Trigger errors and misfire

[MartinS]

Thank you Adam, I very much appreciate the effort you're putting into this.  No worries on the timing, I've got plenty to do getting the wiring harness back on the engine and no real time to do much until the weekend coming up.

[Adamw]

I've had a look through this whole post again to refresh my memory.  No new ideas are coming yet.  But can you do me a fresh log of it running and showing trigger errors etc, and if you've still got the scope can you do us another couple of scopes.

[MartinS]

Will do.  Will take a day or two to get things back together.

 

[MartinS]

Sorry Adam, got things back together but I've screwed something up and have to troubleshoot.  Might be able to get it done on Wednesday, if not it'll be on the weekend.

[MartinS]

Ok, back up and running.  Have a series of 'scope pics and a run log.  This is all with new star ground setup, rebuilt engine harness, RH cam position sensor switched to LH cam.

First, engine unplugged at engine harness.  Cranking:

https://drive.google.com/file/d/1_vS65YGb-xJLy4_X_Zja7TRynZJBsLvL/view?usp=sharing

Engine plugged in to ecu harness, ecu fuel injectors programmed off (ecu and everything else on):

https://drive.google.com/file/d/15ukH276R-if6W9lJFljYzD8QK1PO1l2A/view?usp=sharing

Idle:

https://drive.google.com/file/d/1r_V3FNw01fJm8ISBFaLRPox63yRiqhSC/view?usp=sharing

Running smoothly at approx 2000 rpm:

https://drive.google.com/file/d/12I-oaZy40L35MdMMamE1Mws8V1So6CX7/view?usp=sharing

Missing at approx 3000 rpm (is smoother than before harness rebuild but still missing, not sure what that means):

https://drive.google.com/file/d/1tcIT8UqYnLunJZI_QDk-Y1B9aac5lxBY/view?usp=sharing

Logging file from engine start to approx 3000 rpm:

https://drive.google.com/file/d/1EaAPzyINt1kaeRg_7DuVJOeAC5FPRUie/view?usp=sharing

 

[Adamw]

Do you think it is possible the crank wheel is damaged?  Or its been welded or magnetised or something?  There are the same two teeth that have low voltage in all your scopes - One is a couple of teeth before the single missing tooth and one is the tooth after the missing tooth gap.  

Im not sure how accessible it is in your car if if you have already inspected it, but I think it would be worth having a peak if you can.  Sometimes you can remove the sensor and see down the hole with a snake cam or mirror.  

 

I think it would be these teeth:

[MartinS]

You know, I also wondered about that but it's a brand new flywheel which has the trigger as part of the casting so I rejected it as a possibility a while ago...before nothing else has helped.  Easy to inspect as there is a cover play on the bottom of the bell housing.  Will look on Saturday.

I guess even if the tracing is asymmetrical, if those waves could be brought up to the same level as the rest, there wouldn't be any trigger errors?

 

Sorry, cover plate on the bottom not play

[Adamw]

Yeah the symmetry is a bit unusual but dont think it is the cause of the problem.  I think it is those two teeth that get shorter (or at least they dont increase with the others) as RPM increases so they get to the point were they are not reaching the arming threshold.  

And I dont think it is just the arming threshold set too high as in the 3000RPM scope it looks like one tooth barely exceeds 0V.  

Even if the flywheel teeth physically dont look different there are other possibilities which are difficult to test such as a casting flaw under the surface or the teeth have been magnetised somehow.  I have seen welded teeth do weird stuff for example and have heard stories of triggerwheels messed up just from welding done in close proximity.  But it would still be worth starting with a visual to see if there is anything obvious.

 

[MartinS]

Ok, this is a different line of enquiry altogether.  I'll have to wait until Saturday to have a look.  Anything you can think of I could do to test the structure or magnetism if there's nothing visual?

[Adamw]

You may be able to feel if a tooth is magnetised by dragging something like a paperclip or oldschool compass over it and comparing to other teeth.  Voids and other ferromagnetisim properties are more difficult to test though.  

[MartinS]

Ok, it's good news/bad news.  The bad news: I can't really inspect the crank sensor wheel with the transmission in place.  I have a small bore scope and tried today; there's just not a space big enough to get the scope in and look.  I can see it when I remove the crank positions sensor but really only one tooth at a time making inspection impractical and nearly impossible.

The good news is twofold: it's not a horrible job to remove the transmission, and the sensor plate is not part of the new flywheel.  It is a pressed metal plate which sits over the end of the crank with the rim bent over 90 degrees and ridges pressed into it for the sensor to read.  I honestly did not inspect it closely when installing the new flywheel so my bad.  

It most definitely could have been damaged by the wreckers yard when they removed the engine from the vehicle, separated it from the transmission, and moved it around in preparation to ship.  

I'll order a new plate on Monday morning.  Hopefully will have it by the end of the week and will install and test.  I'll let you know if there's anything obviously wrong when I take the old one out.  

Fingers crossed!!

[MartinS]

Ok, trans off, trigger plate/wheel off.  Looking at it, looks fine.  I definitely did inspect it initially; all coming back to me now I'm in there.  Looking more closely though at the two teeth you identify, they are definitely not ok.  Subtle but definitely slightly out of line.  

First pic, the teeth are in a straight line instead of on a curve.

https://drive.google.com/file/d/15ErdoFo8FGWkAyHm0oXhqqCmbwLnu1MR/view?usp=sharing

This one's a little more obvious but still subtle:

https://drive.google.com/file/d/1JB0iY2wWCrSjMRNAMFxwvtQU5n6DGi8u/view?usp=sharing

There are none of these in NA and Subaru won't commit to a delivery time.  Scrap yards don't want to take one off the engines they're trying to sell.  

So, I've bent it as straight as I can get it and will get it back together and run it tomorrow. 

https://drive.google.com/file/d/18BYLcgyB6Q1-7GakWHgdkaVunOjk5DyZ/view?usp=sharing

https://drive.google.com/file/d/1JdhFwpCKx_x6ZY81nxECQ6wESPHUDCVt/view?usp=sharing

Since the teeth which are slightly damaged are the ones you identify, I'm actually optimistic this is the problem and the fix.  Just hoping I've been able to straighten them enough.  Will let you know tomorrow or next day.  Fingers crossed...

[Adamw]

Yeah there's not much "magnetic mass" behind those thin sheet metal teeth so I suspect it wouldnt take much airgap variation to have a big influence.  Im optimistic you might have found the problem.  

[MartinS]

Back together, PROBLEM SOLVED!!!  First 'scope is cranking, overcorrected a little on the tooth before the flat, teeth after the flat pretty good:

https://drive.google.com/file/d/1RFacToRtrQYvwfWi_C6PackMtupQHNlN/view?usp=sharing

About 2400 rpm, tooth at beginning of flat (which didn't look like an issue before) not increasing in voltage much but faulty ones look good:

https://drive.google.com/file/d/1i46Dj4rmUqwSTuwemPbDM_0AVVo1sUTf/view?usp=sharing

This is about 3500 rpm, same deal with tooth at flat but no trigger errors, no misfire:

https://drive.google.com/file/d/1m1PoVHsTuFIjTxXAdju1sUyhdajFBP0g/view?usp=sharing

Had the engine up to 5500 rpm (which, realistically, it will rarely ever achieve in day to day use) and no trigger errors.  Ran a bit rough but I think because of lack of tuning, not trigger.

I'm actually in shock.  After 9 months of chasing electrical gremlins, I can't believe it was this.  

Do you think that tooth at the beginning of the flat will ultimately cause problems?  I'm thinking maybe I'll order the new trigger wheel and put it on in the future.  

Anyway, thank you so much for your help on this Adam.  You figured it out despite me!

[Adamw]

1 hour ago, MartinS said:

I'm actually in shock.  After 9 months of chasing electrical gremlins, I can't believe it was this.

 Dont forget there was originally a ground issue that was sending a big spike through your trigger signals which you fixed so the wiring work wasnt all in vain.  

 

1 hour ago, MartinS said:

Do you think that tooth at the beginning of the flat will ultimately cause problems?  I'm thinking maybe I'll order the new trigger wheel and put it on in the future.  

It looks like the shortest tooth is still about 5V high so I expect that should be fine. 

[MartinS]

Does the ECU measure from the zero point or does it measure the total swing of voltage in the trigger signal?  Just out of interest to understand it better.

Of interest, and a little too late for me, my friend with the oscilloscope has a friend who's a retired Subaru tech.  He called him to see if he had a timing wheel I could have in his pile of accumulated parts (didn't).  When my friend told him the problem he said something to the effect: "oh yes, the wrecker's yards around here damage timing wheels all the time.  They are still throwing engines around like it's 1960...".  Oh well, lesson learned.

Yes, there are positives.  I think restoring the engine harness ultimately will serve me very well as will the star point ground setup.  Also made me make sure all my wiring was correct, replaced some questionable plugs, and made sure all my sensors and actuators were in good condition.  Tidied up a lot of the routing of the wiring in the engine bay and checked a bunch of the systems...sooooo painful though!

Still, while it's 9 mos later than I'd hoped, that engine purrs.  Very exciting that I'll be able to move forward with the project now.  Can't thank you enough Adam.  

I'm sure I'll have more questions as I move into the tuning part of the process.

[Adamw]

21 minutes ago, MartinS said:

Does the ECU measure from the zero point or does it measure the total swing of voltage in the trigger signal?  Just out of interest to understand it better.

The VR conditioning circuit only really looks at the positive side (above 0V).  To be considered a "tooth" the signal must rise above the set arming threshold, then cross the 0V line on the way down.  How far it goes below zero doesnt matter.  Although it is usually more symmetrical than yours, im still not sure if that is due to the odd pressed sheet metal wheel or if it is the scope distorting the signal somehow.  I dont see it being an issue however since you now have a good voltage margin above the threshold and are still crossing zero by a decent amount the ecu should be happy.