ECU is deterministic so it will behave exactly the same if you uninstall and re-install the software.

It's not the individual FlashPro. Again everything is deterministic so a different FlashPro will upload exactly the same binary into the ECU.

The WOT compensation table is the base WOT compensation. There's probably another 25 tables and 100 parameters that ECU uses to adjust the AF target, which are used under varying conditions.
So the thousands of revisions I have made and sent to people over the years, including my own car, suddenly the ECUs are adjusting the AF target down into the 11's on an NA S2000 and TSX based on these 25 tables and 100 parameters you mention within the matter of the last five days? Can you give me a some type of idea as to what the ECU would suddenly see after 25 files on the same car and the only thing you change is a couple degrees of timing at redline to do this all the sudden? Would you at least concede that it's a bit strange that his car would suddenly do this after two weeks of working perfect at 12.46 in the table, and my car just starting doing it after two years of tuning within days of the OP starting to have the issue? If this is common, there must be some type of example you can give me or else how can someone properly tune an NA K-series or F-series engine if the AFR can suddenly be dragged down to the low 11's AFR randomly when not commanded to via the base WOT table?

Nobody is claiming you guys did anything wrong, just trying to figure out the issue. I am helping him get away from a calibration from a tuner in which the WOT fueling was used without commanding open loop operation, among other oddities. I've spent two weeks with him showing him how this is the proper way to tune, as you guys mentioned to him in a post he made a few weeks ago. After successfully transitioning the car to utilizing open loop properly he is extremely excited about how all the work is going, now suddenly the compensation is dragging the AFRs super rich and he's concerned. I was just hoping to get some clarification and hopefully get him back on track. If it were one car I could understand something strange, but two cars including my own personal car doing it for the first time after two years seems strange is all.
futureal33
Posts: 43
Joined: Wed Apr 21, 2021 11:28 pm

Re: Commanded AFR changing by itself - table set correctly

Post by futureal33 »

Hondata wrote: Sun May 30, 2021 6:53 am It's not some random change with a software release - we have not changed anything with the S2000 logic for some time and regardless we do not change the logic Honda uses for the WOT compensation.

It's not a corrupt installation file. Everything that is loaded into the ECU is deterministic so it will behave exactly the same if you uninstall and re-install the software.

It's not the individual FlashPro. Again everything is deterministic so a different FlashPro will upload exactly the same binary into the ECU.

The WOT compensation table is the base WOT compensation. There's probably another 25 tables and 100 parameters that ECU uses to adjust the AF target, which are used under varying conditions.
Regardless of what's caused it, are you able to advise what I can to fix it? As you can see, the WOT table is set to all 12.46 and as you can in the log it's commanding a descending AFR down in the 11s...I can't work out why?
User avatar
Hondata
Site Admin
Posts: 10435
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Commanded AFR changing by itself - table set correctly

Post by Hondata »

Sorry folks, I thought I was helping by eliminating some possibilities and preventing speculation unintentionally becoming internet fact. And receiving an email 'bug report' implies that you believe it is a bug, does it not?
Hondata
futureal33
Posts: 43
Joined: Wed Apr 21, 2021 11:28 pm

Re: Commanded AFR changing by itself - table set correctly

Post by futureal33 »

Ok, so do you know what could be causing it?
Thanks
User avatar
Hondata
Site Admin
Posts: 10435
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Commanded AFR changing by itself - table set correctly

Post by Hondata »

Yes, as I outlined in my reply above.
Hondata
futureal33
Posts: 43
Joined: Wed Apr 21, 2021 11:28 pm

Re: Commanded AFR changing by itself - table set correctly

Post by futureal33 »

Could you give any suggestion what in particular might have triggered it to go richer than the WOT Lambda Adjustment table, could it be the knock control for example?
I'm just a little worried that it could have gone lean, I've no idea why it went rich, so it's left me a little bit concerned that it may just as well have gone lean... Without understanding what else could be influencing it
User avatar
Hondata
Site Admin
Posts: 10435
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Commanded AFR changing by itself - table set correctly

Post by Hondata »

Knock control or secondary o2 sensor would be my first two guesses.

Formulate a driving test which will consistently reproduce the AFCMD going rich. Then unplug the S02 and see what happens. Then reduce ignition timing and see if that makes any difference. If either do, put back to the original conditions and test again.
Hondata
futureal33
Posts: 43
Joined: Wed Apr 21, 2021 11:28 pm

Re: Commanded AFR changing by itself - table set correctly

Post by futureal33 »

Thank you :)
I had wondered if knock control could influence it, so you are confirming that a high knock control can make the car run richer through a richer commanded AFR? Secondary O2 sensor is disabled, the only thing different on this log to previous logs is a higher knock count and control %, but I didn't realise that knock control could affect AFR?
EFICU
Posts: 3295
Joined: Mon Feb 25, 2019 10:14 pm

Re: Commanded AFR changing by itself - table set correctly

Post by EFICU »

Hondata wrote: Wed Jun 02, 2021 10:07 am Knock control or secondary o2 sensor would be my first two guesses.

Formulate a driving test which will consistently reproduce the AFCMD going rich. Then unplug the S02 and see what happens. Then reduce ignition timing and see if that makes any difference. If either do, put back to the original conditions and test again.
Can you clarify a couple things for me Hondata, just so I can understand how the k control does this.

S2000s appear to have a higher k control range, it looks like 0-138? Where say an 8th gen Si's have a 0-100 range? Basically it looks like the S2000s have a values over 100, and I've never seen that on Civic Si's even when they came to me with a lot of knock from an old tune.

Looking at the OP's datalogs with another S2000 I did, the k control does appear to be causing this fuel commandment to be richer with a 110+ k control and it kicks in at 6000 and richens the afr to redline if needed? Both cars seem to kick in the additional fuel right at 6000 and 7000rpm from 12.46 to a richer value. Does this sound right, kicking in at 6000rpm or so?

Last question, is there a way to stop the rich commandment without bypassing the knock control? Or say cap the richness to 12.0 instead of going to to 11.0, or would you just allow it to do whatever it needs to do with a k control above 120 to protect the engine? I'm all for protection, so if you say to let it add fuel and pull timing when things get that bad I'm fine with it if that's how Honda intends it.

We literally went from a file with the same ignition/fuel values and the k control was at 56 with no knock, to a 120 k control with knock and a super rich AFR which is why all these became a concern. Went back a couple files and there is no knock again and k control is fine. Plus my 07 TSX started doing the same thing with a 66 k control, which I've never seen before.
User avatar
Hondata
Site Admin
Posts: 10435
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Commanded AFR changing by itself - table set correctly

Post by Hondata »

The Civic and S2000 have different knock control ranges; I don't think the reason matters. Lower is better on both.

First step for you is to verify what the cause of the AF change is. Most vehicles enrichen with knock but I can't say this is the cause yet.

Then the controlling logic and tables for knock enrichment are a rabbit hole of unknown size and depth.
Hondata
EFICU
Posts: 3295
Joined: Mon Feb 25, 2019 10:14 pm

Re: Commanded AFR changing by itself - table set correctly

Post by EFICU »

Hondata wrote: Wed Jun 02, 2021 4:48 pm The Civic and S2000 have different knock control ranges; I don't think the reason matters. Lower is better on both.

First step for you is to verify what the cause of the AF change is. Most vehicles enrichen with knock but I can't say this is the cause yet.

Then the controlling logic and tables for knock enrichment are a rabbit hole of unknown size and depth.
That's the kind of info I needed, thank you! Yeah I never want someone to run with a knock control that high, just wondering how it was doing that. It makes sense that Civics don't do it in my experience because they don't go over 100, but the S2Ks do it over 120 from the cars I was looking at logs for. Both S2000s have the issue with k control above 120, but don't do it when k control is 106. So there must be a threshold in there between 107 and 120 that causes the S2Ks to do this, which is fine, now that I know why it's fine. As mentioned, I'm a fan of OEM protections, so it's a good thing. I guess why all the sudden one file is causing issues over a previous file is irrelevant, now that I know that it does it for a reason is all that matters.
Post Reply