Major arcing with new stator panels

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
After spraying about 16 panels this weekend with 2 thick coats of automotive clear coat I reluctantly performed an arc test on one of the panels. I was shocked to learn that these panels arced through the clear coat enamel like crazy...

My question is that could it be possible that the cleat coat is still drying and conducting. I has only been about 24hrs since the last coat and even though we used an accelerator, we did not bake it. Has anyone had a similar experience.

...Will it eventually become less conductive and stop arcing, i don't want to create carbon trail by continued testing.

Or, am I screwed and need to paint some sort of insulating paint such as Glyptol or something?

FYI, there is also 2 coats of Hylux on the panels already! (it was heat cured and also failed miserably).

Thanks,
Doc.
 
How thick is the coating?

The coating need to be quite thick.

Most of the data I have found for pants and generally clear coats seems to be about 400 to 500 volts per mil sometimes a little more as they have been starting to be inculding this data in the info sheets due explosions at the gas pumps cuased from static discharge to the body of the car through the paint.

I just picked up a little sprayer gun my self and I will be researching paints again.

jer :)
 
nylon powdercoat would be the number one choice of material (Nylon 66) as it has about the lowest percentage of creep you are likely to find in a powdercoat (that was what ML used at one time I cannot say what right now) polyester would be an ok second choice for shear dielectric strength but good old enammel is probably well worth a try. Next time test first though I am sure you have already established that. Once you have punched a hole through your insulation you can pretty much kiss that stator goodbye as you now have a carbon track through the dielectric which may as well be a piece of copper at these voltages. Most powder coaters have absolutely zero experience coating for dielectric purposes. Just ask them, they most all are concerned with cosmetic or corrosion applications. Consistant uniform coating thickness is important. Guess that`s why I have for the most part gone the insulated wire route as it is tested and backed by the manufacturer. You are going to have to do the R&D and figure it out with your coater. I do know that ML spent about a year working with their coater (a major company) before they met all their criteria. There are people who can do this for you who have the knowledge and experience but it will cost, you can even purchase coated panels last time I looked. Try here. Best regards Moray James.

default
 
You should check with the paint manufacturer and check to see how it is with respect to creep. As I mentioned Nylon 66 is one of the best you will find. The voltage gradient the maker supplies is tested on a flat even surface with a consistant film thickness. Perf metal presents the challange of sharp edged holes from which the coating will creep back from leaving a thinner coating once set up. This is a project with challanges and you might do well to follow the lead of others who have had success otherwise you will be re inventing the wheele on your own. Good luck and keep us all postedd as to your results good and especially the bad ones so others can have the advantage of your experience. Best regards Moray James.
 
I've had good results with automotive 2-part poly clear coat by "misting" on many coats, with ten minutes or so tack-up time between coats, to first build a foundation, then thickness, until I have at least 10 mils on each side of the panel (I measure it with a vernier caliper or micrometer). I then spray on another final coat (or two) "just wet", for cosmetics. Typically this process exceeds the sprayable pot life of the paint so I end up mixing a second batch of paint to complete the coating.

If the coats are sprayed on thick and wet, the paint will flow away from the sharp hole edges, and that's where it's needed the most. Better to spray it too dry than too wet. And I suspect you don't have enough coating thickness on those panels too. I recommend at least 10 mils on each face.

Even though I've had success with catalyzed polyurethane, the stuff is dangerous to inhale and expensive as hell-- I'm thinking if I build more panels in the future, I will try to find a local powder coater who's willing to do up a few test panels first. I think powder coating is probably cheaper too.
 
Last edited:
Regarding the wholes... I read in an article on electrostatic panels that you have to worry about applying too much paint to where it keeps the charge from crossing the holes?

I would like to see that article too.
I have not heard that, but what I have heard is that excess coating thickness can close up the holes to the point where it impedes the sound coming thru the holes; resulting in lower output.

The bad rap on spray coating perf stators is that you get too much coating where you don't need it and too little coating where you do need it. I would have to say that's indeed true, but if you're going to use perf stators, you need to coat them with the greatest care you can muster to mitigate arcing.

On one set of panels I that spray coated with polyurethane, to 12 mils thickness, the hole diameters had only decreased by about .006" (average); which means that coating thickness on the hole side walls was only 3 mils, and likely even less on the sharp edges.

This happens because it's not possible to hold the paint gun perpendicular to the side walls of the holes, so all you get there is whatever overspray happens float in there and attach. We don't really need much coating to insulate the hole sidewalls per-se, but we do need a good buildup on the sharp edges and the only way to get there is to apply the spray perpendicular to those edges, and not so wet that it flows away from the edges.

Knowing this, when I spray a set of panels, I "dry spray" the coats, holding the gun at a 45 degree down angle (to the panel face) on the first coat, then a 45 degrees up on the next coat, then 45 degrees left on the next, then 45 degrees right on the next... repeating that sequence until I get at least 10 mils on the panel, then spray the last coat "just wet", at 90 degrees (perpendicular), for cosmetics.

The biggest mistake is spraying too wet, which causes the liquid paint to pull away from the sharp hole edges. At best, you're never going to get a uniform coating, spraying a liquid onto perf metal. But with care and proper technique, you can get a coating that will work with moderate voltages.
 
Last edited:
After spraying about 16 panels this weekend with 2 thick coats of automotive clear coat I reluctantly performed an arc test on one of the panels. I was shocked to learn that these panels arced through the clear coat enamel like crazy...

My question is that could it be possible that the cleat coat is still drying and conducting. I has only been about 24hrs since the last coat and even though we used an accelerator, we did not bake it. Has anyone had a similar experience.

...Will it eventually become less conductive and stop arcing, i don't want to create carbon trail by continued testing.

Or, am I screwed and need to paint some sort of insulating paint such as Glyptol or something?

FYI, there is also 2 coats of Hylux on the panels already! (it was heat cured and also failed miserably).

Thanks,
Doc.

I would check the coating thickness with a caliper or micrometer (panel thickness, less metal thickness, divided by 2). If the coating thickness is under ten mils (0.010"), I would hang them up and mist on some more coats until you get there.
 
1/16" is good but if you start to venture any lower than 300hz then you should consider a larger gap.
I can easily clip my diagphrams to the stator at 200hz with a .075" D/S spacing.

I haven't built a panel with no insulation yet But I plan to try it as a control panel to compare my other ones of the same size and different coatings to.

jer:)
 
1/16" is good but if you start to venture any lower than 300hz then you should consider a larger gap.
I can easily clip my diagphrams to the stator at 200hz with a .075" D/S spacing.

I haven't built a panel with no insulation yet But I plan to try it as a control panel to compare my other ones of the same size and different coatings to.

jer:)

Can someone please compare the benefits and drawbacks of insulation versus more gap for air.

My intuition is that you are ahead by going for bigger air gaps and easily attained higher bias voltages rather than whatever may be the reasons for using insulation.

For sure, there are no uniformity issues with air. But then there are no humidity issues with insulation.

Ben
 
Can someone please compare the benefits and drawbacks of insulation versus more gap for air.
I think this is really 2 questions in one.
1) When adding insulation, are you taking up precious air-gap space that is needed for diaphragm motion?
No. Insulation materials commonly used in ESL construction (PVC for wired stators, Nylon 66 for perforated metal stators) have high dielectric constant compared to air, and are fairly conductive compared to air. The end result is that very little (DC) bias voltage and very little (AC) stator voltage is lost across the insulation.
More details here:
http://www.diyaudio.com/forums/plan...truct-cube-louver-acoustat-7.html#post2154621

2) If the insulation is conductive for both DC and AC currents compared to air like mentioned above, why have it at all?
Protection from high current arcing. When stator voltages reach an amplitude such that the air in the gap ionizes and conducts, the air suddenly becomes a much better conductor than the insulation. While the air is ionized, the insulation then becomes the impedance that limits the current of conduction from stator to stator and stator to diaphragm. If the insulation was not there, high current would flow between stators....high current arcing and diaphragms catching fire. I’ve even gotten a nice Jacob’s Ladder effect in one of my un-insulated test panels.

More discussion on the role of stator insulation in this thread:
http://www.diyaudio.com/forums/planars-exotics/154203-role-esl-stator-insulation-2.html#post1967738

BTW, as long as you limit the stator voltages to known safe levels there is no problem whatsoever with using un-insulated stators.


My intuition is that you are ahead by going for bigger air gaps and easily attained higher bias voltages rather than whatever may be the reasons for using insulation.

You intuition on using larger gaps may be missing half of the equation.
You mention that higher bias voltage needed for bigger air gaps is easily attained. I agree.
But, ESL output is proportion to (bias voltage) * (Stator voltage) / (D/S)^2.

So, if you double the gap size and then double bias voltage to compensate, your ESL would now have -6dB lower output. To regain the 6dB of output, you would need to double the stator voltages as well. This would require doubling the step-up ratio of the transformer. This is a problem, and the reason that air-gap is generally chosen to be 1mm - 3mm. Because of the nature of transformer parasitic and power handling, doubling the step-up ratio of a transformer will require reducing the bandwidth covered…either at LF(core saturation), HF(leakage inductance), or maybe a bit of both.
 
Bolserst -

Thanks very much for such a full and clear post. I would only quibble about the big-discharge point by asking if the usual custom of having a giant resistor in series with the bias not enough to control it?

For sure, nobody wants a Jacob's Ladder display, but keeping an eye on bias voltage as the weather changes was, I thought, an ordinary ESL owner task.

Putting things together, I am still wondering if there isn't a gap like 3-5 mm that makes no insulation the best choice?

Small question: does the insulation "eat up" bias voltage across its depth faster than air?

Ben
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.