Sparkplug resistance and dielectric and surface avalances.

[Vtile]

Hello

I wonder how to measure resistance of old small engine spark plug?

[Gyro]

It sounds like a job for a high voltage insulation or breakdown tester. You probably want to be testing it at around 10kV.

A replacement plug is likely to be a cheaper option?

[penfold]

For what purpose do you want to know the resistance? I think it could greatly affect the best approach,

[Vtile]

Hello.

For no particular reason at all. Just academic interest, which I partly satisfied. I originally posted it in metrology section, but this interest was judged as some form silly repair attempt and I was politely asked to get out of the lawn... So I did, here is now the official common spark plug interest group. Maybe one or two plugs will be eventually replaced or cleaned as usual and meanwhile some low current or high ohmic or high voltage measurements are done for no particular reason, but as freetime amusement.

PS. Resistance is not the way to troubleshoot sparkplugs. Use the motor as instructed at any youtube repair video.
(Tech. Note just for interest. HV characteristics are not the same in free air as in working cylinder. Per usual this doesn't matter for usual troubleshooting)

[penfold]

here is now the official common spark plug interest group.

Ooh, I feel like we should commemorate this inaugural meeting somehow

Well, as everyone has already screamed, you need a high voltage source and a means of measuring resistance, if you do this at a few different voltages you can check for linearity.

As Gyro suggested, the best "tool" for the job is to use an insulation breakdown tester which (the newer ones I've used at least) have circuits to ramp the voltage up gradually and find the point at which breakdown just "wants" to initiate by looking at the noise characteristics of the current... you can do it manually, but it needs a very well protected oscilloscope and a lot of patience

[perieanuo]

in fact, typical spark plus most important characteristic is not resistance, but insulation/breakdown voltage
resistance??? you mesure the air or whatever gas you have there, in fact is the resistance of that gas, depending of his humidity, purity and so on

[HighVoltage]

The surface resistance of a spark plug is also pressure and temperature depending, especially when you have black carbon deposit on the surface of the insulator.

To get realistic properties of a used spark plug, it is best to pressurize the plug to about 10 bar air or nitrogen pressure and then apply a DC or a pulsed high voltage signal, so see when breakdown will occur. Before breakdown will occur, you can measure the current and calculate the resistance.



 

[perieanuo]

see, now we're talking

[Gyro]

Yes, temperature and pressure are big factors, humidity is another one [Edit: as mentioned by perieanuo] - a plug in an engine that is being started from cold gets pretty wet with fuel and moisture. High operating temperature of the electrodes will reduce the spark gap breakdown voltage too. I can't think of an easy way to experimentally duplicate this (short of heating the gas in the pressure chamber to high levels). If you play a flame on the electrodes then it will act as an ionization source, allowing current to flow - the principle that is used for pilot light detection in gas boilers.

Being pessimistic, there is no way that you will get a specific relevant resistance value, there are too many environmental variables. The best you will possibly manage is a bunch of interacting curves based on each of the variables and their combinations. Establishing breakdown voltage is a more tangible goal, but again, no single answer due to the variables.

You can see why you were gently nudged out of the Metrology section - they like precise predictable values over there! 

[Gyro]

In the interests of science, I just grabbed an old, well used Bosch car spark plug from the garage and stuck it on my 10kV breakdown tester...

At room temperature and pressure (probably a little damp because I brought it in from the cold garage), it breaks down at a repeatable 5.2kV (centre electrode -ve).

At 5.1kV  I can't see any resistive leakage current on the 1uA FSD scale - no detectable meter deflection, which would probably put it well below 10nA.

[Vtile]

As quick calculation between work, the equivalent resistance would be like more than 500 Gohm Tohm at 5.1kV/10nA. Assuming DC.

 

Edit. Counted zeros incorrectly, oops. 

[Gyro]

I have to admit, that's a much higher value than I would have expected! It's educational in terms of modifying spark plugs for use as low leakage feedthough insulators / terminals.

I think I'll leave the plug in situ in a cold, compressed, damp, engine cylinder test to somebody else. 


P.S. Some updated readings...

After stabilising to ambient for an hour or so, the breakdown voltage has increased to 7.8kV. If I eyeball the meter very carefully, I would guess at about 1nA 10nA current at 7.7kV.

I'm not sure about the increased breakdown voltage, maybe some moisture was originally evaporating off into the gap. I don't see any obvious evidence of corona discharge - the centre electrode must have lost any sharp points after around 40k miles of use but it might account for a little current too (visible / audible corona shows up very clearly on the meter when present).

770G ?

[Vtile]

I have to admit, that's a much higher value than I would have expected! It's educational in terms of modifying spark plugs for use as low leakage feedthough insulators / terminals.

Same to me, I originally expected something like a few gigaohms maximum when I did my experiment, how wrong I were.

I think I'll leave the plug in situ in a cold, compressed, damp, engine cylinder test to somebody else. 

  Unfortunately I only do have 1kV DC source, which is far too low voltage for these kind of things.

P.S. Some updated readings...

After stabilising to ambient for an hour or so, the breakdown voltage has increased to 7.8kV. If I eyeball the meter very carefully, I would guess at about 1nA 10nA current at 7.7kV.

I'm not sure about the increased breakdown voltage, maybe some moisture was originally evaporating off into the gap. I don't see any obvious evidence of corona discharge - the centre electrode must have lost any sharp points after around 40k miles of use but it might account for a little current too (visible / audible corona shows up very clearly on the meter when present).

770G ?

Yep. I did calculate the zeros wrongly. Oops. The moisture (breath) seems to have "dramatic" effects (relatively) to resistance as does some liquids. While IIRC moisture at air increases the dielectric strength, but it usually is other way around with solids and fluids (what I know). Which also were one part of the 'what if' and "hmm.. how this behaves" thinking I had.

What HighVoltage replied about the pressure and suggested to increase the pressure to 10bar air or nitrogen. For spark cap the increase of pressure usually decreases the dielectric strength and decrease in pressure will increase the dielectric strength, with certain window. This is counter intuitive. The physics involved are highly non-linear. Paschen's law gives more information of breakdown on gases.

[penfold]

This is counter intuitive. The physics involved are highly non-linear. Paschen's law gives more information of breakdown on gases.

If you add in to that how metal-insulator-air junctions interact at varying pressures it becomes very "interesting"/frustrating

[joeqsmith]

It's educational in terms of modifying spark plugs for use as low leakage feedthough insulators / terminals.

Interesting thought.

What does "avalance" mean?  I wasn't sure if the OP was asking about the embedded resistive element or the insulator.  If you wanted to measure the element, an ohm meter would get you close enough.   If you want to measure the insulation resistance, that's going to be more difficult.   Used plugs are going to vary a lot.    If your want to measure the breakdown, that's also going to vary a lot.   You may want to measure the complex impedance at these frequencies.   If you want to measure the breakdown voltage, again, seems like there are a lot of variables to consider.   

A few pictures of my test jig.

[penfold]

You may want to measure the complex impedance at these frequencies.   If you want to measure the breakdown voltage, again, seems like there are a lot of variables to consider.   

Interesting you should say that, I had frequent problems with some mixed HV and RF structures (ion traps and reflections) which behaved very unpredictably when breakdown tested at DC vs having an HF component such as a steep pulse rise time or RF, and it wasn't due to resonances or overshoot. I unfortunately it was cheaper to change geometry than investigate, but I do wonder if there's any significance of what was going on there to how to spark plug may arc over

[joeqsmith]

You may want to measure the complex impedance at these frequencies.   If you want to measure the breakdown voltage, again, seems like there are a lot of variables to consider.   

Interesting you should say that, I had frequent problems with some mixed HV and RF structures (ion traps and reflections) which behaved very unpredictably when breakdown tested at DC vs having an HF component such as a steep pulse rise time or RF, and it wasn't due to resonances or overshoot. I unfortunately it was cheaper to change geometry than investigate, but I do wonder if there's any significance of what was going on there to how to spark plug may arc over

https://lss.fnal.gov/archive/other/ssc/sscl-539.pdf

[Gyro]

A few pictures of my test jig.

Ha, I should have guessed that you would already have a high pressure  'torture chamber' for the job! 

[Vtile]

This is counter intuitive. The physics involved are highly non-linear. Paschen's law gives more information of breakdown on gases.

If you add in to that how metal-insulator-air junctions interact at varying pressures it becomes very "interesting"/frustrating

ok, my english just let me down.. By metal-insulator-air junction do you mean conductor-insulator-gas system or metal-oxide-gas boundary.

Interesting never the less.. 

Oh, how should one count radiation and light as ionizators on these spark plug repairs.  Maybe using slightly radioactive gasoline would solve some of the ignition problems when spark is weak.

[penfold]

My apologies! TBF it was my English letting you down - conductor-insulator-gas(or vacuum) (the alternative is calling them triple-points but that's even more misleading!),
The presence of absence of gas has some interesting effects and HF voltages tend to defeat the shielding measures that work well at DC

[Vtile]

My apologies! TBF it was my English letting you down - conductor-insulator-gas(or vacuum) (the alternative is calling them triple-points but that's even more misleading!),
The presence of absence of gas has some interesting effects and HF voltages tend to defeat the shielding measures that work well at DC

Which one of my hypotheses are more wrong (or both) that it is related to (read beginners random thoughts)..

A) relative wave lenght and relative wave phase on crossection of insulator.
B) relative risetime (V/t)

I have two books on these HV and ionisation, one is written 1949 (physics of electricity and matter) and 2nd 2015 both have encouraging statements (this and this) phenomenom or physics is not exactly understood. 

Ps. ETA needed to use some google fu to get the idea of triple point (assuming I did find right articles).. Maybe tomorrow I will read a bit of those, if not mathematic simulation printed on paper.

[joeqsmith]

Maybe using slightly radioactive gasoline would solve some of the ignition problems when spark is weak.

With increased intake pressures, I just tighten the gap.   My test jig uses nitrogen which obviously is not going to behave the same.   So it's a bit of tongue and cheek.   

I had a couple of NGK CR9EK plugs on the shelf.  One used, one new.  Measuring the internal resistive element, I got 7.50k for the used and 5.72k for the new plug.  I doubt this has anything to do with the plugs being used or not but rather just sloppy tolerances.     Looking at the conductance of the insulator, I would need to set up the electrometer.   Using my handheld meter, there's a measurable difference.    I would want to clean the outside of the insulators first (same when using HV resistors).  I think I may have one of these that I fouled when a coil pack started to fail.  That should easily show a difference as well.   

[Vtile]

Maybe using slightly radioactive gasoline would solve some of the ignition problems when spark is weak.

With increased intake pressures, I just tighten the gap.   My test jig uses nitrogen which obviously is not going to behave the same.   So it's a bit of tongue and cheek.   

I had a couple of NGK CR9EK plugs on the shelf.  One used, one new.  Measuring the internal resistive element, I got 7.50k for the used and 5.72k for the new plug.  I doubt this has anything to do with the plugs being used or not but rather just sloppy tolerances.     Looking at the conductance of the insulator, I would need to set up the electrometer.   Using my handheld meter, there's a measurable difference.    I would want to clean the outside of the insulators first (same when using HV resistors).  I think I may have one of these that I fouled when a coil pack started to fail.  That should easily show a difference as well.

So you are actually studying these spark plugs for a reason or another or are you just using them for sacrificial parts for the HV-equipment itself? Ok, I have no idea what is the direction of the slope with these tensions and pressures and cap-distances, from your post I could read between the lines that those are downslope, ie. things are behaving in a way that makes sense, smaller cap smaller breakdown tension/voltage.
 
I'm wondering the first picture of yours, it looks like "low-voltage" (<1000V) front panel, is it function generator maybe? 

Been thinking for years to build or buy HV-tester, but frankly I do not like the idea to get zapped, while doing something quickly at late at night when someone should be already at sleep.  As it is not only the wires and DUT, but the damn tables and chairs and everything starts to charge up (been there, done that).