Closed loop at idle

[msaunders]

Noticed an issue with closed loop at idle today with a stock S13 KA24DE, stock bin, 53F01.
I'm emulating the factory NB02 with an LC-1 and it's working fine in closed loop while cruising. O2 voltage in Nistune fluctuates between ~.2 -.95v as you'd expect, with stoich AFRs being reported. Once coming down to idle from a run, the AFR remains ~15 and O2 voltage fluctuates indicating closed loop is working; however, after about five seconds or so of idling, AFR drops to ~13 and O2 voltage stays at .95v and does not move. If I give it a rev past idle it will get back into closed loop and the O2 will start fluctuating again, only to stick back at .95v and 13 AFR in a few seconds. Is the ECU somehow ignoring the rich reading from the O2 at idle? Any ideas? Thanks!

[Matt]

Appears so... question is why is it dropping that rich after idle? It sounds like its falling out of closed loop and needs a bit of a push to go back up

There must be some hysteresis which is in place when to stay in / out based on injection time or O2 feedback

If you alter your idle mixtures, does that help at all?

[msaunders]

Changing the idle cells does change AFR but it still doesn't want to hold onto closed loop there for some reason... hmm

[raddy]

I experienced very same problem with ca18det, after all it looks that case is big bore downpipe and exhaust what casue (probably) too low gas concentration to be measured by stock lambda sensor at idle...

[CoZZm0]

I have the exact same issue at the moment on my N/A SR20 setup (with a PLMS dual EEPROM board that i'm using as a temporary mesure so we can rule out it being a nistune hardware error). I havent done a lot of fiddling with the tuning at this point (i was hoping to diagnose it this weekend infact). I'll post an update if i locate the issue or have a fix. This also causes mis-fires (perhaps fouls the plugs a bit?) if sitting and idling for extended periods in traffic and its pretty annoying.

I had read before that changing the LC1 setting to have a slight delay rather than instant response has solved some other 'random' issues that people have had (this is done with the LM Programmer software).

I haven't had the opportunity to switch back to the factory O2 sensor and see if this issue persists (indicating a tune issue) or if its due to the narrowband simulation.

[KillerB]

In my experience with a CA18DET, idle will not stay in closed loop. When revving and returning back to idle my AFR logs first show an oscillating mixture and then slowly (after 5 seconds or so) the oscillation stops.
I can use the fuelmap to adjust the idle mixture, so it would definitely indicate it is idling in open loop.

[raddy]

but in standard form, ca18det have to use (using in real life)closed loop at idle, otherwise it cannot pass emmision test...

[KillerB]

it would make more sense to do closed loop at idle, i agree.

But i suppose if you get nice afr's at open loop idle you wont get any problems with emission testing. Mine currently stays fixed at 15afr idle.

[msaunders]

Hmm so it's not just me!

After digging through the FSM for the KA24DE it looks like idle is in fact OPEN loop.
Pretty odd but I guess that is the way it's done. My engine is stock with factory tune and it runs like this.
Would a rich idle then mean an issue with the tune? K adjustment?