Thanks Tirwin. Based on this PP tech article, the first mark is 5 degrees ahead of TDC and the 2nd mark is TDC itself. So I need to advance the engine by 5 degrees (actually 365 b/c I'm one half rotation off at the moment) and then check for alignment of the rotor and the mark on the distributor.

Assuming those are good, I'll see about grabbing a new ignition wire locally to test.

And you are?

Edited image text.

That's just a spammer. Post should be deleted soon I suspect.

Had a few minutes after work, wrapped the wire from distributor to coil in electrical tape. Seems to arc less.

I'll set it up for the TDC check later tonight if I get a few more minutes. I'm suspecting this is just a case of failing wires. Sadly, no one local sells just that one short wire. A full set is not cheap!

The alignment of reluctor and so forth was close, but not exact. The engine runs smoothly, and there is no slop or play in the distributor parts. With the 12 year old wires, I am thinking I can replace those first and see if that solves it.

I'm going to do a little research on if these shielded wires are good or bad - seems like Porsche was trying to solve a problem in the wrong way...

The "final" air box explosion/backfire fix was when Porsche added an idle injector fuel distribution manifold such that fuel due to flooding was distributed equally to the 6 intakes rather than pooling.

Have you yet disconnected/reconnected, SEATED, the CDI connector?

I'm wondering just what the CDI would do with/if the 12 volt supply/grounds input were intermittent, randomly fire with each intermittent power event?

Can you try supplying the CDI power from a alternate source, say the fuse block in the engine compartment. Even in parallel with the ignition source.

Ignition switch....SmileWavy

Very common source of failures.

I removed and re-seated the CDI cable with no effect. It seems to be nice and tight.

Got my new Coil to Distributor cable from PP today and hooked it up. I still get an arc between it and nearby grounds, same as my old picture showed. This is not a complete new wire set, just the 1 wire.

I did more watching in the dark, and there are two "modes" to the event. "Normally" it will have occasional arcs in one of the two places. It isn't very noticeable and you can't spot the effect on the tachometer.

Then it will have little sessions of 1-3 seconds with lots of arcs. If I use a wood stick to push the coil-distributor wire away from nearby grounds, I think there is a 3rd arc (in addition to the one between the coil's connectors) amongst the other wires. More interestingly, during these little sessions the whole distributor assembly vibrates noticeably. These sessions are also when the tach really bounces.

The little sessions are sometimes spaced more than 20 seconds apart, sometimes closer. Also, the engine seems to run smoothly throughout them.

The "wire" from the magnetic sensor in the distributor to the CDI "box" is a shielded coax and if not properly grounded on both ends will make the signal especially subject to external emi/rfi noise.

I think that one is green, right? I can take a look tonight, I recall it looking fairly old and tired.

Yes, the dreaded green wire.

If you search for green wire you will find plenty of info. The "hard" part with replacing the green wire that it is two wires -- inner and outer -- as wwest notes. Timmy2 just made me a new wiring harness and he installed a two pin female connector like the WUR and the AAR use. There is a place where you can buy a new green wire that uses this same connector. Makes future replacement extremely easy. I can post a picture later.

If this does turn out to be the green wire then it is interesting to note this as a failure case.

Also, I see that you just bought the coil to dizzy wire. If your plug wires are 10 years old you need to be thinking about a new set of plug wires anyway.

I'm planning to clean the contacts for the green wire tonight. Maybe order a full new set of wires. The fact that there is arcing through the new coil to dizzy wire bothers me.

I'll do some "green wire" searches.

The green wire is the magnetic pickup the CDI uses to know when to fire so if it is not functioning properly that might be causing CDI "misfires".

So I took out the distributor and checked the fun green wire from the CDI box. The wire has some issues, particularly around the connector. See the following pictures. Further down, the wire itself has a few areas with small scrapes and discolored areas, although nothing through the insulation.

http://forums.pelicanparts.com/uploa...1428017021.jpg
http://forums.pelicanparts.com/uploa...1428017041.jpg
http://forums.pelicanparts.com/uploa...1428017078.jpg
http://forums.pelicanparts.com/uploa...1428017109.jpg
http://forums.pelicanparts.com/uploa...1428017134.jpg

Is the distributor shaft drive gear TIGHT to the shaft? Roll pin?

The driver gear had no play at all. Once I removed the distributor entirely, it spun smoothly and freely.

I'm not sure what the roll pin is. The bolt for adjusting the rotation by small amounts was tight and in the correct place I marked 3 years ago when I last had it removed.

My wire is long and green! When I look up this wire online, I find pictures of short wires with a connector on the other end. On mine, the other end disappears into the wiring bundle with a bunch of other wires. Is mine abnormal?

http://forums.pelicanparts.com/uploa...1428027207.jpg

http://www.pelicanparts.com/catalog/...290701-M14.jpg

Yours is not abnormal. This is what I was talking about in an earlier post. You're going to need to cut in a new terminal to connect the replacement wire to that. The other end of that green wire is 2 of the 6 pins in the CDI connection.

Roll pin is that which holds the drive gear onto the distributor shaft.

As an exercise in determining the green wire's damage, I wrapped it in electrical tape and reinstalled the distributor. The spark plug wires are shielded by their braided covers in the area where they run near it, and that combined with the electrical tape and a bit of distance makes it seem likely that it is not receiving any interfering signals.

There was no change - there are still occasional spurts of arcing I can hear, and the distributor vibrates during these episodes.

However I have some more information that may be relevant. After the engine has been running and dropped down to normal operating RPM, a low 'whoooo' noise begins. It sounds sort of like someone blowing across the top of a beer bottle. If I remove the oil refill cap, the noise goes away immediately (and the engine begins to run choppy). I had previously assumed this was unrelated, but after noting the vacuum hose to the dizzy, I'm not so sure.

Is this the noise of a vacuum leak? Perhaps this is confusing the distributor, and thus the ignition system?

Distributor VIBRATES???

Start right there, it shouldn't!

End play in the distributor shaft?

Drive gear "torque" drives the shaft upward and the rotor tip begins contacting the rotor cap?

OsoMoore,

EMI is not directly the issue (EMI may not even be an issue). Let's break this down.

If we slowed down time and watched the moment the CDI fires the coil, the rotor will be at the optimum point in it's travel to create a complete path: coil->coil wire->dizzy cap->rotor->plug wire X->plug X.

This is the normal case. There is some resistance along this path. We'll call it resistance A.

The we have the abnormal case, where the current is arcing along some abnormal path. We'll call that path B and resistance B.

In your case, when you first discovered the problem the resistance along path B was lower than path A.

Normally path A should have much lower resistance and always "win" given any other available choice of path.

So the question there is why is your normal resistance path "losing" given that you have an abnormal path that includes air?

You can temporarily fix the problem through insulation and increasing the air gap by moving wires around but that is not a long term fix.

Address that first. It's time for new plug wires. You will have done some regular maintenance you're due for and eliminated a variable.