Help! Car won't start when cranking under 180RPM

[robh]

I need your guys' help, I really don't want to take a half day but my car is not starting!
I'm running COP on a Microsquirt, have put on about 3500 miles on this setup.

My car starts fine at any temp (tested up to 32 F) when the starter cranks at ~ 180 RPM.
When it's really cold and I don't have another car to jump me my starter turns my crank at ~138 RPM (I just replaced my battery, so I know the new one is good). When this happens my car does not even TRY to start. I've recorded a datalog of this event and attached my msq.

Any ideas? is there a setting somewhere that says uS won't try to start below a certain RPM? I've found maximum starting RPM, nut no minimum.

[grippo]

You should have no problem starting at low rpm. The limit is somewhere around 60 rpm. So don't change anything in your msq except possibly your Cranking tolerance - try raising it to 60 or 70% and raise the After Start tolerance from 10 to 25 - 70%. That or ignition weak signal/noise are the only two things that I can think of. The lower the cranking speed the less signal from the VR sensor, so there could be no detections or just detections every once in a while. This would show up on the datalog as no detections, meaning rpm flat 0, or trigger+/- counting down. If you have a large tooth to sensor gap you could try reducing it a bit. If trigger+/- is counting up, that would indicate ignition noise maybe due to extra cold weather strain on the starter/ alternator.

[24c]

... If you have a large tooth to sensor gap you could try reducing it a bit.

This was the issue for me too...I found a direct correlation between detecting cranking rpms and my cam sensor gap.

[robh]

Thanks for the replies. I changed my Tolerances accordingly.

I took a datalog today over lunch so the engine was cold (not as cold as yesterday morning). In comparing the two the problem seems transparent. Trigger +/- is fine but note that on the old datalog injector pulsewidth stayed 3.4msec but the new datalog (attached below) shows injPW going to 16.3 msec during cranking. Not getting adequate fuel to start in first datalog.

Why is it that the injector pulsewidths didn't change in the first datalog when the car didn't start? Only noticeable differences are change in ambient temp and change in cranking speed.