Engine starting fine when cold, but not at all when hot?

[B1ack_Mi16]

This is a rather strange one.

Normally when I crank the engine cold I'll get the cranking RPM up in the Tunerstudio gauge.

However this evening I finally seemed to get the coolant system leak free and let it run till the water was 90deg C.
It was running perfectly fine. I then stopped it and was to re-start it to check cranking enritchment etc. at this temperature. But it seem to me that it didn't even register the crank sensor signal?
RPM field stays just at 0 rpm even when cranking. Not any data in the composite logger either, but the MS3 Pro is certainly turned on as it shows all the sensor data in the gauges...

How can this be?

Here's a small vid. of the cranking and composite logger:

http://s428.photobucket.com/user/B1ackM ... sort=3&o=0

[nathaninwa]

What kind of engine, sensor type? I've seen Toyota vr sensors not start when hot from age

[B1ack_Mi16]

It's a Peugeot 2.0 16v Turbo engine stroked to 2.2litres, not that the engine spec really matters for this issue though I think.

I have tried 2 different sensors, as the first time it happened I thought the sensor was shot.
However when left alone for a bit it starts again just fine.

It is this sensor btw.



http://www.aahmotorsport.com/fiat-ducat ... -494-p.asp

[Matt Cramer]

Is this sensor type a VR or Hall effect?

Have you installed any resistors in the sensor wiring?

[B1ack_Mi16]

I did install resistors on both crank and cam input.

I've spent some more time on this one, and I think there is some wiring issued or creep current going on.
I'm using the standard engine loom and it incorporates a electronics box that is used for fan control in addition to a separate sensor, also got some input for activation of AC or not.

So I presume that if I turn the engine off, and the water temp is above the limit for the orginal fan control switch, the car will not start.

Happened once yesterday and I unplugged this box, and it started straight away. So MS3 itself is not of any guilt it seem.
Not sure how to fix it, maybe I just need to remove that box.

[B1ack_Mi16]

I've got the engine running again today after soldering in the new steppermotor chip.

Anyway it starts normally from cold, having good crank signal.

But suddenly it starts misbehaving and it shows no crank signal what so ever in the composite logger or nothing on the "RPM Sync" "light" in the software either. Everything just stay grey.

Can back-feeding of current through either PWM1 or PWM3 outputs lead to such effects?
I think it's really strange that this is happening, but something must be amiss.

I use the PWM1 and PWM3 to trigger the orginal fan control relays. These relays can still also be triggered by the orginal loom and orginal sensors. That's why I ask....

Maybe it would be beneficial to install diodes on these outputs to make sure this really ain't a problem?

[B1ack_Mi16]

Regarding this issue, tried a third VR sensor today just to see if the sensor affect the bahaviour.

But it didn't. Once engine starts to get hot (80-85C), it still runs nice and trouble-free after started from cold.
However if I stop it and instantly try to start it again, no crank signal what so ever is picked up. Trigger Logging is totally empty.

Is there some way to use Tunerstudio as an oscilloscope to see what signal it actually recieve at the VR input pins when this occur?
It is a mystery to me, I thought it could be related to some back-feed of current through some outputs to cooling fans, speed sensor etc. but I have disconnected all these and problem still persists...

This one is making me quite lost.

[Matt Cramer]

The tooth logger lets you see the signal that makes it to the processor, but not the analog voltage coming in from the wiring harness.

[DaveEFI]

The signal from a trigger wheel is a pretty simple one as these things go, so even the most basic scope should be able to read it well. Even a scope adaptor for a PC, etc.

The snag being that MS can only tell you what is getting to the processor. So won't tell you if the problem is the sensor, or the sensor input circuitry.

[B1ack_Mi16]

Yes I suppose I need to get my hands on a scope to log the actual input to the MS3.
The processor is definately not seeing anything as the tooth trigger log is empty and the light for "cranking" is not turning green either.

[B1ack_Mi16]

Could it be as simple as I've wired the sensor the wrong way around? Or would that show up in some other way?

[billr]

You said this is a VR sensor? That should show something in the tooth-logger if the wires are reversed.

Edit: However, the picture you posted a while back shows a 3-wire connector on the sensor. That may very well be Hall. You really need to determine, for sure, which type it is.

[B1ack_Mi16]

Indeed a VR sensor. It is just a 3 pin connector in the loom whre the shielding is connected to the third connector in the plug. Probably extending the shield in the sensor cable I would've guessed.

I'll try to make a new trigger log, which one is most suited to log regarding determining if it is wrongly wired?

Trigger logger, Composite logger etc.?

[grom_e30]

tooth log

[B1ack_Mi16]

Added some tooth logs... but not that I understand a lot from it, apart from that the start attempt after engine shutdown at 90 degreeC coolant is just bollocks.

Toothlogger.zip

[slow_hemi6]

Stroker crank modified original or after market? Any chance the relationship between the wheel and sensor has changed? Might just be marginal when cold but as the engine heats and expands it might be too much gap/ misalignment to get a signal at slow cranking rpms.

[B1ack_Mi16]

Crank is just a different OEM crank from a diesel, the trigger is machined into the orginal flywheel as can be seen here, and cranksensor sits in OEM position in the gearbox clutchhousing.

Should be more or less failproof.



Sensor hole you can more or less spot here. I run the exact same setup on a different EMS and it has proven failproof and stable as long as the sensor is not defect.

[slow_hemi6]

Probably need to grab a scope and test it hot. Hopefully it's not something nasty like a crank thrust issue......

[B1ack_Mi16]

Ok, I tried today with connecting the VR sensor through a different cable, high quality screened one, but it didn't help my starting issue when hot.

I shot a small vid though, here: http://s428.photobucket.com/user/B1ackM ... sort=3&o=3

It looks like all the gauges actually freeze once the starter motor kicks in. And the it looses communication.
Gets back online when I let go of the starter, and somehow doesn't loose communication in the second attempt, but the sensor values do jump when I let go of the starter the second time....

I'm getting a bit fed up, but does it make sense to hardwire the ECU 12v feed just to see if that might be the problem?

[grom_e30]

it does sound like its cranking slowly when warm how healthy is the battery?

during cranking a hot engine is the ecu still getting 12v as the going off line seems odd maybe try running a new dedicated power and grounds to the ecu for testing.