Airflow mode / LTFT histogram

[dhirocz]

Been chasing my tail for months and I think I?ve about got my closed loop idle issues about figured out.

Still having issues with fuel trims lining up with the WB. It tends to dance around a bit and runs better in OL, but the car sees weather that changes enough that I don?t think PE mode at idle is a good choice. I?d like to build a histogram to log fuel trim error to the airflow table and dial it in. However, Im having issues figuring out the right way to set it up; that or there?s more than one way to do it and Im getting confused.

What I would like to do is dial in the closed loop airflow mode table and then tune the 02 switch points.

Last question; should this be done with stock proportional and integrator settings, then tightened up afterwards, or does that matter?

Car is a c5 z06 with longtubes, cats and a very mild cam. It is running hotter 02?s to keep them from getting lazy at idle.

[SiriusC1024]

I'm not sure what you mean. LTFT is part of the stock config in scanner. Looks like this:
ltft.png

Tune in CL, run in CL. Set switchpoints to 451mV while tuning. Use LTFT, STFT, and LTFT+STFT.

Usually the O2 PID's don't need adjusted. If they are lazy do incremental changes.

[dhirocz]

I want to dial in the closed loop airflow mode table. It’s mentioned here:

http://www.hptuners.com/forum/showthread.php?t=11333

Then I want to set my switchpoints. Hopefully it’ll tighten up my 02’s and keep them from dancing around so much. Engine runs great either tuning SD or MAF, the tune is tight; but reenable everything and go into CL and it all goes downhill.

[SiriusC1024]

Oh I see. To do like in post #12 from that link it would look like this:
ltft vs airflow.png

Have to log Dynamic Airflow. The airflow column values are taken from the table in the tune. Right click on the column values, Column Axis>Copy Labels.

I'm not sure what utility it has if there is already fuel trim error. Look for groups of like errors? That post continues by touching on the debate of narrowband vs wideband accuracy.

I think it would at least help define airflows modes are sluggish since you'll be able to view the table live.

Ah ok here. At the cursor point is where the narrowbands are stuck lean. According to that LTFT vs Airflow table...
lean narrowband.png

...then cross referencing the tune that equates to airflow mode 9.
mode vs airflow.png

So that particular one was at idle, between modes 8 and 10 (obviously) in the Rich/Lean vs Airflow tables. Adjust switch points 8 and 10 upward I'd say. That or redefine airflow mode 9 to 10 at 127lb/hr and adjust just 10. For this one if it was 451mV starting make it 480mV and see what happens.

[dhirocz]

Log dynamic? I thought it was dynamic for SD and mass air flow for MAF guys? Does it matter?

I’m going to take a closer look later today when I get a few minutes to play with it.

LTFT needs to be actually working so I guess it’ll be logged in CL, correct?

[SiriusC1024]

Log dynamic. That's what determines final fueling. It's always there.
- In SD, Dynamic = VE or VVE airflow [mass]
- In MAF, Dynamic = Mass Airflow [mass]
- In Hybrid, Dynamic = combination of MAF, VE, TPS rate terms interpreted by the dynamic airflow algorithm.
https://forum.hptuners.com/showthrea...ynamic+airflow
https://forum.hptuners.com/showthrea...ynamic+airflow

Yes need CL. Why else would you be using the narrowbands?

Heres another example. Say you consider the reaction too slow here at the cursor point (it's really not, but just for example):
slow action O2.png

Adjust Integrator Delay. Doesn't need proportional since there's a full sweep in voltage. It needs integrator because there is a delay in switching back to the switchpoints. Proportional and Integrator both are in effect. Proportional is how much response to how much error. Integrator is how much response to how much error for how long. It accumulates over time, and as enough time passes enough response is accumulated to go the other way.

So, here this occurs 63-95 lb/hr. That equates to airflow modes 5-7. Adjust the time allowed for error downward, focusing on mode 6. Or, reassign airflow modes so that you can focus changes to integrator delay for certain modes.
integrator delay.png

[dhirocz]

Whew. Get to learn something new. That’s ok, just want to tighten it up… and it looks like I have plenty to do here since my 02’s act plenty weird.

So I understand that it may sometimes make more sense to join different airflows together under the same airflow mode to simplify it. At what point do we reach the minimum number of modes I should be using? For example, I see that I tend to idle at 95, 32 and 63 is deep deceleration, up to 413 is light throttle and cruise and everything over that is moderate to heavy throttle. Or am I looking at this too hard? Other than the 02’s themselves, I am not sure what else these modes might affect in the calibration… just wondering why the need exists.

I’m a sucker for theory.

I’ll take a closer look today. I’m kind of eager to get off work and play with it now.

[edcmat-l1]

Then I want to set my switchpoints. Hopefully it’ll tighten up my 02’s and keep them from dancing around so much. Engine runs great either tuning SD or MAF, the tune is tight; but reenable everything and go into CL and it all goes downhill.

That would typically mean everything is not "tight". This is the problem with open loop tuning. Use the trims to dial it in. Don't worry about the fuel trims lining up with the wideband. Chances are they won't.

[dhirocz]

Is this typically how tunes are done? SD, MAF, spark, then walk it in with the narrow bands using fuel trims in CL? Or is my case one of those ‘dammit’ cases?

[edcmat-l1]

Is this typically how tunes are done? SD, MAF, spark, then walk it in with the narrow bands using fuel trims in CL? Or is my case one of those ‘dammit’ cases?

Skip the open loop tuning completely. Use the wideband for WOT only. The PCM doesn't care what the wideband says and most of them are wrong anyway.

[SiriusC1024]

I thought you were having difficulty with narrowband sensors at certain points that was preventing CL tuning. Yeah, don't do open loop. Tune in CL.

[dhirocz]

No it’ll stay in CL. I am having some difficulty with the sensors but I’m pretty sure it’s my settings. I did compare my headers with a friends headers tonight and those things are LONG. The collectors are where the original cats were located. No wonder I have these issues.
Should the 02’s switch fairly quickly at idle too or is is somewhat dependent on engine RPM as to how fast they should switch?

[SiriusC1024]

Hard to make a determination without seeing a log and the tune.

[edcmat-l1]

FFS post a data log and tune file.

[dhirocz]

FFS post a data log and tune file.

Hey man there’s no reason to get short with me. I’m not just trying to get this done, I’m trying to understand what I am doing and why and we all learn different ways. If that bothers you, if it annoys you, whatever, you are not obligated to answer. Just go about your business.

I will post a file at my earliest opportunity. I have a business to run during the week so I apologize for the delay. I want to make sure I’m giving up good data.

[edcmat-l1]

Hey man there’s no reason to get short with me. I’m not just trying to get this done, I’m trying to understand what I am doing and why and we all learn different ways. If that bothers you, if it annoys you, whatever, you are not obligated to answer. Just go about your business.

I will post a file at my earliest opportunity. I have a business to run during the week so I apologize for the delay. I want to make sure I’m giving up good data.

Hey man, it's part of the process. If you want help you post a data log and tune file. It's even a sticky at the top of the page.

[dhirocz]

The FFS part is not part of the tuning process. Unnecessary. Nobody is denying the need to post a log and tune file. That was not what I was talking about and we both know it. That being said, I do appreciate what help I’ve been given here considering I’m just learning this stuff. Just don’t come at me like that, please.

Like I said, I will post one as soon as I am able. I’m generally pretty busy during the week and my new 02’s are overdue for delivery so it will have to wait for another day or so.

[dhirocz]

Finally got a chance to play with it.

This is with the hotter post cat 02's in the front position. It's definitely more active, but I dont know what to make of it. I see that at the end, the 02's stop switching...the B1 02 eventually gets stuck lean, where you can see on the log the B2 02 gets stuck in the middle. Odd.

On top of that, my wideband sensor decided to crap out. So I guess I have something to do tomorrow other than figure this out.

z06 11_18_23 reenable 02_1.hpt
02 11_18_23.hpl

[dhirocz]

...aaand they both just stopped switching entirely. removing all changes and reverting back doesn't seem to help. It's like once they get hot, both tend to flatline at .450 mv...02 11_18_23 2.hpl

I'm going to let this cool back down all the way and see if this is still an issue once the car has completely cooled down, and if it happens again when it comes up to temp.

Do STFT's play any role here with 02 sensor tuning or is it just the LTFT's? My STFT's are still disabled, as it was in the original calibration.

[SiriusC1024]

If STFT's are disabled that makes LTFT disabled by default. LTFT is based on them.