OK so its a "W" engine with a VCM engine computer and a discrete ignition module separate from the distr.

First some general guidance on modules and sensors:

Always check power and ground to the sensor (if used) or module. Then look for the signal into and/or out of the module at both ends of the wire. If no signal then disconnect both ends and check the wire for a short to ground, a short to power, a short to another signal wire or a break. If after all of that, you have not found anything wrong then the module or sensor may have failed and not the power, wiring, connectors or the signal in.

For your case, you have the second picture so:

Confirm no spark. Check ICM power and ground by testing for voltage across pink/ignition and black/ground. Be careful there are two pink wires. Then check power to the coil between the pink wire and chassis or battery ground. Do both of these while cranking.

If those are OK we need to look for the signal pulses into the ignition module on the white/ignition timing wire at the ignition module and then again at the white wire ground control wire at the coil. Do you have a scope or an incandescent test light?

If the truck is running the first time you try this thats OK because you will see the proper voltages and timing pulses so you will recognize what has failed when the truck dies.


George

 

I have gotten the truck back home. Unlike previous times, it is still dead even after 24+ hours. That may change.

I do have a test light as well as a portable oscilloscope. Is there an issue with overloading any of the signals going to the coil/icm with the incandescent light? Also, I don't have a "spark tester". My mechanic friend who came over one day did have a spark plug with ground wire attached which he used as a spark tester, however I've also heard that's not recommended and could damage things if used... But he has been using it on various autos for many decades.

 

Forgot to ask.

If the camshaft position sensor in the distributor has failed in some way, would that be enough to cause it to not start, or just run rough and set a code? When I replaced it 1.5 years ago, there was a code set so I just replaced it. No codes now as I stated, but I was wondering if there might be a failure mode in the sensor that wouldn't set a code but would cause the car to die and not start. It seems to have an input to the ignition system so I was wondering.

 

Couple of options on a spark tester:

A proper gap tester which will allow you to see a 1' strong blue spark as apposed to a spark plug gap
A properly attached and positioned screwdriver to get a 1" gap. Be careful not to lose the spark gap, best to start near the ground point
A spark plug. This tells you that you have spark but not how strong it is.

Test power feeds and grounds with a volt meter, use ICM ground for the module

Test the coil ground side pulses with a test light or see if you can see the pulses on your volt meter

Test ICM input pulses with an o scope probe or see if you can catch the pulses on your volt meter. A test probe here could overload the pulse driver from the VCM.


George

 

Crank no start can be:

Fuel pressure
Ignition
Fuel injector control pulses
Security issues
Timing

Probably in that order of probability. Lets find out first why there is no spark, such as no pcm control pulses to the ICM or ICM control pulses out to the coil. Then once we have that symptom then we find out why.


George

 

May have found the problem, after nearly a year.

The dying problem has been getting progressively worse over the last year. First it took a very long drive, and recently has been getting shorter and shorter. Hard to debug a problem when it's not currently failing and no codes. But now it has been dead for 3 days.

I read last night that a failing camshaft position sensor will possibly cause ignition cut off as well as fuel cut off. Since I had seen behavior pretty consistent with both of those, I focused on that today.

I looked at the signals going into the sensor. It seemed that power might be inconsistent. I seemed to be gerting power to the wire below the plug, but not right at the plug. Cranking, I saw something that looked like a signal at times. Ground seemed problematic too.

I noticed that the wires going into the male plug end were bent at quite a severe angle, and the 3 wire harness really is pretty short going there. I bent the wires a bit the other way as they entered the plug, and the car started up fine, after not starting for 3 days.

I cut the plug out, disassembled it, and saw the wire connections are crimped. I soldered those and reassembled. I stripped the wires right up to the plug and soldered on some longer ones. Then I soldered this assembly to the wires behind the distributor (very difficult to do).

The Blazer again started right up, idled just fine for 5 minutes, then I took a 4 mile or so test drive and it still worked. That's 2 miles further than anything it would do up to this point. I'm going to take it for a long drive soon and see how it goes.

New: Drove it for 30 miles to pick up a table in another town. Ran just fine, started fine even after sitting/heat soaking. More long drives coming up tomorrow, then 130 miles on Wednesday.

 

Well, not all is fixed.

The fix I made to the camshaft position sensor plug has fixed the problem with the car just dying all of a sudden. But then instead, it would run fine, and then start bucking like it lost fuel or ignition (probsbly ignition).

I replaced the camshaft position sensor itself, and now it is very down on power but doesn't buck any more.

Trying to fix multiple problems is difficult.

 

Why do you say "probably ignition"?

I can tell you that after a career of technical and business problem solving that assuming the cause with no evidence or symptoms is how you stay stuck on a problem and get frustrated. Test and eliminate in the most likely order is how you find this.

Intermittent problems are tougher but still solvable, you just have to get more creative. Here is how I would proceed:

Any codes and/or freeze frame data?

Watch the dash gauges when it stumbles, does anything lose power. If so it may be the ignition switch. If you suspect this You can then verify by running a wire from the pink power feed to the ICM to a voltmeter in the cabin and see if the ignition components lose 12V power during an event.

Get a fuel pressure gauge on the schrader valve and tape it to the windshield. What happens to fuel pressure when it stumbles? If it drops then its either the power feed to the pump, its ground or the pump.

Watch short term fuel trims when it starts running like crap. If its the pump and it lasts long enough you should go lean. The opposite for ignition.

This could also be a sensor problem. I would first check the +5V reference voltage and lo ref (high quality ground) to a couple of sensors. Then look at live data values to see if they all make sense - ECT, MAP, MAF, TPS, BARO (if equipped), ...

George

 

I just remembered that you said you have an O scope. You can also monitor ignition waveforms to see if they crash during an event. You can look at low voltage signal pulses or high voltage spark if you have a HV probe.


George

 

After driving 260 miles with the car not dying but significantly losing power and running lean on at least one bank (it finally through a code and I got a mechanic to scan it), it has returned to dying and not starting again.

The car has spent 45 working days at a mechanic in town (long time respected repair shop who has been maintaining this Blazer for 12 years) and they couldn't figure it out. I've spent 10 months trying to debug the problem, I get it apparently working for a while, then the behavior changes, then reverts back.

I'm pretty much giving up at this point. I don't have many of the special tools needed to monitor the car while running, and it doesn't throw codes unless I remove electrical connections. Until this latest behavior. I thought I was making headway then it died away from home again and I had to leave it. It might start tomorrow, it might not. But it's too intermittent and variable to ever trust again.