Intermittent Closed Loop issue in the wet?

[vincentav8]

Hi all,
I'm a bit confused by this .... I have been running the same map for a while now without too many issues
(R33/RB25DETs1/z32ECU/z32AFM/550cc)

the map is set with o2 feedback temp set to max (disabled), and with all cells set to o2 ON and then the values set from there (ie: > 192).
Usually, it operates in open loop mode all the time, but in the last few mornings with the weather getting colder and wetter I'm getting closed loop type AFR behaviour right across the rev range under load.

The car surges and slows roughly every second, in time with the wideband which is bouncing up and down around stoich as if closed loop is enabled.

I'm confused because the map hasn't changed (well, I haven't changed it) and I don't really know where to start looking?

any help would be much appreciated.
Cheers, Andy

*****Update *******

ok this is weird ... in the morning, cold and wet, this is happening ... it's a 12 mile run to work and car is running at normal temp according to Consult but it happens all the way. No error codes.
On the way home from work in the evening it is running fine

[Matt]

Take a log (stream mode) with the O2 sensor enabled and watch the feedback from that

Also enable AF Alpha/Alpha Learn (Short term/Long term trims) to see if they are at 100%

Disabling the O2 feedback temp should stop any changes to the trims from occurring. Post the logs here if still unsure

[vincentav8]

ok thanks Matt,
I'll give that a go

[vincentav8]

Hi Matt, it was cold and wet again here today and I got a log done.
This was with closed loop disabled (O2 feedback temp set at max), yet the wideband and narrow band gauges I have on the dash showed o2 open loop behaviour ... this can also be seen in the logs.

any ideas?

I'm guessing something is getting wet and earthing, causing the ecu to go into to closed loop mode somehow?

cheers,
Andy

[vincentav8]

I did the same again this morning, but this time with the alpha-learn box ticked on connection to consult.
What I saw was:
on idle and very low rpm = alpha-learn was about 102
on light acceleration = alpha-learn dropped to about 96, and dropped to around 90 on further acceleration

I'll have a look at the logs tonight and will post them up (I can't load them at work)

During the run this morning, when cold starting, the AFRs were slightly rich as expected and not oscillating (also as expected)
once the engine got up to temp, the AFRs started to oscillate as you would expect if O2 feedback was enabled ... but I double checked and O2 feedback temp is set to 250.
The temperature gauge reading went up slowly from cold and settled at around 85, so I'm assuming the temp sensor is ok.

I then pulled over and turned ON O2 feedback by dropping the temp threshold to 52 and gave that a go ... and it was no different to the previous run.

The bizarre thing is, when I drive home tonight it will be fine ... ?

[bachig24u]

I have a customer with a similar setup.

Z32MAF, 740nismo's, Z32 ECU, RB25-S1.

the only way for the ECU to stop using NB sensor is unplug it.

often the mixtures vary from idle to 750-2000rpm causing fluctuations in engine speed.
Tweaking TTP tables can smoothened it out enough but I really think something is missing in the
calibration tables to disable feed back entirely. changing the temperature or highlighted values seems to have no effect.

[vincentav8]

hi, thanks for the info ...
I've also had a look at a few threads talking about disabling O2 feedback.
But what is unusual for me is that it only happens after a cold start.
Once the car gets up to temp and onto the standard maps, then I see O2 feedback oscillation with it disabled (by max temp threshold).
If I pull over, switch off and restart the car, it all works fine .....

[bachig24u]

I'll see if that's the case here also.

check if the cold start and warm start tables match.

maybe some influence there that is causing a conflict against TTP min/max values.