Spark plug questions for non-auto application

[cryomark]

I'm investigating characteristics of Pt spark plugs for use in a gas contamination detector.(Not gasoline)

Our current instrument uses a continuous arc drawn between chisel tip tungsten electrodes powered by a 1KV DC power supply.A spectral peak of the arc varies with the degree of contamination (N2 in He). The light is focused on a scanning monochrometer tuned to the relevant wavelength.

The stability of the arc is critical to accurate measurements. The large size of the electrode chisel tip(~4mm) allows the arc to "wander" out of ideal optical calibration, especially if physical/chemical contaminants are introduced into the arc chamber. Mechanical shock will also upset alignment. Machining and parts cost play a factor in our desire to replace this design.

Questions are:
1. In an auto,is the applied V at the plug center electrode + or -, relative to the grounding electrode?
2. How does our continuous I of 11mA compare to the arc current of normal automotive applications?
3. Power rating of internal R in resistor type plugs?Are Pts available w/o R?

The physics of arc propagation are not completeley understood, and we are concerned about erosion/deposition on the plug electrodes.
With its small cross sectional area, a flush trimmed center electrode would seem to offer the more
precise and stable arc point.

 

[sciguyjim]

Cryomark,
That's a nice idea, definitely worth a try. I'd say that the 11 mV of the continuous arc is WAY below the peak current of a spark plug discharge. Spark plugs in engines need much higher voltage and current because the spark has to jump in a high pressure gas and this is harder than in a low pressure gas. Be prepared to have to change your arc current and voltage to suit the new electrodes. Are you going to use 2 spark plugs, with center electrodes only, as your electrodes? I would. Cut off the plugs' ground electrodes. They are not as resistant to arc erosion as the center electrodes. In fact, this is the place you may run into trouble. In many continuous arc systems, the electrodes are expected to be gradually consumed by the arc. You are lucky that the arc can run on only 11 mV, it may be what saves your electrodes from burning up too fast. Nevertheless, I say give it a try!

 

[GregLocock]

Quoting from Bosch

Typical spark duration is 1.5 ms , which consists of an initial ignition spike to 15 kV followed by a steady state spark at 1.5 kV.

Typically the coil stores 60-120 mJ of energy, and could deliver 30 kV

11 kV for 0.3 ms is typically needed to initiate a spark.

Does that help?

Cheers

Greg Locock

 

[evelrod]

The opposed spark plugs idea sounds doable. In fact I kinda like it and it has the benifit of adjustability while in operation. NGK makes some rather exotic racing plugs that might stand up to continuous arc (momentary arcs in the 40kV range and up are common). How long is anybodys guess. To my knowledge your in virgin territory here. From using the old fashioned spark plug testers as a kid, it takes a lot less voltage to initiate an arc at low pressure.
Would contamination of tested samples by eroded electrode material (tungstun, platinum, copper etc.) be a problem?


Rod

 

[sciguyjim]

Evelrod,
"...Would contamination of tested samples by eroded electrode material (tungstun, platinum, copper etc.) be a problem?"

Yes, this would always be a problem, even if brief sparks were used instead of a continuous arc. This is why the light goes through a monochromator, it separates wavelengths that would interfere with the analysis so only the wavelength for the contaminant is monitored.

 

[cryomark]

Thanks for all the response.
To give a bit more background,
The device, called an arc cell, is used for real time monitoring of process gas contamination in the compressor discharge of a large liquid helium refrigerator. Everything but He freezes out at our operating temps(4.5 Kelvins). We have N2/He heat exchangers and portions of the system run at sub-atmospheric pressures so there are a number of avenues for contamination.
The cell pressure is regulated down to ~15psig. The flow rate is ~1 scfm. Gas is vented to atmosphere from the cell.
Arc across .125" gap initiates at ~1.1KV, we generally run the PS from 700-800V. ~550V across a 50Kohm ballast R, arc V~200V.
This gives arc power ~2 Watts.
The environment is benign compared to an engine-low pressure, low ambient temp, inert gasses (we have a couple devices calibrated for an argon spectral peak)

Sciguy-we had considered just replacing the tungsten chisels with plugs per your idea. Alignment difficulties with the optical path remain. The re-design of the total package includes replacing the monochrometer, a large and expensive device at ~150mm focal length with a monochromatic filter/ photo diode at ~35mm. Since we're looking for reductions in cost, size and complexity, a single plug with a pre-set gap would be ideal.

evelrod-conductive contaminants could be a very big problem-that's why we vent to atomsphere.

Greg-Those are encouraging numbers; assuming no losses and using the maximums you gave. 120mJ/1.5mS=80W. a high revving engine will barely approach 10% duty cycle for the arc, giving an equivalent DC power of 8W.
This says the internal damping R is capable for this DC application.
I'll have to think about the current.
Any opinions about the polarity of the applied V?

 

[sciguyjim]

Cryomark,
For polarity, the center, platinum, electrode should be positive and the ground is negative. During operation of the arc, positive metal ions will be vaporized off the positive electrode. The platinum electrode is supposed to resist this. The ordinary metal ground electrode would erode faster.

P.S. I'd use a "cool" heat range plug so it can dissapate the arc heat as fast as possible so it doesn't burn up too fast. It may not make much difference considering what arc temps run, but it couldn't hurt. Are your electrode mounts water or air cooled?

 

[gbent]

Cryomark
What would happen if you resharpened your tungstun electrodes from a chisel point to a sharp point? A sharp point is used in TIG welding with AC current, and will hold up to currents of over 300 amps. The point needs to be sharpened with axial grinding marks, not circumferencial.