spark plugs and ignition timing

[budgerian]

The car is starting to annoy me now..

As posted in a previous thread, I changed out all the wear items on the ignition and reset the timing. initial timing at tdc, as per manual. dynamic timing at 22degrees btdc at 2500 revs again as per manual.
The car ran like crap before, like not firing on all six, right from the moment I got her home after buying it, hence the ignition tune up. After renewal of parts and resetting the timing the car still runs like crap, got worse after driving it a few miles. Pulling the plugs on cilinder 5 and 6 simultaniously doesn't change idle, so both are not playing along. Pulled the plugs and here I got lost: not one (new) spark plug looks the same (see picture)

6 is wet with discolouration on bridge going almost towards the threaded part (which would indicate too much timing, which is odd, as the car runs on LPG, so it should actualy be able to go with MORE advance) and the porcelain is brown. Half of the tiny ridge on the top of the porcelain is gone, indicating preignition or detonation.

5 looks textbook perfect, heat travel on bridge ends nicely on the 90 degree bend, porcelain is white.
4 looks identical.

3 looks goodish, but heat travel on bridge goes all the way, again indication too much advance.
there is an oil film below the threads, so outside the combustion chamber

2 is oily wet. 1/4 of porcelian is brown, rest is white. heat travel as 3, all the way down the bridge. again oily film on part below the threads. look closely and you'll see the film on the right. the film does not go circular however, just one side of the plug, same as plug no 3.

1 looks unused like out of the box, bridge is still silver with slight heat travel which ends before the bend of the bridge, porcelian is bright white. plug is completely dry.

Before all of this I did a compression test, which was good, as in excellent. This was the reason I deemed the engine usable and started working on getting it going.
Before, I also adjusted all the valves to .30. No problems there either, all valves could be set within spec.

So, while I can read plugs just fine, I can not explain, why I got almost six different readings, some of them not spelling good news. However compression was good and valves could be adjusted, basically ruling out leaky stems and worn rings, or a burned or sticking valve.
Cooling fluid looks good, though not tested yet, temperature readings were and are ok, but before pulling plugs engine felt extremely hot, taking into account I drove less than half a mile.

Car runs equally bad on gas and LPG, which sort of rules out mixture problems.
Timing issues would not explain six different plugs right?

I'm sure my explanation leaves something to be desired, so fire away any questions, that might lead to an answer that helps me solve my woes...


Roger

 

[Sven]

Have you checked that you are getting a consistent spark at each plug (ie, remove the wire from the plug and hold it against the head)? Those plugs do not look fouled enough not to fire. Have you swapped out the coil?

On my 1972 carb version I set the 22° timing running the engine at 1,700 rpm with vacuum lines disconnected from distributor.

 

[budgerian]

Sven,

That's because carb versions should be set at 1700.
3.0 injection models should be set at 2500.

All wires fire (they are new, but I did check anyway).

I just installed the old plugs again, reset the timing and took a test drive. After a mile or so, above 2000 rpm, the engine started firing on all six and behaved like it should, not so under 2000, it would however appear only one cilinder was not firing, as it was still running reasonably smooth. After about 3 miles the misfiring got so bad I suspect 2 cilinders stopped firing, so I drove back home.
I'll pull the plugs again, but expect to see 6 and 2 wet and fouling again.

One thing I noticed while setting the timing is that I can't get the timing mark to stop moving back and forth a little.
I first thought about the timing chain, but further investigation showed that there is minimal play in the distributor shaft.
How much play would throw the timing off?
Is it a known thing with Bosch distributors to wear the shaft enough to cause timing issues?
Would a worn distributor explain the issues I described before?

Questions, questions...

 

[Sven]

The timing mark may be moving around because of the rough running of the engine.

Have you ruled out that the injectors could be the problem? Have they been cleaned and tested recently?

 

[budgerian]

Sven,

The engine runs on LPG, meaning the injectors are not being used. The liquid petrol gas is mixed in with the air through a mixer in front of the throttle body. Besides, car runs identical when switched to gas, so mixture is not the problem.

I'm interested in what you all think about my distributor questions, as I think that's where the problem might lie.

If anybody has suggestions for replacement distributors other that the original, please advice!

 

[Sven]

Sorry - i cannot advise on the distributor tolerances, other than to say, that if the distributor was uniformly worn, the car might run rough, but it would not only be the same two cylinders that run poorly. Being cylinder specific would indicate a problem specific to the cylinder such as the wires to the plug, the valves not sealing, the head gasket leaking, etc (all items which you have ruled out).

There is a new distributor on ebay in Germany. A bit expensive, but then most items for the car are.

http://cgi.ebay.com/BMW-E9-3-0-csi-...ehör?hash=item4148a493b7&_trksid=p3286.c0.m14

 

[budgerian]

Sven,

Thanks for the link. That would have been a bargain, as W&N sells a new one for a whopping 1150 euro's plus change.
Unfortunately, as usual on ebay it is not a CSI/Si distributor, easily identifiable by the rotor.

The seller is obviously even unfit to copy the part number from the distributors body. To top it off, the last picture shows a completely different distributor!
Sellers like these should be banned from ebay for life.

 

[MMercury]

The car is starting to annoy me now..
Compression was good and valves could be adjusted, basically ruling out leaky stems and worn rings, or a burned or sticking valve.
Cooling fluid looks good, though not tested yet, temperature readings were and are ok, but before pulling plugs engine felt extremely hot, taking into account I drove less than half a mile.
Car runs equally bad on gas and LPG, which sort of rules out mixture problems.
Would a worn distributor explain the issues I described before?
Roger

Although your picture is helpful to the extent I can view it on a crummy monitor, and I am hard pressed to explain the various plug readings nor do I challenge your interpretations as 5,4,3, and 1 look pretty darn normal.

Given what I remember about your master cylinder adventure, is it possible that you have overlooked something very obvious? You may have renewed or tested the distributor wires, but are they all in the correct firing order? Is it also possible that they are bundled too close together to induce cross firing? Is the coil getting sufficient voltage or is the ballast resistor too restrictive? (You could try bridging the ballast just to see if there is any difference.)

It is true that a sloppy/worn distributor could result in misfiring, as could poorly adjusted or worn ignition points or a worn, loose, or poorly made rotor or cap and a whimpy coil. Why you suspect the distributor as the problem is unclear. I used to have a distributor machine that tested the distributor and permitted resetting or at least checking the advance mechanisms with the distributor out of the car with strobe lighting. See: http://www.dutra.org/pictures/misc/distributor-machine.jpg Because it took up too much space it was relegated to the North American equivalent of the Zuiderzee. This machine would confirm or resolve your suspicions and maybe an older shop near you might have one. An oscilloscope could be helpful too. Using a dwell meter to determine if you get a regular dwell reading might be helpful too. Still, you really do not need any or these tools nor do you need to be quoted a specific metric of lateral movement to detect distributor wear. You know it when you see it and/or feel it.

Also, if the main distributor shaft or its uppermost component is worn, you may likely see oil leaks and have a rotor that is eating into the cap terminals. Worn, missing, or broken pieces that are part of your distributor’s centrifugal advance is even more likely. Yet you could find out if this is functioning even close-to-normal by feeling the action with your hands or checking the movement of the advance with the use of a timing light.

As far as swapping distributors, I fail to see the need for a d-jet-specific distributor unless you are still using the d-jet injector system to trigger the injectors. If the LPG system is not triggered by the fuel injection points, most any six cylinder traditional-points-equipped distributor from an M30 engine should work in a pinch. (As long as the drive gear correctly meshes with the camshaft.)

This has been said many times before in this and other forums. The stock fuel injection system could just as easily account for poor running or misfiring.
The fact that you are running liquid propane is interesting. Perhaps you could describe the system in more detail. I think you have described a system that enables you to switch from or bypass the normal fuel injection system. Does this mean the normal petrol system is still pressurized when the engine is running on propane? I ask because if you were running on petrol, I might want to eliminate the possibility that the cold start valve is not leaking fuel unmetered fuel into the intake manifold and possibly flooding a few of the cylinders. Leaky fuel injectors could, to a lesser degree, have a similar effect. Even if an energized fuel pump does not pressurize the petrol side of things, an engine running on LPG might still create enough vacuum to draw additional fuel from these still connected sources.

Given your use of LPG, the following analogy may not be appropriate. I recently lent a hand to someone who wanted to borrow a starter for their e24. As it turns out, the engine would not rotate despite engaging the starter and going through two batteries. Over objections I pulled the plugs and found the engine flooded with fuel and suffering from hydro-lock. Removing the plugs permitted the engine to turn and sprayed the entire garage full of petrol. After the plugs were blown dry, they resembled yours. Starting the car required much cranking and it was only after the engine warmed up that the engine ran smoothly. Starting the engine cold that night resulted in the same problem. Pulling the electrical connection to the "new" cold start valve resulted in an improvement, but the major improvement occurred when the fuel supply to the valve was completely capped. If you are running the cold start system, just something for you to consider. Other issues (or non-issues if you are using pure propane) might be to check the pressure sending unit (next to the all too familiar brake booster) and the connections to it plus the wiring attached to the ECU under the rear seat.

I am admittedly unfamiliar with your LPG setup and exactly how you meter the fuel. I am aware however, that you can still overfeed an engine with LPG enough to affect mixture with symptoms such as yours.
Good luck!

 

[budgerian]

Mmercury,

Thanks once again for the very extensive answer. Once again it contains excellent food for thought. Very much appreciated!

Here's my thoughts on it:
Distributor wires are new ones from W&N, tested separately for spark and together for crossfiring. Even reinstalled the old ones: exact same problem, so I would rule them out as the source.

The reason I suspected the distributor is because of the wobbly timing mark, when setting the timing. I don't think the timing chain is causing it, the plug from the timing light on the no 1 spark plug does not interfere with another plug, which leaves the distributor. As I can actualy physicaly move the housing about a mm or so, I wondered if this is acceptable. I am not sure if a worn shaft would through the timing off equal for all cilinders or different per cilinder depending on where the rotor is when the shaft wobbles. I would indeed need acces to proffesional equipment to figure this one out.
Pulling an ignition at some pick/n/pull might give me the answer cheaper but:
The propane system works as follows:
Car starts on gas (thereby using the d-jetronic), and after settling on idle switches to propane (injected through a nozzle on the throttle housing, ignoring the d-jetronic.
So I will need the dedicated distributor I'm afraid, or choose to alter the system to work on propane. I'm reluctant to do that as now I keep the fuel system in good order and using propane only might put more wear on the engine because of the characteristics of propane.

Both vacuum and mechanical advance work as they should.

Your idea about the cold start injector is something i was thinking about as well (including possible vacuum leaks on the intake) but wouldn't that affect all cilinders equally?
There is no vacuum leak on the inlet port of the 6th cilinder.
I thought about the injector of cilinder no. 6 firing on after switching to propane, but that would not explain why it also runs crap on gas, both when starting and switching to gas for test purposes while driving.

I have a bad feeling that I have either a headgasket leak, causing the wet part, but I can't figure out why the compression test was good, or worn or broken oil scrapers on the pistons, causing not enough trouble to show up on a compression test, but enough to foul the plus after a few miles of driving. Either way, I'm in for a treat...
This would however still not explain the wobbly timing mark.

I also thought about the vacuum line to the brake servo, but the brakes are 120% now, with full servo assistance and I am not loosing any brake fluid.

To complicate matters further: The car starts on the button, undermining the above theories. It starts immidiatly, but runs on 5, take it out for a drive and after a while above 2000 it smooths out, firing on all 6. A few miles on, loss of two cilinders.

Talk about being puzzled..

ps Throwing something into the Zuiderzee equals throwing it on land, as we turned that one into our 13th provence some time ago :-D. But I got your point.

 

[MMercury]

Is a puzzlement?

The reason I suspected the distributor is because of the wobbly timing mark, when setting the timing. I don't think the timing chain is causing it, the plug from the timing light on the no 1 spark plug does not interfere with another plug, which leaves the distributor. As I can actualy physicaly move the housing about a mm or so, I wondered if this is acceptable.
Your idea about the cold start injector is something i was thinking about as well (including possible vacuum leaks on the intake) but wouldn't that affect all cilinders equally?
There is no vacuum leak on the inlet port of the 6th cilinder.
I thought about the injector of cilinder no. 6 firing on after switching to propane, but that would not explain why it also runs crap on gas, both when starting and switching to gas for test purposes while driving.

I have a bad feeling that I have either a headgasket leak, causing the wet part, but I can't figure out why the compression test was good, or worn or broken oil scrapers on the pistons, causing not enough trouble to show up on a compression test, but enough to foul the plus after a few miles of driving. Either way, I'm in for a treat...
This would however still not explain the wobbly timing mark.

To complicate matters further: The car starts on the button, undermining the above theories. It starts immidiatly, but runs on 5, take it out for a drive and after a while above 2000 it smooths out, firing on all 6. A few miles on, loss of two cilinders.

Talk about being puzzled..

If you have a leaky cylinder head, all bets are off. But what causes you to think this? Low compression? Oil/Exhaust in the coolant? Coolant/Steam out the tailpipe? Plug number 6 is wet, but with coolant or oel? If its wet fuel or fluffy-sooty deposits that does not indicate either a head gasket or cylinder head problem.

I am puzzled by your description of movement in the distributor body. Is this true even when it is clamped in place on/in the cylinder head? Doesn’t it leak oil when under pressure? It will be loose if it no longer has the rubber “o” ring oil seal intact. Is the drive gear at the bottom of the distributor securely attached to the shaft? Even though it is held by a roll pin, I suppose things could wear and come loose (I have never seen one though). Although I still have doubts that the distributor is the villain, I would expect the problems would be more likely to happen if the hold down clamps for the cap were loose or if the cap did not seat right on the distributor body.

If your distributor is still equipped with points, what is the condition of the plate upon which the points are attached? Tight? Smooth? Loose? Sloppy? If it is sloppy, it can affect the points and the points setting, in turn altering the timing. As you know, that plate rotates with the vacuum retard mechanism. Because of this, it has to have a certain freedom of movement. I have not looked specifically at the d-jet distrib innards lately, but I believe there is an adjustable cup retaining a ball bearing that keeps the plate snug.

When you describe some plugs as wet, do you mean: wet with fuel or wet with oel, or merely covered with velvety carbon/soot? What does the tail pipe look like? Is exhaust sooty grey or oily wet or ?? You say no. 6 cylinder rarely fires. It has to be air fuel spark or compression. Your earlier posts leave only fuel and spark. How about running the engine and taking the no. 6 lead attached to a plug grounded on the valve cover. Purpose - for you to visually determine that the plug is producing an attractive blue arc. If we are at code “yellow(arc)”, something in the ignition needs to be addressed.

As far as the cold start valve or any injector for that matter, with the ignition disconnected, you can pull them and observe them spray. You may find that injector 6 produces a dribble or even a gusher instead of conical spray.

The cold start valve dumps quite a bit of fuel in that manifold. Since its operation is usually for a less than a minute, a poorly functioning one is easy to overlook. When they go bad, as in leaks, nothing seems to run right. Even if the valve operates properly, it is controlled by a temperature sensor II (I think it is the lowest sensor closest to the oil filter). If the sensor is bad or has a bad connection, the valve can continue to over fuel the manifold after things warm up. I mention this because of the hydro-lock story. I have also seen someone run their engine directly off of the cold start valve - without the injectors functioning. The amount of raw fuel that came out of the cylinders was scary. You are naturally correct about the valve fueling all cylinders equally. Still, if the valve opening is deformed with deposits, it might deflect the spray pattern so that the fuel favors one or more cylinders. Once one plug fouls due to an overrich mixture coupled with a weak spark . . . you get the picture.