Strange spark plug problems

peter-h

Line Up and Wait
Joined
Sep 24, 2007
Messages
613
Location
UK
Display Name

Display name:
peter-h
This is a bit of a mystery which has popped up on a few US forums... I hope I am not missing a recent thread here.

Cirrus (turbo) owners have been reporting a high failure rate (due to cracked insulators) of fine wire plugs, especially. One example is here

Somehow (I cannot find the reference) people have decided the resistance should be about 3k. Support for that may be e.g. this

The traditional manufacturer of these plugs, Champion, does not publish a value.

A new mfg, Tempest, does, and is making the best of Champion's apparent misfortune.

My fine wire plugs (RHM38S) are shorter thread than the Cirrus ones and I have not had any insulator failures, but I have just measured about a dozen of some 700hr ones and all but 1 read open circuit (>20Mohms) with a normal DVM, with 1 reading a few hundred k. These plugs were installed in 2003 and removed in 2008. The erosion status of them all is as-new which is typical for fine wire plugs of that life; they wear extremely slowly.

I also have one conventional heavy electrode plug which reads a few k.

But I've just re-measured all the fine wire plugs with a 1000V insulation tester and except one (which remains open) they read completely differently at the higher voltage. None were over 1M, and this was true at 250V, 500V and 1000V, though the three voltages produced slightly different resistance values.

So clearly the "resistor" is not a resistor as we know them. Whether this is intentional, who knows? I suspect it's a cockup. Maybe there is a tiny air gap at the ends, of the order of 0.01mm, which flashes over easily - the expected result of repeated expansion and contraction of the plug assembly if there is no spring pushing against the resistor. But this is clearly not intended.

It could explain all kinds of weird engine behaviour in the field, because all these plugs test OK in a spark plug tester.

The plugs I tested were c. 10 years old. So unless there is some good explanation for this, it appears that Champion were making rubbish spark plugs for perhaps a decade or more, and looking at the US data they still are, getting away with it only because a magneto puts out enough kilovolts to flash over the resistor, or flash over any air gap at the ends of it. And fine wire plugs are ~ $100 each so this is outrageous.

I should add that the engine worked apparently fine, even with the plug which was totally open circuit at 1000V. No high altitude issues either.

Many years ago I used to work in high voltage (up to 100kV) and one quickly learns that any air gap leads to a rapid degradation of the component. Applied across a stack of differing nonconductive materials, voltage distributes itself inversely proportionally to the relative permittivity of the material and any air pockets thus end up with a very high potential gradient across them, causing them to flash over and gradually carbonise. Having a "loose" resistor inside a spark plug is a stupid idea.

I need to find some unused Champion fine wire plugs plugs and test those... any volunteers?

My feeling is that the reported cracked insulators in turbo SR22s are a separate issue, perhaps connected with the longer reach of these plugs in combination with a high EGT. Regarding the resistors, I fly peak-EGT all the time (non-turbo) and I found all of my plugs to have a massively elevated resistance too.
 
I hear of broken plugs and first suspect improper operation of the engine,
 
How come I have nearly a dozen RHM38Ss which are a few megs, but not damaged?

The two are separate issues, I am sure.

I think the cracked insulators on the Cirruses merely drew attention to the resistor issue.

I have just opened up two of the plugs: the one which was open circuit even at 1000V RMS, and the one which was ~300k.

Both had the resistors in a reasonable condition externally, both were adequately spring loaded, and testing the individual resistors yields the same resistance values as before. So the resistance-shift issue is with the resistor alone.
 
I hear of broken plugs and first suspect improper operation of the engine,

That is Champions explanation for the cracked insulator issue.

There is a little dispute between two vendors going on here:

- TAT has issued a MSB that requires all Champion fine-wires to be removed from TAT equipped planes.

- Champion in return has issued a letter stating that they have seen no unusual warranty returns and that the problem must be related to pilots 'not operating their engines in accordance with manufacturers recommendations'.

To some extent this looks like a pre-emptive blame game for the day that someone with a TAT equipped Cirrus full of investment bankers hits the dirt due a spark-plug failure. Both companies want to be on the record with a 'told you so' to improve their position in the ensuing litigation.
 
I guess the question is, what's special about the 22 turbo/ champion fine wire combo.
 
I reckon 3 things:

1) Peak EGT at around 85% power

2) That EGT is maintained for a long time on any given flight (due to turbo)

3) The plugs have a longer reach, so a longer insulator (?)

Plus the inevitable occassional clumsy engine management which is bound to be seen in a significant population :)

The resistors are interesting. The body is totally an insulator. The terminals are the ends only, which have some sort of "silver" plating on them. So the resistor itself is actually inside that component, which would be the correct way to do it (to prevent flashover down the outside at high altitudes, etc).

But something is causing the resistor to go rotten, over time. The fact that at low voltage it reads open circuit but at 250V+ it reads "something" suggests that the particles of the material become physically separated, by a microscopic amount (250V RMS won't jump very far; air at sea level will hold of the order of 3kV per mm, depending on the electrode geometry).

Evidently, Champion have just got totally lucky with this rubbish design, due to the typical magneto delivering some 20kV+ and just flashing over the whole thing :)

It's amazing what you can get TSOd if you know which paperwork to push :)
 
Last edited:
I guess the question is, what's special about the 22 turbo/ champion fine wire combo.

That they tend to pull high power settings (80+%) for prolonged periods of time based on the knowledge that they can't possibly hurt their engines the way they do it.
 
It's not uncommon to pull 1 or 2 broken fine wire plugs out of the TSIO 520/ 550 engines or the TIO 540 engines used on the PA46 at annual. I pulled one out recently during a pre-buy that only had 6 hrs since the annual. Many times over half of the ceramic insulator will be gone (I wonder where it is). Most times there are no operation issues. It doesn't seem to make any difference if it's operated ROP or LOP.

The Tempest fine wire plug has a much larger ceramic insulator than the Champion. I don't think I have found a cracked Tempest plug. The Champion has a very long narrow ceramic insulator nose. A number of years ago Champion told me that it was detonation. Later told me that it was because abrasive blasting media was getting stuck between the center conductor and the insulator.

The RHB36S part is the primary plug choice for the Mirage. This is not a problem that I saw it's first years. Something has changed.
 
there is nothing special about the Cirrus combination. In fact, this problem is found in all engines, turbo or not, that have Champion spark plugs. Just today I found three cracked insulators on a normally aspirated SR22. I have some from an SR20, a Piper Seminole, and a Cherokee Six. If you want to see plugs without cracking, but with phenomenally high internal resistance, which causes harness spring burning and magneto flash-over, I have them too. They are also made by Champion, and come in fine wire and massive electrode types.

http://www.flyplatinum.com/blog/?p=768
 
Anyone know what the resistor really is in a champion plug?

Auburn too for that matter.
 
I have put a pic here

The two resistors above are from a) the plug which was totally O/C at 1kV and b) from the plug which was the lowest resistance. One of them shows a little bit of burning at the ends.

BTW the resistor is described as "silicon carbide" here, page 3
http://www.championaerospace.com/assets/AV6-R-Nov2004.pdf

A google on silicon carbide resistors reveals these are not normal resistors, e.g. here


Resistors which pass a current proportional to the fourth or fifth power of the applied voltage are now widely used in many fields of electrical engineering. The first part of the paper discusses the behaviour of these resistors in terms of the characteristics of single contacts between silicon-carbide crystals. In the second part, the construction and properties of the resistors are described, and some typical examples are given of their uses for protection on transmission lines, in d.c. inductive circuits, in radio, for spark quenching at relay contacts, for voltage regulation, for field control of electrical machines and for various other purposes.


[my bold]

It's no wonder that these don't measure as anything useful at low voltages, but one would expect more consistency...

However it is quite possible that these resistors do basically nothing useful at sub-kilovolt voltages, so measuring them as described may well be a complete waste of time.

And the turbo Cirrus failures have to be a separate issue anyway; there is no way the cracked insulators can be anything to do with the resistors.
 
Last edited:
It's no wonder that these don't measure as anything useful at low voltages, but one would expect more consistency...

However it is quite possible that these resistors do basically nothing useful at sub-kilovolt voltages, so measuring them as described may well be a complete waste of time.

This is what some of them do at 25kV :

http://youtu.be/DERIHonZT-k
 
That's not a good test because a stun gun discharges a capacitor into the "subject" without a current limit, so if e.g. this spark plug "resistor" has a resistance of say 100k then the stuff will just flash right across, because to maintain 25kV across 100k you need a current of 250mA which one can easily get from a high voltage capacitor.

The output impedance of a magneto is nothing like as low as that; I don't suppose a mag can do more than (of the order of) 1mA.

Also, if this kind of arcing was really taking place inside the spark plug, it would be extremely obvious from the condition of the internal parts.
 
That's not a good test because a stun gun discharges a capacitor into the "subject" without a current limit, so if e.g. this spark plug "resistor" has a resistance of say 100k then the stuff will just flash right across, because to maintain 25kV across 100k you need a current of 250mA which one can easily get from a high voltage capacitor.

The output impedance of a magneto is nothing like as low as that; I don't suppose a mag can do more than (of the order of) 1mA.

Also, if this kind of arcing was really taking place inside the spark plug, it would be extremely obvious from the condition of the internal parts.

This element read 'open' on a DVM.

If you look up the 'TN Bonanzas with fine wire RHB32S spark plugs - MSB' thread on beechtalk, you will find the post Max made on 4/4/12 detailing what he did for testing on the Champion fine-wires he pulled out of his 421 . Some of his findings and thoughts mirror what you stated at the beginning of the thread.
 
Champion plugs have been having similar problems and failure rates as long as I have been around aviation. It was bad enough I always used Auburns.
 
What exact problems?

Cracked insulators?
 
How do the resistor issues manifest themselves.

I am really curious because I have a quantity of RHM38S plugs, extracted from a perfectly working IO540-C4D5D, on which all the resistors were pretty high impedance, and (at 1kV) one seems open-circuit, yet all obviously worked fine.

That's just one data point, sure, but it is the exception that ...
 
How do the resistor issues manifest themselves.

I am really curious because I have a quantity of RHM38S plugs, extracted from a perfectly working IO540-C4D5D, on which all the resistors were pretty high impedance, and (at 1kV) one seems open-circuit, yet all obviously worked fine.

That's just one data point, sure, but it is the exception that ...


The plugs wouldn't fire correctly under pressure.
 
The plugs wouldn't fire correctly under pressure.

Exactly, and if one takes a Champ that is eroded beyond limits but still sparking under pressure, and takes that resistor out and puts it back in again, it probably won't fire anymore. I have no idea why that should be.

Auburn's plug had a molded-in carbon (graphite?) resistor that seldom made any trouble at all. Unison copied it but didn't get it quite right, and they gave me trouble, too.

Dan
 
Another piece of data:

I've measured both the removed resistors with an insulation tester whose voltage can be varied from 0 to 2500V RMS and the current is measurable 0-100uA.

The "low Z" sample reached 100uA at just 30V, corresponding to its ~300k resistance previously measured at both low voltage and 250/500/1000V.

The "open" sample actually conducts fine at higher voltages. It reached 100uA at 1.5kV.

I also tested one of the unopened plugs (one of the set which all measure ~1M at low voltage) and it reached 100uA at 350V, and was linear below that.

None of the samples exhibit traditional voltage dependent resistor behaviour (i.e. a sudden fall in resistance above a certain voltage threshold) at the voltages tested i.e. below 2.5kV RMS.

So this explains how you can get the silly resistance readings but the spark plugs all work just fine.
 
That they tend to pull high power settings (80+%) for prolonged periods of time based on the knowledge that they can't possibly hurt their engines the way they do it.

It would be nice if their knowledge was based in facts.
 
Another piece of data:

I've measured both the removed resistors with an insulation tester whose voltage can be varied from 0 to 2500V RMS and the current is measurable 0-100uA.

The "low Z" sample reached 100uA at just 30V, corresponding to its ~300k resistance previously measured at both low voltage and 250/500/1000V.

The "open" sample actually conducts fine at higher voltages. It reached 100uA at 1.5kV.

I also tested one of the unopened plugs (one of the set which all measure ~1M at low voltage) and it reached 100uA at 350V, and was linear below that.

None of the samples exhibit traditional voltage dependent resistor behaviour (i.e. a sudden fall in resistance above a certain voltage threshold) at the voltages tested i.e. below 2.5kV RMS.

So this explains how you can get the silly resistance readings but the spark plugs all work just fine.


What kind of difference happens when you add pressure to the equation?
 
I have no way of testing that, especially with the bare resistors :)

I'd be amazed if it made any difference.

What pressure will affect is how much voltage across the gap is needed before the gap flashes over.
 
I have no way of testing that, especially with the bare resistors :)

I'd be amazed if it made any difference.

What pressure will affect is how much voltage across the gap is needed before the gap flashes over.

Any pressure change will affect the dielectric properties of the air gap.
 
I glued the two resistors previously mentioned to a piece of tufnol and milled them down with a carbide cutter.

Pic is here
http://i101.photobucket.com/albums/m74/peterh337/RHM38S/2resistors.jpg

The slightly damaged one "felt" harder to mill. Unfortunately I can't now tell you whether it was the low-Z one or the hi-Z one :)

But the texture looks identical, and there is no obvious internal damage in either.

So how could they be so very different in their behaviour? Could of course be the composition, or it could be the hi-Z one had microscopic cracks in it.

You can see the thin nonconductive coating on the outside, which explains why one cannot get a resistance measurement along the resistor.

Tomorrow I am going to pull out the top 6 plugs which have been in 2008-present and measure them at both low volts and with the insulation tester. With 6 years between the two purchases, the result will be interesting...
 
I have another data point... rather interesting.

I've pulled out the six top plugs from the engine and measured them. They all measure 15-25k. I have no reason to think the six bottom ones will be any different and since I swap them all round every 40 hrs or so, it would be statistically astonishing if they were. These plugs were fitted in 2008 and have done ~400 hrs. Note the previous set, all ~1M, had done 700hrs.

So we have these possibilities:

1. The resistors were always 15-25k and never changed.

2. The resistors were say 3k and they degrade gradually, taking a jump from a few k to 15-25k before 400hrs, and then a big jump to say 1M between 400hrs and 700hrs

3. Champion changed the resistor mfg process between 2002 and 2008

4. Some combination of the above :)

My money is on 3. definitely, and 1. possibly.

I numbered the plugs and will recheck them at the next check, in ~20hrs' time. That WILL be interesting.

I didn't bother testing them all with the 250/500/1000V insulation tester because values like 20k just read as close to zero.

It would be really interesting to get a brand new RHM38S and measure it. But anybody who does that is going to blow away ~$100 because they come in sealed packages.
 
new resistors are 700 to 1500 ohms. In any brand spark plug. They exist to prevent premature electrode wear, nothing more.

there's a blog with more on the topic at p=690"]http://www.flyplatinum.com/blog/?p=690[/URL]
 
As an update, I have ordered Tempest URHM38S and will chuck out the Champions at the next service. The Champions are obviously all defective, with resistances several times bigger than they should be. They still work (have just flown several hours at FL200) but clearly are on their way out.

Tempest did have some manufacturing problems with these but they are now solved. I got a good detailed response from the chief exec there which is very reassuring.
 
Back
Top