Raised from Steve's HD/Delphi EFI tuning thread
I think Steve pointed out 3 things from that example log
Hardtail picked up on O2 voltage and pulse width later, which funnily enough is what I've been looking at. I'll explain this in a minute.

I don't actually think this is a problem is it? - you have a long term adaptive value that used to think that cell was lean but the new CLI this time thinks it is rich. This strikes me as just sums and adaptive values changing over time, no?

My thinking being that initial fuel is calculated as VE*AFF but with the closed-loop O2 sensor feeding back a CLI that moves the new target fuel to VEnew=VE*AFF*CLI

The other two things he said related to from line 13541 onwards which is when EITMS kicked in. That deserves another thread because for me certainly it raised another couple of questions

 

So Steve said to look at the O2 voltage on lines 524 - 528 then again at 532 - 534

Basically the O2 was reading Lean for most of that particular second of time, but the closed loop fuel calculation was asking for lower NewVE ...which would seem to be wrong ?!

This, presumably, has to be skewed data or just outright wrong.

Having said that, remember I said my idle wasn't the smoothest, so I wondered if what Steve had pointed out was something needing fixed ...like a faulty O2 sensor. Funnily enough hrdtail mentioned this later but I was already looking at O2v vs Pulse width and came up with the following...



The X-scale is in seconds and the PV 14 samples per second should be more than enough for 8 cycles per second the engine was doing at idle.

I think I see something of a difference between the front and rear O2 sensor voltage traces but I'm not sure if either is actually wrong at all?

The reason I say that is if you look at the dotted trend line in both, the average O2 voltage over that recording was 0.6 for the front and 0.62 for the rear.

If stoic you expect the sensor to switch back and forward equally you would average 0.5 volts. My idle is at 0.977 lambda, so I'm thinking an average of 0.6 volts could be expected?

I think I see the PW responding to the O2 readings to some extent but it doesn't seem to be particularly aggressive?

What I'm saying is that I'm not sure I see a problem here. Having said that, idle does cough and surge ever so slightly so has this got something to do with it??

 

Nice. Thanks for sharing this. Gives one something to consider. I have a similar situation on mine. Got to thinking about it and started to make little changes on the CLB tables. Definitely made a difference. Still playing around with it but on hold for the time being due to a mechanical issue.

 

One thing Steve didn't get to was idle control, and this is something hardtail noticed in the example log.

If you look at the spark timing it is flailing around between 14 and 22* and this is what I thought might have been wrong with the smoothness of the idle. The timing tables ask for 20* at idle (I have left closed throttle timing table disabled as was stock)

From what I can see (in the PV anyway) is there are two tables called "Idle Spark Control Gain" and "Idle Spark Control Max" that use timing to help smooth the idle.

I don't think they work!!

2014 stock is gain=4 max=6, 2016 stock looks like 8/12, and I've also tried 1/4 and 2/4 and couldn't see any significant difference to idle stability.

 

I am not up on the PV. But I do know the closed throttle spark table in another producr does make a huge difference. Took a while for it to sink in when I could not figure out what was going on with idle spark. Then the light bulb lit. I also used my VCI to help adjust the TPS.

 

I don't seem to have CLB tables in the PV for a 358 map??

 

You mention TPS, I read an old thread on HTT about adjusting the throttle position sensor voltage at fully closed and setting IVC steps correctly to improve idle ...I wonder if that is still relevant??

 

I am using the TTS VCI. Blue one. I also have the black one. I would think setting the throttle accurately would still be relevant. I have the cable system on my bike (05). I find it very easy to dial in the Throttle position for closed this way. It has a big effect on idle including that slight idle flare. At leasy using this method, every thing is as consistent as I can make it.

 

I would adjust all spark tables so they match. I would turn on closed throttle spark and make this match. I would make sure idle and surrounding cells are also in open loop. Perform warm up test again. Better/ worse? If this smooth's out idle. Try again set in closed loop. Better/ worse? Then I would leave closed throttle spark on and 1 by 1 start setting tables back to original starting with main.


This is going to take some time, and I wouldn't worry so much starting at the extreme cold going to temp management. BUT, this might tell you when/ why and what table.


Here is another thing to look at also. IAC. What is this doing? Is it staying steady or is it jumping around. If IAC, and spark is jumping around. Seems the ECM is chasing it's tail and this might be making the idle speed doing what it is doing.

 

Cheers Hrdtail I've been trying some of that already...

Front and Rear spark tables are more or less the same, certainly anywhere around idle or low load

I've not tried enabling closed throttle spark table - mainly because it is disabled in Stock and without a darned good reason I was for leaving it that way ...I don't actually understand the need for that strategy?

I tried increasing fuel by 5% in the low range which included around idle (which also dropped it into closed loop). I half thought this was slightly better, but I switched back to my normal map and it didn't feel any difference.

I read about the IAC getting dirty and doing funny things, so I cleaned mine (as well as the rest of the intake from a touch of oily residue after I realised the dealer had overfilled it again after the service)
IAC seems to move with temperature, but gently, no jumping around. (that can be seen in that sample log as well)

One thing on temperature - the idle is actually ever so slightly better if the temp is up around 280 compared to 230, and this is repeatable.

Another thing about temperature I see is that CLI at low/mid revs is within +-5% at or below 230, but up at 280 it is out to +-10% Maybe related, maybe not? Hints of instability somewhere?

I'm still working on it