Erratic spark timing problem

[PotatoSalad]

Hi, I've been reading the posts here over the past week as i've been working on the latest repairs, but this will be my first post. Forgive me if my reply interval is long. I'm often away from the computer.

I have a 91 Beretta with the TBI 2.2L (VIN G). After having a problem with intermittent stalling and severe hesitation and "chugging" on the highway, I replaced plugs and wires as they were way past due anyway. Compression testing (and my nose) indicated a blown head gasket which disassembly confirmed. Head shaved, valves ground and all new gaskets later, I reassemble the top end and the problem is still there.

The intermittent behavior has now become constant and makes the car barely able to idle. Timing light indicates an erratic spark angle of about 20 deg, an oscilloscope triggered on the crank sensor and observing the ESC signal from the PCM to the ICM indicates the same erratic phase shift. The engine sounds like it's running with severely retarded spark timing, stumbling, missing and backfiring throughout the rpm range

So far i have done the following (recalled in random order):
Ignition: consistent spark presence verified with both inductive pickup and spark tester
Vacuum: varies with angle and degree to which misfiring/backfiring changes rpm, but as high as 20", but as low as 10-15" when it's stumbling bad. No discernible leaks (new manifold gaskets and hoses from head rebuild also)
PCV valve: cleaned during rebuild, works as intended at least on the bench
EGR valve: (this is a vacuum operated one) cleaned during rebuild, bench test ok
Test intake air temp sensor: matches spec curve at at least two temperature points
Test Coolant temp sensor: matches spec curve at at least two temperature points
Test/replace crank sensor: matches impedance and volt spec; even tried replacing with spare (no change)
Test MAP sensor: 4.77V at atmospheric, 1.6-1.8v drop at 10" vacuum (within spec); no spurious output
Test TPS: 0.5v at idle, 4.2v at WOT (floor mat is all curled up) also within spec
Test Oxygen sensor: warms up and starts crossing 450mv like expected; it's a bit biased to lean, but i figure if it's not burning completely due to late ignition or misfire, the excess oxygen would cause that. I haven't yet been able to wrench it out to test a replacement
ICM: 6 months old, replaced with known good unit anyway, no change
PCM: swapped with known good unit, no change (still had my original rom in it though!)
Harness: visual inspect and wiggle (dunno what else to try if signals show up when backprobed)
Timing chain: I figured maybe the tensioner broke but it didn't skip? (i'm grasping here) no discernible lash between manual crank movement and cam position (as observed from rockers and oil pump drive)
Lifters: I don't know how this would effect timing, but i'm just throwing this out there. During the rebuild, i ran em all through a hand press to check for leakdown and didn't observe any, so i doubt one is collapsed.

Things of possible relevance that i don't have a way to test:
PCM: I don't have any way to read actual process variables (no OBD1 tools or anything) fwiw, it registers no error codes. Without a connector face diagram (my manual is in the mail) i can't effectively backprobe signals at the PCM since there's a multiplicity of same-color wires in the harness.
Fuel pressure: I wouldn't think this could cause erratic behavior, but i can't test post regulator fuel pressure (i guess the regulator is integrated into the TB and there's no test valves)
Information: Until my helm manual comes in next week, my only reference is a haynes manual (garbage), the internet, and the helm book for my deceased oldsmobile of remotely similar vintage. The specs i've used to "pass" the sensors may be incorrect.

I probably forgot a few things as this is basically a summary of the last three weeks. Apologies for the long post, but i figure an excess of information is better than not enough. I'm kind of at a loss for determining what the problem could be. I wouldn't expect a wiring harness issue to disrupt the engine control loop in an erratic manner that wouldn't set error codes. It would either work or not work at all. My other current thought is that I am still using the same ROM as before. i really have no idea if that's a possible cause. It's probably something embarassingly simple, but I am out of ideas at this point. Anyone have any ideas?

[PotatoSalad]

Today the unburned fuel lit up in the catalytic converter while running it. The orange-hot cat was pretty obvious and i figured it may be plugged so i torched it off. It's maybe 30% blocked up and the vacuum seems to stay more around 20" now and there isn't a lot of unburned fuel in the exhaust any more. There's still a constant stumbling at idle and acceleration and occasional misfires at constant throttle. It basically runs the same, but with better vacuum.

I'm going to pull the rocker cover and try to make timing marks to figure out where the spark angle is and how much it's varying.

[PotatoSalad]

As if it matters, the issues were:

1: blown head gasket with prior long-term coolant burning (had precipitated other problems)
2: plugged cat and increasing backpressure problems causing temp-dependent vacuum and intake backfiring problems
3: slow and intermittently spurious oxygen sensor (found as soon as i decided to use a scope instead of a stupid DVM)
4: one of the ECM's was never throwing codes and seemed to ignore the bad oxygen sensor entirely. (new one immediately dumped 44/45 errors)
5: two of the new spark plugs had internal cracks

the last one was the most frustrating because they pass any test outside of the engine. only under certain mechanical shock do they shunt internally. this produced a misfire/weak spark condition that varied with temperature and tended to appear or grow worse under certain rpm/pressure conditions. when i made the first post, the effects of the cat and the questionable ECM behavior were causing just about the perfect storm for absurdly erratic engine control stability. i sliced the top crimp off the threaded plug body using the lathe and pulled the porcelain core to find a completely fractured ceramic/resistor element inside.

now it's time to fix the exhaust. compared to my delta 88, the beretta doesn't sound as good with a straight pipe in place of the cat.

[Money pit Beretta]

Most info ever and no one wants to help?

[3X00-Modified]

Well it sounds like he's found a lot of his issues so far and is still trying to trouble shoot a possibly erratic spark... Until we have verification he is still getting an odd spark pattern then there is no reason to blow up the post yet... But I was gone all weekend and the activity of help on this board seems to drop lower and lower each day.

[Money pit Beretta]

It's not just here, it's like that on every site I'm on.

[PotatoSalad]

That's okay.
It's running for the moment, but i haven't put too much time on the road to be sure it's okay.
Running with intermittent headlights and no exhaust but a downpipe is a bit risky.

Having the helm book makes me feel better for future repair prospects though.
I'm not sure if I want to invest in a ALDL link kit given the expected remaining life of the car.
For what it's worth though, does anyone have a recommendation for a practical data interface tool/kit/software that would be compatible with these PCMs?
I was a bit confused about the USB ALDL link cables saying they're compatible with all 160 or 8192 baud PCMS, but a cursory survey of WinALDL indicates it's only intended for 160 baud systems.

[themixer]

that's really annoying, you sound like your chasing gremlins. I have had a miss in my car for 5 years now that i have been working on. I have had not much help on forums diagnosing the issue

Doing a compression test is worth it, just to see if anything is obviously mechanical causing a misfire
since you are tossing parts at it, a CPS wouldn't be a bad idea. They are a bitch to replace on 2.2s but they are cheap and it does control some aspect of the spark.

[PotatoSalad]

Doing a compression test is worth it, just to see if anything is obviously mechanical causing a misfire
since you are tossing parts at it, a CPS wouldn't be a bad idea. They are a bitch to replace on 2.2s but they are cheap and it does control some aspect of the spark.

The compression test is okay now that the head has been rebuilt.
The crank pos sensor is known good (i have two and they're easy to test with a scope and DVM)
it's just a variable-reluctance PMG

At the moment, the flip-flopping timing symptom seems to have gone away.
Like i said, it seems to have been a product of the PCM trying to stabilize engine control with a speed/temp/torque dependent misfire, a spurious oxygen sensor, and restricted exhaust all being coincident. It's hard to tell wtf is going on when you're blind to the PCM's state or process variables.
The only remaining issue has been difficulty in getting the new oxygen sensor to keep grounded. It seems to go open circuit occasionally when it's hot, but today i took it out and ground & doped the manifold seat. The possibility of any other intermittent connections on that ckt should be easy to diagnose now that i have the helm book.

Also:
fwiw, the CPS is easy to get to on my 2.2L if you've memorized where everything is on the back of the block.
when i rebuilt the head i had to make new fuel line stubs for the throttle body, so i opted to route new flex lines behind the vacuum booster instead of out in the way like the original QC nylon jumpers are arranged.
this makes it a lot easier to reach everything on the backside from on top (ICM/coils, CPS, oil pressure switch)
of course, bending those throttle body stubs to lock into the intake manifold support clamp was fun...

[RobertISaar]

I was a bit confused about the USB ALDL link cables saying they're compatible with all 160 or 8192 baud PCMS, but a cursory survey of WinALDL indicates it's only intended for 160 baud systems.

that's because WinALDL is only intended for 160 baud units.... tunerpro will do either.

[Hawk2010]

You get this problem fixed? I got the same problem on a 2.2 1996 s10 I'm working on for a buddy. I think it is a fuel volume issue. I'm pulling bed tonight to pull pump and look at screen. Good luck

[Hawk2010]

You get this problem fixed? I got the same problem on a 2.2 1996 s10 I'm working on for a buddy. I think it is a fuel volume issue. I'm pulling bed tonight to pull pump and look at screen. Good luck