[Lightfoot]

Had an episode of cranking, no-start in the parts store parking lot. Ordered a replacement ECM. Three hours later came back to try starting and it fired up, so I drove it home.
After installing the 'new' ECM the next day, it started and I drove it down the road half a mile, returned to the door yard, shut it off and restarted it after waiting a minute. Everything seemed good.
Two days after that, I tried to go to town but the engine wouldn't fire again. It cranks great guns so it's not the battery.

The last check of the codes (using the SES light flashing method) before I pulled the old ECM, all I had was a code 31 Park Neutral Switch.
Unfortunately, I didn't re-check for codes after I drove it that short ways with the new ECM. It had no trouble codes right when I connected it up. But I have a habit of removing the fuse for the ECM overnight for theft protection in the summer time. This may have erased a potential code 41 or 42 for the Cam Sensor or Crank Position Sensor.

Now, since the engine will not re-start, I don't know if it needs to be running for the ECM to set the codes for either a faulty Cam sensor or Crank sensor, or if it would do it from just cranking the starter for more than 10 seconds. I suspect not, for the latter, but I'm going to start doing some testing at the ignition module plug for voltage to & from the sensors.

I replaced the (EST)
module four years ago, hoping it would cure an intermittent stalling & rough running after re-start issue that's been going on for a long time. I posted another thread about that sometime ago.
The issue continued with the new module, so I suspected the ECM might be the culprit, so that's why I decided to try that first when this no-start issue cropped up this week.

The vehicle is an '89 LeSabre with 3800 engine.
It has always run great, in between these intermittent stalling issues. I've put roughly 30K miles on it during nine years of ownership. That's another reason why I haven't suspected the crank sensor or cam sensor to be failing over the long run.

[Lightfoot]

P.S.
I had fuel spit out of the valve on the injector rail after trying to start the engine numerous times. Also listened to verify the pump was running when I turned the key ON.
The fuel pump was replaced 8 years ago when it completely failed.
The car has 150,600 miles on it.

I think the Code 31 might have been set the day the engine wouldn't start, on the way to the parts store. I could feel the engine acting like it was going to stall while the transmission was 'hunting' between 4th and overdrive. Then it did stall when I came to a stop at a yield corner, and I threw it into Neutral and re-started it (standard procedure whenever this old glitch happened in the past).
From there I drove it to the parts store to air up the tires, which was another 2 miles. The no-start issue was totally unexpected.
I was on my way to go someplace else, not related to diagnosing car problems.

[salem1912]

Many moons ago I remember replacing a lot of Ignition Control Module (ICM) under the coil packs for no start/ intermittent starts.

[Lightfoot]

Originally Posted by salem1912

Many moons ago I remember replacing a lot of Ignition Control Module (ICM) under the coil packs for no start/ intermittent starts.

Thanks for the reply.
Yes, this is what I replaced in 2016. I just called it EST (Electronic Spark Timing) per the GM Service Manual.

Before I changed that module, I was shopping for a new ECM. The parts store salesman at the 3rd or 4th store I'd been to suggested the Ign module after I explained the symptoms my car was experiencing. I had never realized there was a seperate module for that, other than the coil packs. I had checked them all out with an ohm meter and they had passed the test.
So I changed my strategy and went with the ICM, even though it was more expensive than a new ECM.

I have a $200 scan tool that was showing 15 codes from my OBD I port.
I reasoned there was NO WAY my engine would even run if all that was accurate. So I was suspecting that my ECM was just mis-behaving, while still allowing the engine to function 95 % of the time.
When the car just wouldn't start at all this past week, I figured the ECM had probably just sh;t the bed, finally.

Whenever I'd had the stalling problem in the past, the engine would always, always re-start. It would just run extremely rough (like on 3 cylinders) and not idle below 2,000 rpm until the ECM went into open loop.

[CraigGT37]

The crank sensor will not always set a code. I used to replace them frequently when i worked at the dealer. Try cranking it while watching the live data screen on your scan tool. it should show rpm while cranking.

[salem1912]

Originally Posted by CraigGT37

The crank sensor will not always set a code. I used to replace them frequently when i worked at the dealer. Try cranking it while watching the live data screen on your scan tool. it should show rpm while cranking.

Also if car has a tach you'll see rpm (quick check).
How is the harness look? They start to deteriorate on older cars, grounds look good?

[Lightfoot]

Originally Posted by salem1912

Also if car has a tach you'll see rpm (quick check).
How is the harness look? They start to deteriorate on older cars, grounds look good?

That's a helluva good tip about the cranking rpm, but I don't think my scan tool shows data like that. It is designed mostly for OBD II but it came with adaptors for OBD I plugs for Ford, GM, Chrysler and a mainstream Japanese make. I think it only displays the trouble codes and what they signify. Anyway, that thing has never shown any less than the 15 codes for that car when I plugged it in five years ago. I've never been able to clear them out (of the scanner). Even when it's not plugged into it, if I enter in the Year, Make, Engine Digit, the bloody thing just shows those same codes.
So anyway, I hadn't checked the actual codes from the dash light readout for a long long time because it's such a PITA. It's virtually impossible to ground out the connectors unless you remove the whole kick panel under the dash.

I wasn't really paying attention to the tach when I was cranking it. I wanna say it was reading something, but I'm not 100 % sure.
The next question would be: If there is a bad spot in the wiring harness and the CPS is OK, would the tach still register? In other words, wouldn't an open or grounded wire between the CPS and ICM still stop the tach from working?
Maybe the tach only uses the 18X band signal?

My testing method should determine if there is a fault in the harness, if I extend the voltage checks to both ends of the circuits. There is a bit of engine grime on some sections and I can see that it has deteriorated the conduit where it goes close to the engine. It woudn't surprise me if that's where the problem is.
Also, I'm going to be testing the Cam sensor for voltage signal at the ICM terminal. I'm not sure if that is as likely to cause the no-spark, no-start condition. Would the factory tach show cranking rpm if the Cam sensor failed? Or if bad wiring from Cam sensor to ICM?

I already pulled all the spark plugs to make it easy to rotate the engine while I do the voltage signal checks at the ICM.
It's going kind of sketchy because I'm working with a rinky dink multimeter made in China. The test leads don't seem to have good conductivity. I made a makeshift tool to probe the back of the terminal connectors, rather than trying to prick the wires. So I'm not always sure if I am making contact with the wire ends.

Not Sure I want to try spinning the motor over with the starter at this point.
Maybe after I finish the voltage signal checks and put the plugs back in.

I appreciate you guys' input.

[AC445]

Hello, my buddy had a late 80's Park Ave years ago with a no start problem. Buick had crank sensor problems back then. Evidently they would heat up and fail so I had him pour cold water thru a long trans funnel onto the sensor when it failed.The car would start right up so I changed the crank sensor and problem solved. Good Luck

[Lightfoot]

I got through most of the voltage signal checks with my crude terminal probe tool (filed down paper clip) today.
The crank sensor is doing what it is supposed to, as far as sending a signal back to the ICM at proper intervals. There is one minor discrepency from the procedure I found online: The voltage coming out from the sensor is showing around 9.0 volts instead of the 5.5v on the page. It could be because their test was written for a '93 Century L27, vs my LN3.

I did continuity test on all seven harness wires between Crank/Cam sensors and ICM. They passed with flying colors, 0.00 ohms. So that meant I could double check the volt signal at the back of the crank sensor plug, which was much easier to probe for a reading than the ignition plug.

The only thing I didn't get done was verifying a volt signal from the cam sensor. That one takes two full turns of the crank to get the cam gear magnet to line up with the sensor. I had the probe tool set and was ratcheting the crank with one hand while holding the other lead to the battery Neg post.
Didn't see any voltage after I turned the crank possibly three times, so I gave up until I can take the sensor out and look for the magnet in the hole.

One thing I'm not sure of is, if there supposed to be a brief volt signal when that button lines up, or is the signal supposed to be there for 700 degrees of crank rotation and the only time the volts 'drop' to zero is when the button lines up?

I had it backwards in my head about the 3X crank sensor pulses, thinking they were only present for the short duration intervals. Found out it was just the opposite.

I also checked the ICM block for continuity to ground, since there is no seperate grounding wire to the module on pre L27 3800s.
I'm glad I don't have to unbolt that and clean the mounting base.
And I double checked the ohms across the coil packs today. Rock solid at 13 ohms.

[mgarblik]

Been reading along. That engine will start and run without a camshaft sensor position signal. It will just give you a long cranking condition. If a cam sensor signal is not present, it will just fire all the injectors in a batch fire sequence, 3 per revolution and the engine will run and store a cam sensor code. So it's something else. Does your engine have a Motorola or AC Delco coil pack? The Motorola one has all three coils in a single unit, the AC Delco one has 3 separate coils. Have had allot of no-start situations with the Motorola unit. Not the Delco one.

[mgarblik]

Originally Posted by mgarblik

Been reading along. That engine will start and run without a camshaft sensor position signal. It will just give you a long cranking condition. If a cam sensor signal is not present, it will just fire all the injectors in a batch fire sequence, 3 per revolution and the engine will run and store a cam sensor code. So it's something else. Does your engine have a Magnavox or AC Delco coil pack? The Magnavox one has all three coils in a single unit, the AC Delco one has 3 separate coils. Have had allot of no-start situations with the Magnavox unit. Not the Delco one.

Please see corrections in quote above. I meant Magnavox not Motorola. Would not let me edit.

[Lightfoot]

Thanks for the reply. It's nice to know the car would still run with a faulty cam sensor.
The first tests I did (not from the Buick service manual) showed that the Crank Sensor and Cam Sensor were interupting the voltage signal as they are supposed to. I also checked inside the mounting hole for the Cam Sensor to verify that the magnet had not fallen off the cam sprocket.
My car has the AC twin coils, which I double checked with an ohm meter to verify that they are not fried.

After tediously going through the four part test procedures in the service manual, and comparing end results in all senarios, I ended up re-installing the C3I (Ignition) module that I changed out for a new one in 2016. It road tested OK the first time I tried it, but on a trip to town last Friday, the engine stalled again after the third shut off and start cycle.
This time, I watched the tach while I tried to refire it, and it wasn't registering anything. I thought it wasn't going to start so I let it sit 15-20 minutes before I tried again. Finally it did start again but quickly stalled when I tried to put it in gear to take off.
It was acting out the same issue I've had in the past when it re-starts but won't idle. Only now it's more nerve racking because I've had these instances where it won't re-start at all, and I'm afraid it might die on me when I'm stopped to make a left turn in the middle of a busy two-lane highway.

I had a code 22, and code 42 in the ECM, but I think they might have been triggered by some of the tests I was performing, and they didn't get cleared out before I took the car on another road test. I did clear them out after I got home on Friday.
Let the car sit all day Sat, and took a chance driving it to church on Sunday. It didn't cause me any trouble that day, for just that 11 mile round trip. When I got home, I checked for codes again and there were none, so that's encouraging.

At some point, I may have to start digging into my harness to look for detereorated wires. Maybe after I do basic continuity checks on all the critical circuits between ECM and TPS, Crank sensor, and C3I module.