(P0336, crankshaft position sensor circuit/The ECM detects that the CKP sensor signal is inconsistent for 10 seconds)

We had the privilege of taking in a twin turbo GTO owned by a member of our local club, and built by another member of the same club. This member that built the car, did not build the engine.

Just to cover the specs on the build, it’s a forged 6.0, DART pro1 heads, custom cam, Victor Jr intake, elbow, APS twins. New clutch. All the “supporting mods” done. The car is beautifully built. Excellent attention to detail. Very well constructed.

I didn’t get in on this thing until just recently. I was on our local board, and saw that the builder was having problems getting it to run. Seems it was setting this 336 code repeatedly. The builder and I went over a few things. I explained that a crank relearn had nothing to do with it and wouldn’t fix it. He said he was done with it and was sending it to me. He had spent a great deal of time trying to figure it out.


This is where the fun begins. Just a FYI, this is the kind of challenge I thrive on.

So, the car comes in and we begin by running it, and seeing exactly what it does. It runs a bit rough, and the check engine light is blinking at me. It has no codes at this time. I pull the tune and it has the P0300 turned off. OK, turn that back on, run the car, and it sets the P0300. Easy enough. Do a little tuning, smooth it out some, and adjust the misfire tables .BOOM, that’s fixed.

We let it run for a while, and it sets a 336 after about 10-15 minutes.

OK, next step, lets check the crank sensor airgap. We pull the starter, the sensor, and use our trusty little tool we made up to check the gap. We came up with about .050. Right about what it should be.

Next, before we put the sensor back in, let’s check the reluctor for run out. We turn the motor over by hand, and see there’s one spot in the reluctor that is bent towards the front of the car. I make a mental note of this, and we put the sensor back in. We tighten the female pins in the connector because they had spread open (typical GM stuff) and put the car back together. We back it out and let it run. And run…..And run………….And run. The whole time it’s running, I’ve got my o-scope pinned into the crank sensor signal return. Everything looks lovely. Nice clean pattern. I rev the motor several times, sounds great, no codes. I think man, we’ve got this whipped. Loose pins in the connector. It ran for over an hour with no signs of failure. I go to move the car, push the clutch in, and it almost dies. WTF???? I take my foot off the clutch, it returns to normal. I try it again and same thing. WOW!! This is getting interesting. I try it a few more times, and it does it repeatedly. I put my foot on the clutch, and hold it there, and it sounds like it’s on a stutter box. After a few seconds, it sets the 336.

So, we do a wiggle test on the harness and don’t find anything. I check under the dash to make sure the clutch pedal isn’t rubbing anything. Good there. Next I check the clutch pedal switches (there’s 2) for their impact on the computer. We tried jumping the pins, and even moving the switches without moving the pedal . No affect. Car ran perfectly .

Now we’re onto the mechanical affect of pushing on the clutch. We check the crank end play, as best we could with the engine in the car. That seems fine. Now we’re mentally focusing on that bad spot in the reluctor wheel.

Not one to guess, or swap parts, or do something and “hope it fixes it” I have to be able to verify everything. I believe there is a way to measure and quantify everything. So, now I’m racking my brain to figure out how to pin it on this one spot in the reluctor………..GOT IT!!!!!!

OK, if it’s one spot on the reluctor, it SHOULD only affect 2 cylinders, right? It takes 2 revolutions of the crank for 8 power strokes. There’s a firing event every 90 degrees. So, that one spot will affect 2 cylinders. In a DIS application, this would be referred to as companion cylinders.

Now, if we can figure out what it’s doing, and to which cylinders. With the o-scope again (my bestest friend) I scope the ignition, and we operate the clutch. I find 2 cylinders that are flat lining completely ignition wise, in other words the coils weren’t even being triggered, and 2 more where just looking kinda funky. Weren’t dropping out completely, but had a distinct change in the firing.

So now, will it repeat? It did, over and over. Same exact cylinders. How do we figure out if they’re companion cylinders? The same angle of crank rotation? We take the firing order, and place the second half under the first half. 18726543….Like this
1872
6543

Guess which 2 cylinders were dropping out completely? 7 and 4. The other 2 that were getting funky? 2 and 3

Next question is, how do we match up those with the bad spot? That’s pretty easy, we pop off number 4 (easier than 7 to get to) and get it to top dead center, and pull the crank sensor. So we do, and the bad spot is just a little out of the way. We had to rotate the crank just slightly, and the bad spot lined right up in the crank sensor hole.

We’re figuring the reluctor is out just enough, that when warmed up, and clutch depressed, that it’s pushing the crank just a few thou forward, and interrupting the signal.

We’re going to take a video of it using my bore scope and a VCR so the customer has some visual evidence to go back to the engine builder and crank company, hopefully for some help with the $$ it’s going to cost for figuring this out, and repairing it. We’re going to try bending the reluctor through the crank sensor hole, but I’m not real optimistic. They’re pretty rigid.

I’ll make sure and post the results as to whether or not we are able to get it repaired easily, and what the final outcome is.

 

Lil video.

http://s125.photobucket.com/albums/p...t=reluctor.flv

 

LOL, almost the same thing I went through about a month ago except we verified the reluctor wheel on the crank was straight but the crank was moving a whole heck of a lot fore and aft. Guess they forgot to check basic tolerances when they built it.

 

That's what I like; good, logical, diagnostics!

Not many know how to do it anymore (or would take the time).

Fantastic bore scope video too!

 

Nice work Ed! It is always nice to quantify the diagnosis with repeatable results.

 

Survey SAYS..............It's fixed. Bent the reluctor wheel back. It was a bit springy. We got it to where you can hardly see any variance in it. Runs like a champ. Push the clutch in and no change at all. Let it run for an hour, push the clutch in and perfection.

 

good job Ed,now put an ice pack on your head so the swelling will go down

 

That made me LOL. I s'pose I deserve that. It does feel good to conquer **** like that.

My head hurts just from thinkin that hard........

 

Very nice catch! And you thought my car was bad.

 

Your car WAS bad. I got this one done in less than 2 days. Had it figured out in one.

Luv ya Kevin. Good luck tomorrow.

 

Ed,

Excellent diagnostic detective work... ...illustrates understanding how a system works is something we should all strive for.

Can you post any waveforms captured from the crank sensor...?

Cheers
Joe

 

I don't have the software for downloading my scope shots. I lost the disc, and Vetronix wanted to charge me another 500 bucks. I told them to F**K off. You'd think after spending 10 grand with them they'd give ya the damn software. Anyways, I tried to take a picture of it, but it just ended up hard to see.

 

Could you see a distinct variation (just by eyeballing the waveform) in the CKP waveform that corresponded to the the bent tab on the reluctor wheel...?

Yeah, I hate software that makes you pay each time you reinstall it... I've had a few problems like that.

 

No variation the way I was originally triggering the scope. I had planned on putting it back together, run 2 channels and triggering off #8 coil which most likely would have put the bad spot of the pattern just to the right of the coil discharge on the screen.

The bitch of that though, would be the divisions to get one crank revolution on the screen. At 600 rpms it's making one revolution every 100 milliseconds. So, at 800 it's one revolution every 75 milliseconds. and in one revolution, you have 24 crank sensor pulses. Gets a little tight on the scope ya know? I could have possibly blown it up bigger. It would have been 90* or 6 pulses between the #8 coil event and the spot in the reluctor in question.

But, anyways, at that point I knew what I was after, and kinda ran out of time. I just made the fix and moved on.