Wobbly idle, slow cold start, can't fix, help!

[deadrange2001]

Car is an old Peugeot 206, 122 cid / 1998 cc. Car uses MS3 firmware 1.2.4, BMW 2-wire idle valve, stock MS3 MAP sensor, original MAT sensor. Injectors are Subaru Impreza WRX Denso blues (550cc).
Idle PID is set at P: 100, I:100, D: 100. Seemed stable for 2-3 days, then went haywire. Tweaking the injector dead time from 0.8 to 1 does not improve the situation. No problem with the car moving. I see also the "CL Idle off" flashing on and off (green/grey) when this "wobbly" behavior happens.
What I also find weird is that the car cranks on the first key but doesn't start, starts then dies at the second one and starts quickly at the third key.
Is this a bad tune, bad injector or bad fuel pump? Car has vacuum/boost FPR set at 3.5 bar @idle. Normally it runs with a Eaton M62 supercharger but that's now off until I fix the problems with under-boost behavior.
I attach tune and log.
Many thanks...

[billr]

Your "cranking pulse" is way too low. Did you see the tip about most engines needing 200-400 when cold, and at least 100 when hot?

[deadrange2001]

Hi Bill an thanks for the answer. Increasing the cranking pw did not help, actually resulted in a nice black cloud of smoke behind the car at (yet still) the third cranking attempt. Probably a cranking advance problem...
My worst problem still remains the idle :(

[phil21191]

The cold start behaviour sounds like it needs more After Start Enrichment to me.

I initially ended up doubling the values in the table from what came with the Base map on the MS2 for my 99 MX5

[deadrange2001]

Thanks, will try. Can anyone help me with the wobbly idle? It happens at hot engine though...

[turbo conversion]

The first thing you need to do is get the injector dead time correct, every time you change this every thing has to be retuned.

I would then start by disabling o2 correction below 2000 rpm to start with.

I would also use PWM warm up instead of closed loop idle until you achieve a stable idle.

The reason for this is the o2 correction and the closed loop idle are fighting one another.

Your VE AFR and SAT need some smoothing in the idle areas and ASE needs to be increased.

David

[deadrange2001]

Kind thanks!

[deadrange2001]

So... did all these, worked fine for a couple of days but it came back. Probably the key is in the logs, I can see the idle condition going on and off like crazy. Maybe I have too tight idle return-to-idle conditions?

[turbo conversion]

Can you post your current msq and data log?

David

[deadrange2001]

I have a bit of progress ! What I discovered is while at idle, if I turned the VEAL on, the engine comes to a stable idle.
I did some smoothing in the VE idle area with EGO off and still the idle is rough. I upgraded the firmware to 1.4, still the same.
Then I enabled EGO control in idle area altogether with IAC PID and seems somehow stable. BUT, upon getting to a normal level, it still makes some +/- "pendulations". Like if I have the idle@900, upon throttle release, it goes to 700, then 1000, then 800, then 950 then 900.This in around 4 seconds.
Basically I had P=100, I=100,D=100, sensitivity 500, correction interval 100ms.
I use now P=50, I=100, D-2, sensitivity 1000, correction interval 300ms.
All I want is quick fall to idle, no more sinusoidal pendulation.
After studying PID theory, I understand that:
P - takes you as fast as possible towards the desired target, if there is a difference.
I - approaches the desired target through +/- adjustments, like a helmsman (basically this is where PID was inspired from).
D - flattens the adjustments above.
Are there some predefined PID settings to have a close-to-stock idle falldown?
Many thanks!