Can't get an easy cold start #2

[billjam]

For some time, I have been struggling with difficult cold starting.
Engine is 3.6 Porsche with ITBs, twin COPs and MS3X.
Symptoms are really sluggish stop-start cranking with occasional backfire for the first start of the day. Often it just won't start.
Once the engine has run for a minute or so, it starts just fine for every start during the day and it runs perfectly and pulls very strongly. The running tune is pretty much perfect.
I should point out that the IACV isn't working properly but it doesn't make any difference to starting whether the valve's hoses are connected or disconnected (open).

I have played with cranking settings, replaced battery and starter and tried all sorts of other means of isolating the problem such as new battery connected directly to the starter with temporary heavy leads and rerouting crank and cam sensor wiring away from all other wiring.
You will notice that the cranking pulse settings are currently at 400%. I have varied this between 100 and 400% and it hasn't made any difference to starting. It wouldn't have made any difference during the tow starting anyway.
I have varied the cranking advance in 5 degree steps from 0 to 25 with no effect.
With spark turned off, cranking speed is around 170rpm.
Spark test mode fires all 12 COPs as it should.
Same with injectors.

Attached are MSQ and two logs.
One log is a failed cold start, the other is a run around the block following a normal start about an hour after the engine had been running.
I tow-started it today to see if taking the starter out of the equation solved the problem. It started, but not as easily as I expected. I've tow-started lots of cars and it seemed almost like the engine was locking up on compression (in second gear). I stopped the towing and tried the starter briefly to check that it wasn't locking up with too much fuel ... it turned over slowly on the starter, but no start. Back to towing, I tried third and fourth gears but the locking up was worse, so after a bit more towing in second, it fired up. It ran very roughly with lots of black smoke, so obviously too much fuel in the cylinders. After a half a block it was running sweetly.

Anyhow, can anyone see anything obviously wrong with the settings that might be causing the starting problem?

[turbo conversion]

In the no start log there are 12 sync losses in each attempt, 24 total.

In the log when it starts there are 2 sync losses, this needs to be addressed before moving on.

David

[billjam]

I think the sync losses are caused by the engine speed dropping to almost zero at each compression stroke.
I don't know, but to me, they look more like a symptom than a cause.

[billr]

Post a tooth-log of the "cold start". Have you tried wasted-COP? Tried disabling half of the coils, just one per cylinder?

[billjam]

Post a tooth-log of the "cold start". Have you tried wasted-COP? Tried disabling half of the coils, just one per cylinder?

I'll try and do a tooth log in the morning.
I haven't tried wasted COP. I just tried to find where to set this, but no luck. Where is this setting? I am happy to try anything at this stage, but I don't see how wasted spark could help.
I will disconnect one set of coils and see what happens in the morning. If you are thinking that the ECU might be struggling to fire the coils, I have already dealt with that issue. The Toyota coils I am using aren't driven directly by the ECU because it doesn't have enough power. The spark outputs are connected to MOSFETS that switch a higher power 5v supply to trigger the coils.

[Dennis930]

Bill,

Oh no, not sync loss issues again. I would pull one sparkplug per cylinder and see if you still get sync loss indications. What is the compression ratio? Have you verified the ground strap from the transmission to chassis?

What is your cold start procedure? If the IACV is not working are you holding the throttle slightly open?

[billr]

"I have already dealt with that issue. "

Yes, but did you really find the root cause? I mean no disrespect, but urge you to consider all possibilities.

[billjam]

Bill,

Oh no, not sync loss issues again. I would pull one sparkplug per cylinder and see if you still get sync loss indications. What is the compression ratio? Have you verified the ground strap from the transmission to chassis?

What is your cold start procedure? If the IACV is not working are you holding the throttle slightly open?

Dennis,
I'll leave pulling spark plugs for the weekend. CR is standard at 11:1.
Yes, ground strap is good.
Yes, I usually give it a little bit of throttle to start. Within a minute it is usually idling OK without any throttle assistance. It doesn't often get below 10*C here, so I can live without IACV if necessary.
I tried resolving the IACV problems a few months ago. I got close ... I got it responding in test mode but gave up again because I couldn't get any info on settings that I could follow.
I thought I had fixed the original sync loss problems, but they may be back. Perhaps I never completely fixed them! Anyhow, I'll provide a tooth log in the next post.

[billjam]

"I have already dealt with that issue. "

Yes, but did you really find the root cause? I mean no disrespect, but urge you to consider all possibilities.

Bill,
Sorry, I didn't mean to sound dismissive of your comments. Putting in a relay box (using MOSFETS) was something I had to do way back in 2013 I think, to get the COPs to fire properly. At that time, we couldn't get the coils to fire because the MS3X spark ouputs didn't have enough power to drive the Toyota COPs. Got over that hurdle

I have just taken a tooth log which may help. I haven't had time to look up what it should look like, so I don't know yet if it is good or not.
From looking at recent data logs, I was aware of sync loss during cranking, but I assumed that this was a result of the cranking speed dropping to almost zero at each compression stroke. May be not?

[billr]

No, the tooth-log does not look good. It looks like the focus should first be in that area, but I gotta ask... do you have 12 MOSFET drivers for the coils, or just 6?

Is your CKP sensor VR or Hall (or other)? I'm guessing this is a 60-2 crank wheel, but will go back and peek for an MSQ.

Post a tooth-log from when it is running, too.

[billjam]

There are 12 MOSFETs in the relay box which take power from two 12v/5v converters, one for top set of coils and one for bottom set.
I put LEDs on each circuit so I can easily see if I have 12v, 5v, spark output from MS3 and signals to each coil. When I run the ignition on test mode at any speed, I can see that all coils are firing consistently (or at least getting a signal to fire) and in the correct sequence. Same when it is running, LEDs all flash steadily and in sequence.
When I crank the engine, looking at the LEDs, sparking is definitely random and not all coils are firing, so I guess we are looking at sync loss while cranking. Now just have to figure out why.

splitter box 04a.jpg

Spark output splitter box a.jpg

[Dennis930]

Bill,

I can't remember, but do you have your crank sensor at the flywheel or did you install a missing tooth crank pulley? What brand sensor did you use? It would be interesting to see what a tooth log looks like without compression slowing the engine.

[billjam]

I will run a log on the weekend without compression.
Crank sensor is standard Bosch part for this engine running off the standard 60-2 flywheel. I think this is a VR sensor.

[billr]

Post a tooth-log while running, no need to pull any plugs.

[billjam]

I strongly suspect that the problem is directly related to engine cranking speed. When temp is low, say 10*C, starting is impossible, lot's of stop-start, very slow cranking with occasional backfiring. When temp is up around 20*C (like it is today ) it starts up with a few seconds cranking.
It just seems to me that there is a setting that is borderline OK around 20*C. Once the engine has run, all subsequent starts for the day are OK.
I might repeat that battery, wiring, grounds and starter are all good.

Anyhow, I ran the attached logs today.
One is cranking with no compression and another is with compression. Both of these had fuel switched off, so no firing.
The third is a start and run log.

Hopefully there is something in these that narrows down the problem.

EDIT: Files not working, deleted. Uploaded again in post below.

[billjam]

Guys, any comments on the logs?
Are we looking at R56 and R52 adjustment?

[DaveEFI]

None of those logs will open with MLV here.

[billjam]

Thanks for letting me know. I have no idea what the problem is with the files, they open OK for me in MLV, but I will check them and upload again.

[DaveEFI]

Those load OK on MLV. I'd certainly have a play with R56 to see if it can be improved. It's not usually a critical adjustment - I'd try a full turn CW and see if it improves. And another, etc - but keep a note of where you started from. Ideally, all the teeth pulses would be the same amplitude. Most number of turns from fully CCW I've needed on a V3 board is 8.

[billr]

Those load for me, but sure look screwy. With no compression it looks like a nice 60-2 wheel, but with compression it seems to morph into a 41-6 wheel. Is the missing-tooth area right near TDC, not in a more "quiet" part of the compression stroke? Does the starter really struggle that much to crank it? How about posting the MSQ/MSL, too? Even when idling it seems speed variation is quite noticeable; does this have an especially light flywheel?