A place to discuss hardware/software and diagnostic procedures

Out of ideas and techniques to find ECM fault

More
8 months 1 day ago #62812 by skydawg
Have a simple LS3 boat engine with GM E67 ECM. It’s running standard GM sensors on near stand alone harness. 

One day after years of trouble free performance, long after any prior work on engine, Started getting multiple faults for different systems that come and go, but  almost always have DTC related to CKP (erratic signal….specifically SPN 636 FMI 8 on J1939 DTC list….most GM marine engines use j1939 protocol) and most often ECT (shorted high). Engine also runs rough with erratic RPM +/-200 rpm and will die if throttle is advanced too quick. If I clear codes DTCs stay out and engine runs norm but usually again set


I have considerable experience in designing ECM systems for different applications and never had faults I couldn’t figure out or repair, but I am out of ideas and hoping for some suggestions. 

Here’s where I’m at….

Done all typical trouble shooting for ECM power issues and completed GM service procedures for most of the DTCs displayed with no issues found. All grounds from to battery terminal to engine and ECM good per GM manual. All standard voltage drop test good. Less than 2 volts drop at both ECM grounds during engine cranking. I have disconnected all sensor inputs, load tested 5v ref and sensor ground wires for the CKP and ECT in addition to continuity checks on all other harness ref and ground wires. All 5v ref circuits test good at sensor connector. 

The DTCs set when ignition power is keyed on without cranking or engine running…and can be cleared before engine is started but often return. Interestingly, when I look at freeze info when the DTCs set on key up, sometimes there will be over a 1000 RPM indication. Also, ECT temp is indicating a default temp as it’s getting out of range signal. … harness is standard GM quality with twisted pair on CKP, CMS, and knock sensors. I tried wrapping CKP wires in conductive tape and in aluminum foil with one end grounded,  but no affect. 

Also, ran all 3 CKP and the 2 ECT sensor wires on new set of jumper wires to ECM  passing harness completely, but no affect. 

Besides CKP fault, often get ECT fault….and ECT display remains default temp,… I can back probe ECT low reference/ground wire and jump to known ground and temp indicates correct. I tried doing same on CKP but still got DTC.  I also jumped ground to MAP and IAT ground wires at same time but no effect on CKP faults. The ECT ground wire at sensor when unplugged shows less than .02v but has same as ref (4.98v) as ref wire which explain ECT faults at start up and default display temp until it’s additionally grounded. Again, the ground wire test good between ecm pin and sensor connector under load…. The sensor also test good and reads correct with the additional ground as well as has no effect on CKP fault if both ECT wires are removed from ECM or sensor. 

My latest troubleshooting technique was to disconnect all wires from all 3 ECM connectors except the 12v inputs, both grounds, and CKP,  figuring I would plug in all other wires one at a time until CKP DTC set (obviously got a bunch of new DTCs but was expecting no CKP fault codes) but CKP fault still set with key on before any other wires connected.  Never seen this unless. ECM bad but I tried 3 different ECMs with same results and no ECM related faults ever set. Also, occasionally the engine will not set any codes and start and run normal for period of time which a bad ECM or sensor wouldn’t likely do. What’s so puzzling is still having a CKP fault, on jumper wires outside the the harness, with so few circuits connected and a solid ECM ground. 

I am pretty sure it’s a ground fault but can’t figure out what Else to try next. Expect that the ECM supplied low reference grounds are not sinking loads on power up or on some sensors and CKP signal is seeing key up voltage as a surge and engine rotation. I feel all DTCs are related to common problem, likely ground related or some positive ecm input back feeding a ground.

I am now considering that components on the engine may be magnetized, such as the reluctor wheel, but doubt it as CKP faults still set without CKP connected and engine occasionally runs normal. 

Any suggestions appreciated. 

Please Log in or Create an account to join the conversation.

More
7 months 4 weeks ago #62826 by skydawg
Update…. Ran separate wires from ecm to working test CKP (outside wariness) which I could hold in Hand and simulate a reluctor wheel trigger event by moving a wrench close to CKP with volt meter back probed to signal wire.

When wrench next to CKP voltage drops to about zero (as it should) and back to 5.0v when moved away (as it should). But, occasionally, the voltage would not change when wrench is moved, often staying at zero with wrench moved away from CKP. Moving the wrench produces occasional rpm indication but displays 0 most of time which it should consist at different rpm. It seems the signal is lost half the time because signal volts don’t change while moving wrench.

CKP has a solid ground and ref voltage per load testing with a test lamp, and neither changes when inducing rpm. So, both ground and ref normal but CKP not sending proper signal of 5v with wrench away and 0v near. I would expect a bad sensor but same results with known working CKP.

Not sure what could cause a good CKP to produce there results.

Any suggestions welcomed.

Please Log in or Create an account to join the conversation.

More
7 months 4 weeks ago #62827 by skydawg
Update…. Ran separate wires from ecm to working test CKP (outside wariness) which I could hold in Hand and simulate a reluctor wheel trigger event by moving a wrench close to CKP with volt meter back probed to signal wire.

When wrench next to CKP voltage drops to about zero (as it should) and back to 5.0v when moved away (as it should). But, occasionally, the voltage would not change when wrench is moved, often staying at zero with wrench moved away from CKP. Moving the wrench produces occasional rpm indication but displays 0 most of time which it should consist at different rpm. It seems the signal is lost half the time because signal volts don’t change while moving wrench.

CKP has a solid ground and ref voltage per load testing with a test lamp, and neither changes when inducing rpm. So, both ground and ref normal but CKP not sending proper signal of 5v with wrench away and 0v near. I would expect a bad sensor but same results with known working CKP.

Not sure what could cause a good CKP to produce there results.

Any suggestions welcomed.

Please Log in or Create an account to join the conversation.

More
7 months 4 weeks ago #62828 by skydawg
Update…. Ran separate wires from ecm to working test CKP (outside wariness) which I could hold in Hand and simulate a reluctor wheel trigger event by moving a wrench close to CKP with volt meter back probed to signal wire.

When wrench next to CKP voltage drops to about zero (as it should) and back to 5.0v when moved away (as it should). But, occasionally, the voltage would not change when wrench is moved, often staying at zero with wrench moved away from CKP. Moving the wrench produces occasional rpm indication but displays 0 most of time which it should consist at different rpm. It seems the signal is lost half the time because signal volts don’t change while moving wrench.

CKP has a solid ground and ref voltage per load testing with a test lamp, and neither changes when inducing rpm. So, both ground and ref normal but CKP not sending proper signal of 5v with wrench away and 0v near. I would expect a bad sensor but same results with known working CKP.

Not sure what could cause a good CKP to produce there results.

Any suggestions welcomed.

Please Log in or Create an account to join the conversation.

  • Noah
  • Noah's Avatar
  • Offline
  • Moderator
  • Moderator
  • Give code definitions with numbers!
More
7 months 3 weeks ago #62848 by Noah
I'm going to take a stab in the dark in here, did you change the crank sensor?
And, if this is a highway use motor vehicle PCM, perform the crank pattern relearn procedure?

"Ground cannot be checked with a 10mm socket"

Please Log in or Create an account to join the conversation.

More
7 months 3 weeks ago #62860 by bigdaddyjustin
Have u checked for alternator ripple I haven't seen it to many times but what your seeing sorta does line up with what ive experienced.

Please Log in or Create an account to join the conversation.

Time to create page: 0.220 seconds