Spark energy leak at head?

If this "leak" is in the form of arcing between the plug and head, I'd say your spark plug wires are bad. Replace them with high quality, not the cheapest in the store. If there is some kind of glow or something, I'd have to ask "how is it running and what problem(s) are you experiencing?". If it stumbles under load, that is a symptom of bad plug wires. To verify, you may try wrapping a couple of wraps of electrical tape around the boot that goes over the wire. If the problem goes away, you definitely found the problem, but it may leak through the tape. This is only temporary as the adhesive will not hold up to the heat very well....

If you take a spray mist bottle of water and mist the wires at night, you will likely see that the spark is leaking out of the rubber boot. They get dry and hard and can leak out.

'Sometimes' a good coating with dielectric grease on the boots will stop the leak for a while, but once leaking they really should be replaced.

Mike

LanceM wrote:

Mike Roma> If you take a spray mist bottle of water and mist the wires at night,

Have misted and used ground probe. The ignition wires are good (a couple of new sets trialed). As best I'm able to determine, the (unusual/rare?) leak/glow is between the spark plug bases and head, with stronger intensity correlating to RPM below. I hope to research a safe method of doing a (Kilo) voltage drop, using attenuated probes.

Driveability problems, in brief, are mild to severe bucking around 2800 RPM. Initially only under light load; has progressed into steady-state throttle (still worst @ 2800 RPM), and jerky acceleration w/mild to moderate throttle. WOT is fairly decent. Lots of other testing performed (car has been to a couple of master Saab techs, w/o luck) - am focusing on the leak/glow for now (to see if correlation extends to cause & effect to some degree)

Looks like I messed up the response line

If this clarifies - the leak/glow is coming from directly beneath the spark plugs - right where they contact the head. I'm not seeing arcing from the (higher up) boots or porcelain to the head.

With 0.2 Ohms between plugs outer case and the head that should not be possible. What color is it? Can you shoot a pix and post it somewhere? Are you certain that you do not have some kind of weird insolation occurring at the threads. Is the head aluminum and has it been chemically cleaned? I am thinking of some kind of hard anodizing insulation despite the fact that you measure 0.2 Ohms which is nothing for the high voltage. Measure the resistance between the head and the negative battery terminal. Should be close to zero.

Regards,

Boris Mohar

Got Knock? - see: Viatrack Printed Circuit Designs

I have seen it run down the porcelain to the spark plug base and jump from there to the block. It is not an unusual arc to see with old wires.

The water mist usually shows it up really clear one way or the other.

I can't see how the internal spark can get past the threads to arc on the outside.

Mike

I have old-fashioned (?!) emulsion 35mm cam gear, and marginal scanner, but not a digital camera. Think it would take substantial amt of exposure bracketing to simulate what I'm seeing. But may still give it a go

The color of the leak/glow is healthy blue. The glow is/are rings that completely encircle each of the spark plug bases.

I sampled the continuity of the head/spark plug thread holes and the spark plug to head mating surface, to the negative battery terminal, off another car, and multiple probe locations almost always show virtually no resistance (as it should, afaik)

On the Saab, however, multiple probe locations _do_ show high resistance or an open circuit. However, as stated, the spark plug hexes show no virtually no resistance to ground. I thought the latter might be more definitive...

There is no resistance between other parts of the head and the negative battery terminal (0.0 ohms)

So trying to make sense of the paradoxical data, and whether the former is just probe contact anomalies (but suspicious, given the circumstances)

Thought most definitive measure might be a v-drop of peak firing KV - if that is possible w/a scope at a fast time base (w/spark KV only lasting nanoseconds) - too much better quantify my "visual" assessment of a "mild" intensity leak into some numbers, and to confirm that what I'm seeing is legit. Trying to think thru how to do that

Altho 2dary scope ignition pattern is uniform and appears normal across all 4 cylinders (another paradox to me, if this is an ignition misfire due to shunted spark energy), the area where the spark plug gasket makes contact w/the head on #4 cylinder - the worst "glower" - _is_ noticeably discolored to a goldish hue. Is this possibly hard anodizing? If so, it would only be microns "deep"??

The head is alu, on an iron block. I was thinking I'd best pull the head to chase the threads (or chemically clean?), to prevent whatever crud from falling into the cylinders, and to have the spark plug gasket/head contact areas cut/skimmed/machined. If I can do something w/o pulling the head...?

I've sorted thru a couple of sets of new ignition wires - picking the best individual wires using misting & ground probe (a couple of sets of new plugs, too)

As would be expected, misting increases the intensity of the "glow rings" that completely encircle each of the spark plug bases.

The plugs are semi-recessed - depth perception is difficult - but the leak does appear to be coming from underneath the plug bases. I will recheck again, tho, in good spirit

I wonder if the anti-sieze compound is not correct and is an insulator?

Is there a chance the plugs have the thin threaded contact on top and the plug wires have the cap with a hole for a screw on stud or post spark plug?

Or the other way around, the cap has the two wires in it vs the round hole on the plugs?

I think something is off big time.

It is even acceptable to use anti seize on the plug threads and still have a good ground, so no way can the threads not make a decent ground.

The ground they do make is still a contact ground at the end of the spark point. If electricity can find a shorter route, it will sure take it or at least send some power that way.

Mike

Here's an idea to try,,, how aboutmaking a ground xystem that would connect the spark plug bases to ground directly. Take some heavier copper wire, twist a loop around each base then connect the wire to round. You could also use a copper ground strap with holes in it.

George Kowal

My ISP's news server is down so this is from Google

I should first better detail some previous info: of about 15 probe locations on each Saab cylinder - mostly inside the head thread holes, but some at the spark plug gasket/head contact area - 10 had high resistance or an open circuit to negative batt terminal; 5 had less than 0.5 ohms [in contrast to my other "sample car/head", where all 15 or so probe locations showed virtually no resistance to the negative batt terminal]. But all spark plug hexes show virtually no resistance to the negative batt terminal

The spark plug top studs are fixed, match the plug wires terminations (IIRC SAE 4mm?), and are secure. I have the thin threaded type on another car, and actually prefer them, as seem to make more positive contact

Several installations & removals and a couple of new spark sets ago, I did sparingly use one application of Loctite/Permatex antiseize, which should be fine. But I did clean off everything as best as possible a few days ago (just in case), w/o change to driveability or the glow

I'm going to try to get one of the below, if possible and not too big $$

"Measurement" spark plug, that has a test lead (would think it has continuity to the center electrode, so should be able to perform a scope V-drop w/attentuated probe)

Would think all of the other spark plug companies would make them, too

Chase the plug holes with a tap. It is an obvious thread to ground problem. I don't know where your ignition is located but if chasing the threads doesn't work, I would try on a hunch to use a jumper cable to ground the head to the body and see if there is a change. Take an ohm reading from the body ground to the batt neg and see what you get. Strange indeed.

By "spark sets" do you mean wires and plugs or wires only? If they are the same plugs, are the ceramic insulators cracked? Or some other problem with the plug itself?