Idle STFT confusion

[Mud]

Hi, I'm a bit confused about what's going on with my idle. Pictures tell the story...

Closed loop screenshot (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/idle_CL.png)
Open loop screenshot (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/idle_OL.png)

Closed loop, the NBs are pulling fuel out via the STFTs and making me go lean. Is this normal? The idle gets increasingly unsettled as it leans out. Also, what's causing the STFTs to suddenly jump back to zero?

Open loop, do my NBO2s look healthy? Seems odd bank to bank...or more likely am I down a spark plug/injector? I'm also not sure why it's progressively leaning out, ECT and IAT are almost constant.

Here's the spec:

LS1 with LS2 heads
222-226/115+3 .6"/.6" cam
102mm TB / FAST
Full exhaust, OTRCAI
LS3 injectors

I have just upgraded from COS2 to COS3, thinking I should try open loop around idle, although it doesn't seem to have taken (the closed loop shot is in fact using COS3, with .98-.99 lambda commanded below 1200rpm). I'll do more reading there, but is this a good strategy in general?

Files:
Open loop tune (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS2_0106_OL.ctz) / log (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0106.efi)
Closed loop tune (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS3_0111_SCL.ctz) / log (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0111.efi)

Any help much appreciated - I read a lot, but don't get much time to try things out...so sanity checks/pointed advice where to look would be great.

[Mud]

PS., no MAF.

edit: I guess not having B4205 maxed is why my COS3 semi-open loop didn't take.

[joecar]

...
edit: I guess not having B4205 maxed is why my COS3 semi-open loop didn't take.No, that's not it... something else is going on... try setting B4105 to 450 mV all across



SOL will occur in OL with any B3647 cell that is stoich regardless of B4205 (CL temperature enable) or B3801 (LTFT enable (implies CL)).

[Mud]

Ok, set B4105 to 450mV and put B4205 back. LTFTs are disabled. I'm still getting trims at idle, they only seem to be at zero on DFCO.

Here's the current tune (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS3_0114.ctz) / log (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0114.efi).

[Mud]

I'm honestly confused as to whether I'm CL, SOL or what right now, but the STFTs look active and the AFR is basically behaving. I should stress that the car drives mostly very well, I'm just trying to understand the inconsistencies between the banks and see how smooth the idle can be.

Fresh screenshot (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0114.png).

[joecar]

SOL only happens in OL when table B3647 has cells that are stoich.

[Mud]

So should STFTs stay still/at zero when in non-stoich cells of B3647, or do they vary but are ignored? Am I missing a tutorial on this or should it be obvious? I'm commanding .98 lambda at 1200rpm and below in the above screenshot.

[joecar]

Also log HO2S21, SPARKADV, KR.

[joecar]

If you look at HO2S11 and HO2S21, you will see that STFT go to zero while HO2S11/HO2S21 stop switching;


i.e. STFT trimming occurs only while HO2S11/HO2S21 are able to switch.

[Mud]

I've added those PIDs back in, will log those tomorrow (the above logs were my minimal VE PIDs with few additions).

I can see that the STFTs are zero when the NBO2s go low in DFCO, but I still don't understand - I'm commanding .98 lambda at idle on B3647, so this should be out of SOL - but STFTs are still varying...am I in CL for these cells then? Hence my comment before on disabling CL by maxing out B4205. I want OL on idle.

[joecar]

During idle it is using desired air and spark to trim the idle speed, so HO2S11/HOSS21 may not necessarily be able to switch.


Also, log GM. EQUIVRATIO, EXT.WO2EQR1, EXT.WO2LAM1, EXT.WO2ST1.

[Mud]

Ok, thanks - will report back with more to go on.

[Mud]

Here (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0114_3.efi)'s my morning driving around. Still on this (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS3_0114.ctz) tune.

If my WBO2 is correct I'm a touch lean (and looks like I'm not adding in enough fuel at very high IATs - hottest day for a while here). The WBO2 is new though, maybe it needs recalibrating. The STFTs are still pulling fuel out.

Any pointers much appreciated. Too much spark correction on idle?

Apologies for DFCO and lean cruise being enabled...oh and not updating the VIN.

[joecar]

Your IAT jumps from 84°F to 135°F at frame 24900... something is wrong with IAT sensor, or harness/connectors, or PCM.

Your ECT also jumps in some of the places that align with the jumps in IAT... check all your grounds.

Some of the HO2S11/HO2S21 jumps also align with the IAT/ECT jumps... check that the NBO2 wires are not shorting.

EDIT: ah, I see that those jumps are where RPM has gone to zero and starts up again... lol silly me :doh:

[joecar]

EXT.WO2LAM1 shows some lean places (goes above 1.000), and shows absolutely no correlation to GM.EQIVRATIO...

which wideband do you have...?

[Mud]

Your IAT jumps from 84°F to 135°F at frame 24900... something is wrong with IAT sensor, or harness/connectors, or PCM.

Your ECT also jumps in some of the places that align with the jumps in IAT... check all your grounds.

Some of the HO2S11/HO2S21 jumps also align with the IAT/ECT jumps... check that the NBO2 wires are not shorting.

EDIT: ah, I see that those jumps are where RPM has gone to zero and starts up again... lol silly me :doh:

You scared me for a second there!


EXT.WO2LAM1 shows some lean places (goes above 1.000), and shows absolutely no correlation to GM.EQIVRATIO...

which wideband do you have...?

It's a Innovate MTX-L with a recently-replaced sensor. I agree there's no real correlation between those two PIDs, but all I'm seeing is GM.EQIVRATIO is one unless it has just been started, and EXT.WO2LAM1 has various DFCO, lean cruise and undoubtedly some sub-optimal tune in there too.

edit: Oh yes, and there's always some garbage from the WBO2 where it's still heating up (static [typically high] value).

I think I'll put it back in OL for a bit for troubleshooting and further EQ mod tweaking whilst the weather is so hot.

[Mud]

Here's today's OL tune just to show it isn't a million miles out. Log (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0116.efi) / tune (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS3_0116.ctz). No DFCO or lean cruise.

I maxed out B4205, and I see that this kills the STFTs that allow SOL to work. I still don't get why in CL I'm pulling out fuel when I already seem to be lean though...but given I'm making minor tweaks to EQ mods/VE along the way I will re-enable CL to see what happens...not sure if the O2s are just getting fooled by my cam overlap, although I thought it was pretty minor TBH. I also wonder if I need to fiddle with charge blending a bit, as I seem to need to alter B3633 every time the weather changes.

[joecar]

You scared me for a second there!



It's a Innovate MTX-L with a recently-replaced sensor. I agree there's no real correlation between those two PIDs, but all I'm seeing is GM.EQIVRATIO is one unless it has just been started, and EXT.WO2LAM1 has various DFCO, lean cruise and undoubtedly some sub-optimal tune in there too.

edit: Oh yes, and there's always some garbage from the WBO2 where it's still heating up (static [typically high] value).

I think I'll put it back in OL for a bit for troubleshooting and further EQ mod tweaking whilst the weather is so hot.When you replaced the sensor, did you do the heater calibration (remove sensor, power on MTX, power off MTX, connect sensor, power on MTX)...?

[Mud]

Yup, calibrated properly. It does say recalibrate after 30 days if I'm recalling correctly though, and it is probably going on for that.

[joecar]

...

I maxed out B4205, and I see that this kills the STFTs that allow SOL to work. I still don't get why in CL I'm pulling out fuel when I already seem to be lean though...but given I'm making minor tweaks to EQ mods/VE along the way I will re-enable CL to see what happens...not sure if the O2s are just getting fooled by my cam overlap, although I thought it was pretty minor TBH. I also wonder if I need to fiddle with charge blending a bit, as I seem to need to alter B3633 every time the weather changes.Yea that is strange because B4205 enables CL (closed loop), whereas SOL (semi-open loop) functions only during OL... however, if fueling moves away from stoich, then SOL cannot function, this may be what is happening.

I can't see why fuel is being pulled... try this: increase VE by 3%, and capture a new log.

[Mud]

Ok, that SOL should only be working outside of CL is part of the puzzle I wasn't completely understanding (precedence basically).

I will try bumping VE as you say.

e: not worth bumping, but realised RTFM fail - B4205 should be maxed, and B4108 disabled. That said, I think my idle spark may have been overly high - I'm investigating that avenue as part of locking in my ideal idle (and don't yet know if I'll end up CL or SOL).

[vbhero]

I thought for SOL to work at all, that you had to disable LTFT (B3801) and max out B4205.
If the PCM goes into CL, then B3647 is ignored, or at least that is what the COS tutorial says.


Also, what's causing the STFTs to suddenly jump back to zero?

Mine does this also running COS3 in SOL. It goes to zero, then starts pulling fuel for a bit then bam zero again. I only noticed it the other day while doing a RAFPN...

[Mud]

I thought for SOL to work at all, that you had to disable LTFT (B3801) and max out B4205.
If the PCM goes into CL, then B3647 is ignored, or at least that is what the COS tutorial says.


Mine does this also running COS3 in SOL. It goes to zero, then starts pulling fuel for a bit then bam zero again. I only noticed it the other day while doing a RAFPN...

I'm yet to see SOL activate, if I max out B4205 then the STFTs stay at zero. I've tried setting B4101-B4104 to zero, no difference. Would you be willing to post your tune for me to peruse?

Good to know I'm not alone on the creeping-lean STFTs at idle. I'm not sure if it stumbles slightly and goes into a protection/stall-saver type mode - the commanded AFR deviates from stoich briefly when the STFTs reset.

I got a bit sidetracked into playing with my idle spark and air (sudden warm weather combined with being OL prompted me to fiddle). My O2s are behaving differently, so either I've got some bank-bank variability or the sensors aren't right. I will swap out the O2s before re-enabling CL (/SOL if I can ever get it to activate!) to avoid the strangeness I see with my current O2s. My Innovate WB also isn't to be trusted, so I've got a Zeitronix to swap in. For what it's worth, in OL I don't perceive a problem with the idle, so I'm hoping I have a bad NBO2.

[vbhero]

What I am seeing in your log is SOL working fine. Your trims are working when Stoich is commanded, and zero'd when anything other than stoich is commanded.
I see the issue where your B3647 is commanding 14.39AFR (.98lambda) at idle, but your logs are indicating 14.63 commanded, and SOL working.
I think the root of your problem is why is it still commanding stoich at idle if you have B3647 set at .98.

I haven't seen that before, as I command stoich at idle.

I am new at this myself, and cannot offer much more than those observations. I'll be watching this one to see the out come :)

[Mud]

Which log? The closed loop one in the opening post has B4205 at stock values, and it's coming from a hot start so is just standard CL behaviour I think and not SOL.

[vbhero]

Which log? The closed loop one in the opening post has B4205 at stock values, and it's coming from a hot start so is just standard CL behaviour I think and not SOL.

post #13

But I have looked at them all and I am seeing similar patterns...

[Mud]

That one also has B4205 at stock, and shouldn't the STFTs be zero when hot idling (b3647 is non-unity for <=1200rpm)? In tune0114_3.efi I'm only seeing zero STFTs in DFCO or on cranking.

[vbhero]

That one also has B4205 at stock, and shouldn't the STFTs be zero when hot idling (b3647 is non-unity for <=1200rpm)? In tune0114_3.efi I'm only seeing zero STFTs in DFCO or on cranking.

You will only see zero STFT at idle, IF idle is commanding any thing other than stoich. From what I can see in your logs, you are commanding stoich at idle so the trims will be active.

Have you got a log there that has B3801 Set to Disable and B4205 set to 140 across the board? With all your logs / tunes, I cannot easily see one where you have sent SOL correctly (according to the tutorials) but that could because I am tired too LOL

EDIT: The reason I am going down the SOL path, is that COS3 will ignore B3647 unless SOL is correctly set up. So you're commanded idle fuel in that table would have no effect if SOL isn't working. Again, I am basing this on what the documentation says, I am no expert on the matter.
By rights full OL should also use that table too... if in full OL B4108 should be disabled too.

[Mud]

I'm sure I had convinced myself even following the tutorial to the letter I could not get zero-STFTs where B3647 commanded non-stoich (and as soon as I max out B4205 STFTs peg at zero). However, I've just gone back through the last 15 tunes and logs and there's none where I've got exactly the right mix of parameters required to demonstrate this - I'll go and flash a tune with B4101-4, B4205, B4108 and B3647 all set so as they should enable SOL, and will post a log next time I use the car. I'm hoping it's user-error as that's the easiest to correct/explain - thanks for your help.

[joecar]

I thought for SOL to work at all, that you had to disable LTFT (B3801) and max out B4205.
If the PCM goes into CL, then B3647 is ignored, or at least that is what the COS tutorial says.

...Yes, correct... SOL only occurs in OL.

[vbhero]

I'm sure I had convinced myself even following the tutorial to the letter I could not get zero-STFTs where B3647 commanded non-stoich (and as soon as I max out B4205 STFTs peg at zero). However, I've just gone back through the last 15 tunes and logs and there's none where I've got exactly the right mix of parameters required to demonstrate this - I'll go and flash a tune with B4101-4, B4205, B4108 and B3647 all set so as they should enable SOL, and will post a log next time I use the car. I'm hoping it's user-error as that's the easiest to correct/explain - thanks for your help.

Cool, hopefully it is something like that. This is where a roadrunner would be so handy :) I'll be interested to see how you go.


Yes, correct... SOL only occurs in OL.

Sweet, thanks for that Joecar good to see that this stuff is slowly sinking into this thick head of mine :P

[Mud]

Ok, these are just to show I'm still failing to get SOL: tune (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS3_0130.ctz) / log (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0130.efi)

[vbhero]

Ok, these are just to show I'm still failing to get SOL: tune (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS3_0130.ctz) / log (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0130.efi)

Yeah, you're right there is no trimming there at all. You appear to be in full OL.
What I am stumped trying to figure out, is why your commanded AFR is 14.52 for a large portion of your log. Looking at your tune, I cannot see where that is coming from.
So being the noob that I am, I think that it is time I bow out of this and observe what the Gurus come up with.

FWIW, In mine I have left B4101, B4102, B4103, B4104 & B4108 at stock settings. I am not sure if leaving B4108 enabled would make a difference as far as trimming goes.

[joecar]

Ok, these are just to show I'm still failing to get SOL: tune (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS3_0130.ctz) / log (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune0130.efi)SOL can only happen while GM.EQIVRATIO is 1 (i.e 1.00000, i.e all zero decimal places)...

while your engine is cold, your log shows there are periods (~30 seconds) where GM.EQIVRATIO is 1;

when your engine is warm, your log shows that EQIVRATIO is never 1.00000, log shows it being around 1.06.

[joecar]

i.e. so it cannot do SOL trimming because fueling is far enough away from stoich.

[Mud]

Thanks, understood. I've got a lot on my plate, and avoiding a long story, will return to this, I appreciate the help.

A possibly stupid question: is it ok to flash the calibration from a COS2 file onto a COS3 file (and vice versa)?

[joecar]

Thanks, understood. I've got a lot on my plate, and avoiding a long story, will return to this, I appreciate the help.

A possibly stupid question: is it ok to flash the calibration from a COS2 file onto a COS3 file (and vice versa)?I would first flash back to the OEM OS and then do the COS2 upgrade.

( but, COS3 has all the features of COS2 )

[Mud]

I would first flash back to the OEM OS and then do the COS2 upgrade.

( but, COS3 has all the features of COS2 )

Ok - currently back on COS2 (which was installed according to the tutorial), but for my COS3 OS install I used a COS2 file to flash on the calibration. I wasn't sure if that could induce weirdness. Possibly clutching at straws.

[joecar]

Should be ok to cal-only flash from COS3.

[Mud]

Google is failing me...any idea where these spikes in the commanded AFR are coming from? I notice they occur when the STFTs reset and the NBo2s go high.

Image (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/commandedRichSpikes.png) / log (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/tune079_mod_4.efi) / tune (https://dl.dropboxusercontent.com/u/95930/Monaro/EFI_live/BenLS1_COS2_0079.ctz)

[RADustin]

that is weird.

have you tried going backwards? swap back to stock OS, CAL flash on your COS. Save it, and try again?

looks like you are dropping out of SOL when the STFT max.

[Mud]

That last lot was on COS2, so standard CL.

Something odd is going on, and I'm sure it's daft. I suspect my NBo2s could be better, but after swapping in some cheap densos that wouldn't switch I put the old ones back in for the time being...

[Mud]

For some kind of thread resolution...I made a new COS3 tune based on an OE tune file (i.e., NOT flashing the calibration from a COS2 file), and all seems fine now. The weird dips in commanded AFR haven't recurred and STFTs look reasonable (touch wood...).

[joecar]

For some kind of thread resolution...I made a new COS3 tune based on an OE tune file (i.e., NOT flashing the calibration from a COS2 file), and all seems fine now. The weird dips in commanded AFR haven't recurred and STFTs look reasonable (touch wood...).Good idea :cheers:


Along the way I think I solved my hot start issue. I had used B3633 [after-start intake temp enrichment] to solve heat-soaking of my OTRCAI whilst in OL (mainly for dialling VE). I think my particular hot start problem was B3633 adding a lot of fuel for heat soaked IATs*, and B3628 [after-start enrichment decay rate] (which I rarely/never see discussed) being very high (slow decay) at high ECT by default - these combine to command very rich for very long. I don't really understand B3633 - it would seem on a cold start it takes ~1 min to get to commanded stoich/CL/SOL, but a hot start takes ~5 mins. These timescales seem bonkers; I've reduced B3628 to 4 for all engine temperatures for now...although with B3633 now zeroed this may be unnecessary (and why B3628 never seems to be discussed).

*I assume this means the fuelling model is off for an OTRCAI setup, since hotter intake should mean thinner air, and richer mixture, surely?Good analysis.



Can anyone confirm if A0014 is active in CL/SOL?I think it is, but let me check.

[Mud]

You beat my (not so) ninja edit...no matter, was just on the fence about my altering B3628, and I know A0014 is a table that does get discussed (so should be able to find the answer myself).

[joecar]

( some days I can type quick and accurate, other days my typing is dismal )

[Mud]

Cheers Joe - do you know if A0014 was ever fixed on 01290003? I can't find anything to indicate it was...so I guess if I need that table I will have to move to COS4 or COS5...

[joecar]

I don't think it was fixed, I don't see anything in the forum indicating it was...

COS5 gives you more options (and is fun to play with, if you have time).