Dead cylinder for the first mile

[morsing]

I'd try swapping the powerstage. For a short test, the swapped powerstage doesn't need to be bolted down.

Note: the powerstage is 2 parts, a black electronics module which is bolted onto an aluminum heatsink. Don't separate them - unbolt the heatsink from the fender instead. If you separate the 2 parts, you need to add a layer of white thermal compound between them on reassembly.

Swapped, didn't make a difference.

Does anyone know where it is grounded?

Regards,
Henrik Morsing

 

[morsing]

Turning the distributor will be a bit fiddly, but I might have to try it.

Blowing gas as the injector area didn't help, forgot to try the inlet manifold itself.

It's annoying my ignition system doesn't fit anything in Haynes/Bentley. I was to test the Hall sensor resistance, but which terminals am I meant to check?

I am going to splash out on a new coil. The one I have is a cheap, un-branded one. I measured the resistances and found:

Primary: 1.2
Secondary: 8.3K

This combination does not fit any system in the Haynes manual, but I suspect the Primary is too high.

Regards,
Henrik Morsing

 

[morsing]

LOL! At least I fixed one thing taking it all apart over the weekend!

Sligthly panicked as it was sucked in!

Regards,
Henrik Morsing

 

[tbrcktch]

LOL! At least I fixed one thing taking it all apart over the weekend!

Sligthly panicked as it was sucked in!

Regards,
Henrik Mo

LOL! At least I fixed one thing taking it all apart over the weekend!

Sligthly panicked as it was sucked in!

Regards,
Henrik Morsing

Hi Henrik,
Progress! Congrats, sometimes it's just the little things. You're getting there! Probably never thought you'd be feeling good about having an engine that sucks!

 

[tbrcktch]

Turning the distributor will be a bit fiddly, but I might have to try it.

Blowing gas as the injector area didn't help, forgot to try the inlet manifold itself.

It's annoying my ignition system doesn't fit anything in Haynes/Bentley. I was to test the Hall sensor resistance, but which terminals am I meant to check?

I am going to splash out on a new coil. The one I have is a cheap, un-branded one. I measured the resistances and found:

Primary: 1.2
Secondary: 8.3K

This combination does not fit any system in the Haynes manual, but I suspect the Primary is too high.

Regards,
Henrik Morsing

Hi Henrik,
I think rotating the distributor, and switching the plug wires should be very revealing. Very curious as to the result. Did you get a chance to video your exhaust again? Hang in there, you're gonna get this!

 

[bobxyz]

%**@! I was really hoping that swapping the powerstage would improve the problem.

The distributor pickup looks to the the same as the LH2.2 pickup. You can test it with key on and a voltmeter (doesn't need to be running). Back probing across the outer pins of the 3-pin disti connector (labelled - O + on the outside side of the disti connector) should show close to full battery voltage. If it's significantly lower, probe from the - disti pin to the - battery post (should be ~0), and from the + disti pin to the + battery post (should be ~0).

You could also test the disti output (center pin) at idle. It should be close to a square wave, with an average voltage of [guessing a bit here] 4 to 7 volts.

I found the same TZ-28H diagram that @Apollo kid posted in the '85 wiring greenbook TP30808. The "B" ground connection is on the end of the headlight hi/lo and power distribution bracket:


The distributor drive gear has 12 teeth. If you mark the starting alignment (to get back to original timing), you can pull the disti out. It will rotate ~1 tooth as it pulls out. You then need to rotate it 3 more teeth and re-install to shift the spark to the next tower. After re-installing, the rotor tip should have moved by 90degrees.

One more question - do you have a basic rotor or one of the fancy rev-limiter rotors? If it's a rev-limiter, can you try a basic rotor instead? Maybe the rev-limiter is partially stuck and causing a misfire? It's unlikely, but nothing else has helped.

 

[apollokid]

%**@! I was really hoping that swapping the powerstage would improve the problem.

Why? One of the main rule in autodiagnostic for finding quickly the root cause of the problems says that we have to to distinguish between failures that occour to a single cylinder and those who occour to any cylinder so that it is easier to identify the suspected elements. A failed ignition coil or a powerstage in an ignition system like our it is very very difficult that it could cause a missfire on a single cylinder since they are in common to all cylinders and not to a specific one.
The only reason that I can image for a missfire on a single cylinder is that the spark is too weak (to all cylinders, not only to a single one) and in that cylinder there are other causes of problems that sum togheter (eg. bad fuel mixture or a mechanical problem, but the previous tests made by Henrik of switching the injectors with the hoses attached and the refurbish of the head in addition to the good compression measures taken before seem to exclude problems fuel or mechanical problems so it remain only ignition)
This in my opinion based on what I read.

 

[morsing]

[...] A failed ignition coil or a powerstage in an ignition system like our it is very very difficult that it could cause a missfire on a single cylinder since they are in common to all cylinders and not to a specific one.
The only reason that I can image for a missfire on a single cylinder is that the spark is too weak (to all cylinders, not only to a single one) and in that cylinder there are other causes of problems that sum togheter (eg. bad fuel mixture or a mechanical problem, but the previous tests made by Henrik of switching the injectors with the hoses attached and the refurbish of the head in addition to the good compression measures taken before seem to exclude problems fuel or mechanical problems so it remain only ignition)
This in my opinion based on what I read.

This is now exactly my thoughts. I wanted to do a video of no-spark on cylinder two yesterday but trying it, I what jumping all over the place with cylinder #4 being the worst. In-fact. my thermo gun showed the exhaust runner on #4 to be pretty cold, with the other three being about the same. Sorry I didn't do an actual video, it complicates things being in the middle of things.

I think my coil is either wrong for the car, cheap crap, or faulty. But my spark strength tester should be arriving today, so I will know for sure.

As you says, some slight additional condition, head-gasket related maybe, caused #2 to be worse off, but it has now evened out.

I ditched the rev-limited rotor 19 years ago, hateful things.

I tested power to the Hall sensor, it was 10.8V which is about right when the ignition is on and lights drawing power.

I also spent most of the weekend working on the Saab, it would be so helpful having a spare car.

Regards,
Henrik Morsing

 

[morsing]

Hi Henrik,
Progress! Congrats, sometimes it's just the little things. You're getting there! Probably never thought you'd be feeling good about having an engine that sucks!

It's an amazing difference, the glove was like cow udders before, I wish I'd filmed it.

Regards,
Henrik Morsing

 

[bobxyz]

I was hoping that it was the powerstage because, in the TZ-28H, the powerstage has active dwell control and automatic over-dwell shutoff. If these weren't working correctly, it might have shown up as a periodic no spark. It's grasping at straws, but powerstage was one of the few parts that hadn't been swapped.

[details: a) the active dwell control adjusts the dwell to reach a fixed peak charging current. Internally, it delays from the falling edge of the hall sensor signal to start charging. If it hits peak current before the trigger (the hall rising edge) it delays the next spark's dwell a little more. If it triggers before peak current, it delays the next spark's dwell a little less. b) the over-dwell protection it to prevent cooking the coil when the ignition is turned on but the engine is not started. If the hall sensor is low (normal charging time), the powerstage has some sort of internal timer that detects too long of a dwell period and stops charging.]

Henrik - I pulled out a LH2.2 disti from my parts pile and compared it to one of your pictures. They look the mostly the same:

Yours vs. mine:

+



Two things to notice - it looks like you have a slightly bent vane on the right and your wires aren't tucked under the plastic cover properly. I don't think the bent vane would cause problems if it's not hitting anything. The uncovered wires are OK if they're not getting hit (cut into) by the rotating vanes.

[details: the hall sensor has a fixed magnet inside of the rotating vane and the sensor on the outside - when the vane blocks the magnet-to-sensor path, the output signal is high. When the gap un-blocks the magnet, the output goes low. I'd think that a slightly bent vane would still block the magnetic path.]

 

[apollokid]

@Bob: even if the power stage logics could cause a periodic missfire every 4 sparks it must be very difficult that the the sync is always on the same cylinder. Every time you stop the engine it can stop in any position of the camshaft and the probability that next time it restarts the missfire continues to be on the same cylinder is only 25%.
Too low in my opinion.

 

[morsing]

[details: the hall sensor has a fixed magnet inside of the rotating vane and the sensor on the outside - when the vane blocks the magnet-to-sensor path, the output signal is high. When the gap un-blocks the magnet, the output goes low. I'd think that a slightly bent vane would still block the magnetic path.]

Thanks, I've never managed to find anyone who could explain how it worked.

I assume if the sensor works, it works. It can't have a weak signal?

P.S. So in theory, I could cause a spark moving a piece of metal through the sensor?

Regards,
Henrik Morsing

 

[bobxyz]

I meant to add these to the last post but got distracted:

The magnet and iron wrapper are on the left, the sensor is within the plastic tower on the right. I wouldn't expect a weak signal from this type of sensor, and I don't think a slightly bent vane would cause problems. Next time you have it open, check that the wires are still OK.

 

[ZVOLV]

I am confused. Which cylinder is misfiring now?

I don't see how the primary ignition system could cause cold start misfires that clear up.

 

[apollokid]

Hall sensor does't need that the rotating element reaches a minimum speed so it could be possibile to check the signal with a multimeter and see the voltage changing from 0V to 5V (or 12V? I don't know the voltage output level) according to the position of the metal window in front of the sensor (like if it was a contacless microswitch)
It must only be powered the sensor by turning the ignition key on run postion and rotated manually the rotor of the distributor.
If you want to do this disconnect primary of the ignition coil to avoid the shock hazard.
In this way you could be able to check the signal without a scope, but since the problem occours only for a few minutes when the engine is cold I think that this test doesn't be really useful compared with a check of the squarewave with a scope while the engine is running
You could try it but is not It's not what I would use.

 

[apollokid]

I don't see how the primary ignition system could cause cold start misfires that clear up

ignition coils defectes could be appear and disapper with temperature, but coil is an element in common with all cylinders so it if was a problem iike this the missfires must be randomly to any cylinder and not always the same.

 

[ZVOLV]

It would be fun to just spin the distributor by hand or a drill and check spark at all 4 plugs. Sparking in the air is gonna be different than in a cylinder with compression and fuel.

I also had luck with manually making a coil spark by tapping the ground side with a jumper wire to a good ground with the correct rythmn.

Personally, I haven't ever seen cold start misfires on my redblocks and I can't really follow what's going on with the thread, but keep up the work.

I would play around with cylinder drop tests by disconnecting fuel injectors during a misfire event and see if you can narrow down the issue. I don't really trust timing lights. I have one somewhere that I used to use, but my preferred method is spare plug or even better, a good spark tester, they are easier to clip/connect to ground.

 

[morsing]

Hi all,

Drove to work for the first time since the head gasket change and valve grinding. Was way better to before it went completely wrong, but still not good. Idles on three cylinders and intermittent in traffic but once you're on the motorway it's ok. It's also no longer just at the beginning, but occurs throughout the entire journey.

I am dropping it off at a body shop today so no further investigations till the weekend.

It's annoying my spark tester didn't arrive, but I will have it by the weekend, so should give more solid proof if the spark is too weak.

Spent a lot of the weekend working on the Saab, hoping to get it ready soon so I won't have to rely on the 240 for work. It is incredibly stressful having to do the early morning 32 mile drive.

P.S. I think the failing cylinder is a a bit random now.

Regards,
Henrik Morsing

 

[morsing]

I finally, finally, finally received the spark tester! And I did some videos on my phone, which didn't record, and some on my camera, which I haven't looked at yet.

Conclusion is:

I think the coil is weak/wrong and I have ordered a proper Bosch one now. Just shows you you should never skimp on parts.

According to the instructions for the tester, a standard coil should do 1-1.5cm with a clear, blue, straight spark.

At 1.2cm mine is erratic. Occasionally misses and isn't very straight or blue.
Then tried it directly on the cylinder leads. Not sure what it should be for that, but at 1.2cm only about 1/3 sparked and it was much warmer in colour and jumping around a lot.

I have received my smoke leak tester, leak-down tester and compressor, but I am missing connectors for the compressor. Very confusing to figure out what was needed.

When I get the video from the camera, I will post it and I will post an update once I have the new coil.

P.S. Just wanted to add that it hasn't all been completely in vain. Along the way I have found and fixed:
- Leaky cold start injector
- Blocked flame trap
- Leaky valves (Manifold vacuum is now significantly better at idle)
- Head gasket problem

Regards,
Henrik Morsing

 

[morsing]

Turns out you can't get the proper coil for this car anymore. This won't be easy to fix.