The Mule rides again (sort of) - pics.

[teamlseep13]

In addition to getting started on the 2.5" stainless X-pipe exhaust that I realized I could fit, I swapped out a couple of coil packs.

The "ignition drops" I've mentioned before were getting pretty bad. The car was difficult to drive in higher gears below 2000 RPM under load and pretty much wouldn't pull cleanly from 1000 RPM in 4th or 5th.

I swapped out one of the two coil packs I had out in the shed. It was worse. In addition to ignition drops being at least as bad as the original coil pack, it also had a bad coil and dead miss. Bleh.

The second swapped pack was much better than either original or the first swapped pack in terms of ignition drops and did NOT have a miss. I left it installed. It seems to have ignition drops based more on time than anything. When I was out for the test drive, sometimes it would pull cleanly from 1000 RPM in 5th uphill and not have any problems until it had a cluster of 3 or 4 drops at 1800, then keep pulling. Sometimes it would have 2 or 3 drops at 1200, then keep going. Sometimes it would be clean all the way from 1000 to more than 2000.
The original coil pack wouldn't allow the engine to idle for more than a few seconds before an ignition drop stalled it. The current coil pack allowed the engine to idle for more than a minute when I got back from my test drive. During that period, it experienced a couple of ignition drops, but none of them stalled the engine.

I've never seen the ignition drops happen above 2000 RPM. An ignition drop, if you remember from earlier in this thread, is a momentary loss of ignition. I can be seen by a flicking of the tach, heard in the exhaust note and felt via a split-second power interruption. Since the power delivery is actually interrupted and the tach flickers, the problem must be either in the trigger signal path or power. The trigger signal starts at the crank sensors then goes to the coil pack (ICM) to be translated into something the ECM can understand as well as driving the ICM's tach output, which goes to the tach. Somewhere in that path is the problem.

I guess it *could* be something esoteric with the ECM, causing it to drop the coil trigger signal while keeping the bypass line active. Not sure how the ICM would react to that. I suspect that's a long shot because the tach is showing signs and it is driven by the ICM using crank sensor signals, not by the ECM.

Ideas?

Usually what happens with these early ignition systems is the ignition control module (ICM) that the coils are mounted on top of starts to drop out the crank and sometimes cam signal to the PCM. This ICM is responsible for taking the cam and crank sensor signals, buffering them for the PCM and then send them on their way. PCM then tells the ICM what kind of ignition timing the engine needs and the ICM administers that.

I will bet either your signals(most likely crank) are getting lost either on their way to the ICM or in the ICM.
We have had plenty of ICM's cause drops outs in power as you are describing.

[The Dark Side of Will]

Usually what happens with these early ignition systems is the ignition control module (ICM) that the coils are mounted on top of starts to drop out the crank and sometimes cam signal to the PCM. This ICM is responsible for taking the cam and crank sensor signals, buffering them for the PCM and then send them on their way. PCM then tells the ICM what kind of ignition timing the engine needs and the ICM administers that.

I will bet either your signals(most likely crank) are getting lost either on their way to the ICM or in the ICM.
We have had plenty of ICM's cause drops outs in power as you are describing.

Thanks! Interesting...
What you describe is really what it felt like and it is a very unfamiliar failure mode... Crank sensors usually just die spontaneously after experiencing extended start times; failing coils have problems at high RPM/high throttle, etc.

Is there a definitive diagnosis for this problem?
I can scope the crank, cam and trigger signals at the ICM...
Are both the trigger for the ECM and the tach generated by the same circuit? The tach signal is dropping also, as I mentioned.
I've never tried to open an ICM... If it's able to maintain the signal at high RPM, but drop out at low RPM; that sounds like a capacitor failing. I guess we're not lucky enough for that to be a discrete component which can be replaced.

Good to hear from you. Don't be a stranger.

[teamlseep13]

Ya, I have been a stranger, my laptops HD failed and I have been spending about 200% too much time in other peoples garages working on their cars for them than in my garage on my projects...so I have been out but I try to pop in when I can.

The ICM creates a 4X (cam) and a 24X(crank) signal and sends these to the PCM. The PCM uses 24X to calculate engine speed, which it sends to the IPC for a tach signal. If the 24X signal goes, then the PCM has no idea how fast the engine is turning because it really never sees direct CPK sensor data, it sees a cleaned up signal made by the ICM.

As far as diagnosis goes, make sure the wiring to and from the ICM is sound. If it is, then you want to see what the ICM is putting out when this happens.
If you can scope the 24x signal while its happening and get a freeze frame of the event you might be able to see it.

Good luck!

[The Dark Side of Will]

My understanding was that the ignition and injection are still triggered off the 4x and the 24x is for misfire detection (detect crankshaft speed transients when a cylinder fires by having six impulses between events instead of 1).
The ICM pushes a 50% DC square wave derived from the CPS signal.
All GM's other DIS units have a 1x per event trigger signal. The 18x per revolution signal on the OBDII V6 engines frequently comes from an external crank trigger in the balancer.

How common is this failure mode? The original coil pack did it, and BOTH of the replacements I tried did it. They've all been on the shelf for a while; does the failure develop from sitting, from use or simple age without regard to usage?

You have a garage right now, so you're not doing so bad.
My places are rented out, so I have to go to my dad's garage 2 hours away to work over the weekends... hence things going slowly.

[Nashco]

I agree, wiring/connections sounds like the culprit.

Bryce

[The Dark Side of Will]

Wouldn't expect wiring to have such a strongly RPM dependent characteristic... Would expect wiring to present a constant problem OR to be related to engine movement.

[teamlseep13]

My understanding was that the ignition and injection are still triggered off the 4x and the 24x is for misfire detection (detect crankshaft speed transients when a cylinder fires by having six impulses between events instead of 1).
The ICM pushes a 50% DC square wave derived from the CPS signal.
All GM's other DIS units have a 1x per event trigger signal. The 18x per revolution signal on the OBDII V6 engines frequently comes from an external crank trigger in the balancer.

How common is this failure mode? The original coil pack did it, and BOTH of the replacements I tried did it. They've all been on the shelf for a while; does the failure develop from sitting, from use or simple age without regard to usage?

You have a garage right now, so you're not doing so bad.
My places are rented out, so I have to go to my dad's garage 2 hours away to work over the weekends... hence things going slowly.

The PCM uses both the 4x and the 24x for ignition timing control and the ICM creates a cam high signal for sequential fuel control. I believe it also uses the 24x for rpm calculation.

Common, not super common but the early engines had this problem. I don't think sitting on a shelf does it, mostly heat and age like any engine compartment electronics.

Also, if you aren't getting any drop out from the ICM and your wiring is good then suspect a PCM. '93 PCMs were kinda finicky.

[The Dark Side of Will]

Why have both unless they're used for different purposes?
My guess is that the use the 4x (1x per ignition event) for ignition and injection triggering because the 4 and 6 cylinder DIS units already worked that way and they knew they could make it work (They already had the software for it). The 1x per ignition is so integral to the way the DIS works, sets dwell, etc that they couldn't change it without re-designing the whole system.
The 24x (6x per ignition event) was breaking new ground and they wanted to test it out in a non-critical function--misfire detection--before they brought that function out across the product line. It was an add-on to what was essentially the same DIS used on the 4's and 6's.
I guess the PCM can do whatever it wants with the signals, of course... may trigger ignition based on the 4x and injection based on the 24x. However, my background in systems engineering leads me to suspect it's as I've postulated.

I should be able to scope the crank sensor signals, the 4x and 24x this weekend, now that the car will stay running through a few dropouts.

Rockauto shows one remaining AC Delco ICM on closeout for $220.

If I get that, do I have to get new coils also? I think I recall there being a failure mode in which old coils can kill a new ICM (or vice-versa)...

[teamlseep13]

Why have both unless they're used for different purposes?
My guess is that the use the 4x (1x per ignition event) for ignition and injection triggering because the 4 and 6 cylinder DIS units already worked that way and they knew they could make it work (They already had the software for it). The 1x per ignition is so integral to the way the DIS works, sets dwell, etc that they couldn't change it without re-designing the whole system.
The 24x (6x per ignition event) was breaking new ground and they wanted to test it out in a non-critical function--misfire detection--before they brought that function out across the product line. It was an add-on to what was essentially the same DIS used on the 4's and 6's.
I guess the PCM can do whatever it wants with the signals, of course... may trigger ignition based on the 4x and injection based on the 24x. However, my background in systems engineering leads me to suspect it's as I've postulated.

I should be able to scope the crank sensor signals, the 4x and 24x this weekend, now that the car will stay running through a few dropouts.

Rockauto shows one remaining AC Delco ICM on closeout for $220.

If I get that, do I have to get new coils also? I think I recall there being a failure mode in which old coils can kill a new ICM (or vice-versa)...

I am sure they use the 24X for misfire detection but another huge feature of the 24X signal is being able to locate the crankshaft position in the fewest engine revolutions possible. The 6 pulses per ignition event plus a 4x signal allows the PCM to locate crankshaft position in about half a revolution. Most other engine set ups during that day took much longer to start the engine because it had to spin a couple revolutions in order for the PCM to locate where the crankshaft was in relation to injection and ignition events.

Thinking about it, yes the 4X is probably the primary tach signal but I still think the PCM needs both in order to function correctly and it may be that loosing one or the other or both can cause this issue.
Looking at both will at least show you which is causing your issue.

One thing is, if while running the cam sensor is unplugged, you keep on running, so the ICM must use the two CPK sensor signals to generate a 4X signal in the event of loss of the cam sensor otherwise I would think the loss of your cam sensor into the ICM would kill the 4X signal and you would be outta luck with a tach signal right?

I dont recall putting old coils on a new ICM causing issues but ohm them out and make sure they are good anyway.

[The Dark Side of Will]

Remember that the PCM doesn't control ignition until the ICM switches the bypass signal. The engine starts and accelerates to 300 RPM with ignition controlled entirely by the ICM. What the 32x wheel and dual crank sensors allow is for the ICM to synch to the crankshaft within 1/2 revolution. The 4/6 cylinder DIS requires up to an entire revolution because of the double notch synch pulse on the 6+1 wheel.

Once the PCM receives two of the 4x triggers, it can start injecting fuel. It can actually synch faster on the 4x than the 24x. When it gets a pulse on the 4x line, it knows where the crankshaft is relative to TDC of a cylinder. When it gets two 4x pulses, it knows engine RPM and can begin injecting fuel in batch fire mode. If it's triggering off the 24x, it has to wait until it gets a cam synch signal to know where it is in the revolution.

The 24x and 4x can be generated entirely from the crank sensors. The cam sensor is not needed by the ICM. The only thing the ICM does to that signal is conditioning the analog pulse sensor input to a 0-5V square wave logic output for the PCM. The PCM only uses the cam sensor to synch the sequential injection. Without the cam sensor, it can drop back to batch fire seamlessly.

The GM DIS really is ingenius from the systems engineering perspective... it's too bad they never got to develop it to its fullest potential.

[teamlseep13]

Agreed!

Any luck on diagnosing your issue?

[The Dark Side of Will]

Haven't been back to the car yet. Was in Atlanta for drill last weekend and will be scoping the signals this weekend.

[The Dark Side of Will]

Is there any GM service information regarding this failure mode?

[teamlseep13]

No, just experienced it a number of time on some of the old turds that roll in the door.

[The Dark Side of Will]

That describes my car perfectly

[The Dark Side of Will]

I left the replacement coil pack that ran better than the original installed. Apparently it has a drain because the battery was *FLAT* after two weeks, while it had previously gone over a month and still been topped up. The automatic charger didn't want to charge it, so it looks like CarQuest is going to warrantee a battery for me...

My dad had two scopes at work. We tried to use the Fluke portable unit that "theoretically" could screen capture and output to a computer, but the PC interface software wasn't included with the $2500 unit and, Fluke being Fluke, cost $250 to purchase after the fact. Nice.

After discovering those two items, we switched to his Tectronix, which was a lot more agreeable than the Fluke anyway. At idle the 24x signal is ~250 Hz, so it's difficult to get a good trace on it... I'll have to play with the triggering more tomorrow, but I don't expect to be able to do very well. The idle speed fluctuates, so the frequency of the signal moves around a good bit and I can't get a steady trace, BUT it looked like the 24x was dropping out coincident with the ignition drops. I'll try to confirm that with the 4x tomorrow (greater duty cycle), but i'm pretty confident that I'll see the same thing.

My dad's previous clamp on ammeter disappeared at work and he ordered another one. That should be here this week and let us check for a drain through the ICM.
That was the plant ammeter, not his personal one... As the plant engineer, he gets to order lots of cool tools with company money.

[The Dark Side of Will]

X-pipe has been tacked up for a couple of weeks. Last weekend I was able to get under the car and take some measurements to help me with the layout sketch for the exhaust. I'm thinking I'll be able to build the entire system from the manifolds, including the X-pipe and dual cats with less than 36" of straight tubing.

[The Dark Side of Will]

The bends are tacked up, but neither of those assemblies is tacked to the flex section. I need to complete the manifold modification to the rear bank manifold (in work) so that I can have a solid place from which to start building this pile of 2.5" stainless spaghetti.

[The Dark Side of Will]

I built this fixture to hold a dual wall front manifold at 22.5 degrees so that I could hack it off in a band saw at the right angle to flip the outlet so that it would be pointed straight down in the installed position. This is a front manifold, but it's being modified for use on the rear bank.





Here it is cut with the inner wall removed



Trying it out... it'll need a little bit of shaping to go back on.



Stock configuration



Modified



Unfortunately, in the modified configuration it only has about 1/8" of clearance to the axle. I'll have to look at how to cut it back a little bit more to get more clearance.

[Shaun41178(2)]

I can't seem to remember but weren't you looking for exhaust donuts to cut?