Some people say that the best way to learn is from your mistakes. I disagree. I think the best way to learn is from someone else’s mistakes so I offer this one up. With this diag I tried to make some, what I thought were logical decisions which ultimately ended up giving me tunnelvision and leading me in the wrong direction.
What I knew about the truck:
2007 Nissan Xterra
Running rough under load
Transmission was just rebuilt and vehicle has not left the shop since
Previous shop installed a “new from the dealer” CKP sensor
Intermittently setting a P0335 code for a CKP fault
I test drove this truck and under light acceleration it ran rough but no CEL during fault.
I monitored fuel trims, O2s, and MAF. No abnormalities aside from symptoms caused by the rough running.
Next, I scoped CKP and CMP1 and found that when the engine was loaded, and during failure, CKP signal became erratic.
So, I knew that this symptom was only present when the engine was loaded. Since the CKP has already been replaced with an OEM and the symptom is still present I am thinking of a few likely potential causes.
First, the flexplate is installed incorrectly, loose, or broken in such a way that it is not making any noise and cranks fine with the starter however is shifting the teeth away from the CKP sensor when the engine is put under load. This seems like the most likely culprit as the transmission has just been out of the vehicle.
Or second, wiring is damaged or chafed somewhere in the harness causing an issue when the engine shifts under load. If you look at the failures I first saw it shows that the signal seems to be getting stuck high. I didn’t pay much attention to this at the time but the only way I could see wiring causing this would be if the ground wire was being pulled open. If this were the case it is even less likely for a circuit to be open intermittently.
Third, the sensor could have come loose or the gap could be incorrect.
Going with the easiest tests on the list I checked and verified that the CKP signal looked the same at the ECM connector and at the sensor itself. I also verified a good 12v reference and ground during the symptom. I removed the sensor which was not loose and inspected the mounting surface which looked good. It was easily to identify that this sensor was brand new in contrast to the rest of this 167k truck. These checks verified that my second and third theories were not the cause of the failure.
The only thing left was the flexplate but I wanted to find some more evidence. I was thinking maybe I would be able to see a pattern from the waveform just as it began to fail. Sure enough when I reviewed the recordings I found what looked like solid evidence that the cause of this fault was something related to the flexplate being damaged, loose, etc. You can see in the photo the flexplate has 30 teeth total spaced out in 3 groups of 10.
When compared to a known good taken from the same truck without any symptoms 10 teeth can be seen bringing the signal low. This would mean that, during our failure, the sensor is not seeing the gap between the 8th and 9th tooth on the flexplate. Because the flexplate has 3 sets of 10 teeth but the fault only happens once every 3 sets of 10 teeth it is occurring at the same exact spot in crankshaft revolution. This consistency tells me that it is certainly a mechanical fault as there is no possible way that the ECM, wiring, or CKP could have a fault that was consistent with only one position in the crankshaft.
I also captured this screenshot that I thought was interesting. I have seen the amplitude change in a pick up coil type sensor but never in a hall effect sensor. I was thinking that maybe this was the flexplate physically moving away from the sensor for a few teeth.
I was happy with my diagnosis and printed a few screenshots of what I had found to show the shop. I recommended that the shop remove transmission and inspect the flexplate.
I got a call a few days later that the flexplate had been replaced and it still did not fix the concern. I came back out to look at the truck and took more data. I verified that the concern did still happen under load. I verified that the 12v and signal ground circuits were not dropping out during the symptom and checked the harness again doing a wiggle test. I took all of the data home to study it. The more I looked at it the more I started to suspect it was just a new faulty CKP. When I called the shop the next day to tell them to warranty the CKP they had already installed it and the truck was running great.
In this case I thought I was getting super technical in trying to isolate the fault and I think I was just overthinking it. Either way I overlooked the fact that when it would completely act up the waveform was all over the place. I would be curious if anyone has ever used a similar strategy to successfully diagnose a bad reluctor or flexplate.
I was reading through this and everything you said made perfect sense regarding a faulty flywheel, especially the repeat in the waveform. It amazes me that in todays world we have to worry about every part we buy being a defect, including OEM. My question would be how in the hell would you have made a call to replace a brand new OEM sensor off the data you gathered?
I've seen enough bad waveforms for failing sensors that look like that static.
I've been compiling data from sensors and also can data hoping someday to write some training material of what a waveform may look like when shorted/open/excessive resistance or if a sensor is failing or the terminating resistor has failed. While I don't completely have my head wrapped around the differences I'm starting to learn to identify them.
I just had it in my head that the flexplate was damaged. It was very consistent but I guess I wouldnt expect it to be missing just one tooth but who knows.
Based on your testing results and my experience I would have said it was a faulty crankshaft position sensor.
I have had very similar scope captures from vehicles with faulty CKP sensors. A new sensor fixed every one. I believe the sensors were weak and not reacting properly.
I have been in the same position a few times. Each of my problems like this has been on a vehicle that has not had any parts replaced, other than the position sensors. Each of these problems were caused by aftermarket sensors which were installed.
I have also been in this exact position. OEM crank sensor on a VW. Bad out of the box.
Great case study, I am fairly new to scoping and interpreting wave forms is a bit tricky for me still. But I can definitely see the consistency in the drop out and why it would make sense to call out a bad flex plate. Thank you for sharing your experience, will keep in mind. Looking back at the original waveform, that was when the CKP was acting up under load, not pulling all the way to ground... and all the other waveforms with the consistent failure pattern, were captures of it failing in a different manner?
I guess my question is was this CKP sensor failing in two different ways? From me going through this case study, that is what I am taking from it. Thanks again.
No the captures where the waveform is consistent were taken right as the symptom began or under a light load right before the symptom would present itself. The first capture was taken when the failure was obvious and present.
My thoughts on the consistency in the waveform and why the sensor began to fail under load may have been because the trans was just out and the bushing that the new torque converter rides on (the concern wasn't present before the trans was removed) had just a little more play so under load it allowed the flexplate to shift away from the sensor.
Then at the tipping point where those snapshots of the consistent failure were taken I speculate that the consistency may be caused by that particular tooth having a slightly different metal composition or was worn.
Just a thought. I would love to hear others' opinion on this.