Idle, Air Con and AFR wierdness

[arran]

I've driven it a few times now and the behavior hadn't changed. Maybe once every 10 or 15 times I returned to idle while letting the rpm free fall the car would stall or almost stall when the AC was on. It is perfect with the AC off.

I tried changing the activation method of idle VE and idle advance to be PID, or normal, tried changing the normal activation to be identical and slightly different, no affect. Also tried changing "leave idle valve closed above rpm" from 1800 to 1300 and back to 2000. /seem to dip more from 1800 to 1300 and less from 1300 to 2000.

I then turned off "Auto zero TPS" and after that for the rest of the drive (another 10 mins) it behaved perfectly.
Now as strange as that sounds, what are the chances that there is some occasional zeroing of the TPS that is upsetting the return to idle, just when AC is activated!!!???

I'll check it again through the next warm up cycle in a day or two, in the mean time comments are welcome.

[cmonref]

arran:

I've finally got the AC IdleUp hooked up and running. But I am having exactly the same problem you describe above -- CL Idle works beautifully except when you let off the gas and push in the clutch from higher RPM, after which the engine will die.

This thread dies with some changes and a promise to report on conditions after your next drive. Did the actions described actually solve the problem?

Brian

[cmonref]

Well, after much more reading and study of logs, there appear to be two distinct differences between CL Idle without, and with, AC operating. I'm going out to test this theory in a few minutes.

Both differences are due to the RPM drop rate being much faster when AC is running (after throttle off and clutch in from mid-range RPM, say 3000):

1. The tune state for WITHOUT AC must "catch" the rapidly-dropping RPM as it approaches the designated idle RPM. It takes a LOT more CL-Idle-Initial-Value, I think, and bullet-proof VE table values below the idle MAFload, from the mid-range RPM down to idle speeds.

2. The RPMdot values for WITH AC is much much higher than for WITHOUT AC, requiring a large increase in the RPMdot lockout value. Of course, this only matters if the RPM drop is "caught" and stabilizes the RPM at a representative value before the engine dies.

Brian

PS:
The tuning drive sorted things out. It took higher CL-Idle-Initial-Values than were necessary WITHOUT AC plus more AC-IdleUp dashpot than was in my tune. The RPMdot values did NOT need changing for WITH or WITHOUT AC. The current state of CL-Idle tune is in three ranges defined by throttle closed and clutch in from: low rpm, the rpm catches well above target idle then drops to target; medium rpm, the rpm catches at the target idle, jumps up, then drops to target; and higher rpm, the rpm catches below the target, jumps up above target, then drops to target. More reading about what settings will result in the rpm catching at consistent rpm before the drop to target.

Brian

[Sean123]

I also think is because of the AFR, as you mentioned that it goes full lean on decel with the AC on. cmonref defined the situation well.

[arran]

I have been looking at the tune around idle for many many months now and spent many many hours taking data logs most times I drive the car, and looking at the data logs. Following on from the conversation throughout the rest of this thread I have had a problem with the engine wanting to occasionally stall when it drops back into idle. Maybe once in 20 times it will stall, or almost stall (drop way below target). It is very hard to catch in a log. It is very annoying. I can repeat the same test over and over, cruise at 3000 rpm, lift off the throttle, push the clutch in, and 1 in 20 times or so the engine will stall or almost stall.

The behaviour seems to occur more frequently with the AC on, but it does also happen with the AC off.

With the amount of time I have spent tweaking settings without fixing the problem, I cannot help but think that there is a bug in the firmware but I have not been able to put my finger on what it is exactly that the ECU is doing that it shouldn't. So on the premise that there is some combination of my settings that is causing the problem, I turned off "Idle RPM timing Correction". I have been driving the car for a couple of weeks now and the stalling condition has not been evident. Prior to disabling the feature I tried less and more aggressive correction but this did not change the stalling behaviour.

So my conclusion at this point is that there is some combination of Rotary, sequential injection, sequential spark, staged injection, etc with "Idle RPM timing Correction" activated that causes the weirdness in idle behaviour.

Attaching MSL excerpts, all these were taken from the one data log (that is, the same tune)

[arran]

More MSL excerpts attached

[arran]

Another MSL excerpt, this one shows a stall

[jsmcortina]

I took a look at the last datalog posted.

Around 1304.267 the engine appears to be idling nicely on its own WITHOUT closed-loop idle (status2 = 0)
Then at 1305.221 the battery voltage starts dropping - why?
At 1305.424 closed-loop idle engages (status2=128) the timing jumps up and the battery voltage continues to fall.
At 1305.729 the idle valve starts to open but the RPMs have already started to tumble - opening the valve doesn't solve it.
Despite more idle valve duty and more timing the RPMs continue to fall.

Are you sure that more advance is what this engine wants at idle? I recall that Ken was using a fixed idle-advance of 5ATDC on his RX7.

James

[arran]

I took a look at the last datalog posted.

Thanks

Around 1304.267 the engine appears to be idling nicely on its own WITHOUT closed-loop idle (status2 = 0)

Yes

Then at 1305.221 the battery voltage starts dropping - why?

With my new found power of Custom Fields I can see that the fan has just switched on

At 1305.424 closed-loop idle engages (status2=128) the timing jumps up and the battery voltage continues to fall.

Engine RPM is falling, the alternator is under speed

At 1305.729 the idle valve starts to open but the RPMs have already started to tumble - opening the valve doesn't solve it.
Despite more idle valve duty and more timing the RPMs continue to fall.

Are you sure that more advance is what this engine wants at idle? I recall that Ken was using a fixed idle-advance of 5ATDC on his RX7.

James

No, I'm not sure. What I have not been able to understand is why it seems to return to idle perfectly 19 times, and on the 20th it will play up. When I have disabled Idle RPM timing correction it has been returning to idle every time