Preventing Corona Discharge from Wimshurst Machine

[jg21]

TL;DR Summary

Trying to optimize efficiency of my DC generator. Losing charge from sharp points.

Hey all, new here. Just finished my first wimshurst machine and I’d consider it to be fairly AWESOME! Having a blast so far.

The thing is, I believe that given the parameters of my build (# of sectors 24, diameter 11”, etc) I’m not quite getting max voltage possible and I’ve confirmed this at night in the dark by watching for corona. And it’s somewhat obvious in the daytime to listening for crackling around the caps and combs.

My question is this:
GIVEN the fact that there are some sharp points on my machine that I can’t avoid retrospectively. How can I beat “seal” my machine from leaks?? I’m basically in search of some sorta liquid Kapton tape or something… will painting gobs of beeswax onto all the sharp points improve performance?

At night I can see discharge from various slightly sharp points/burrs/abrasions on the brass or solder joints. And also I can see a little discharge from the ends of the Leyden jar foil to the wooden base about 1” below since they’re open PVC tubes (ugh, now I know, should have sealed the bottom somehow)

Right now I get ~2” sparks in the late spring conditions here in the northeast.

I want to check before I go to town with beeswax or other, because once I apply that to things like the foil I won’t be able to safely remove it. If it’s actually conductive, then I will have totally ruined my machine… anyone have experience with this? How can I seal my leaky wimshurst?

 

[Baluncore]

At night I can see discharge from various slightly sharp points/burrs/abrasions on the brass or solder joints.

Remove the sharp points where possible.

Make a spherical cage from round wire and attach it to enclose the point generating the corona. That will increase the virtual radius of the conductor surface, which will reduce the corona.

 

[jg21]

Remove the sharp points where possible.

Make a spherical cage from round wire and attach it to enclose the point generating the corona. That will increase the virtual radius of the conductor surface, which will reduce the corona.

What about the sharp points where NOT possible to remove? Like the top and bottom edge of the Leyden jar foil. Etc.
there must be some sorta protection I could paint/slather on to hold a higher voltage no?

 

[Klystron]

TL;DR Summary: Trying to optimize efficiency of my DC generator. Losing charge from sharp points.

will painting gobs of beeswax onto all the sharp points improve performance?

TL;DR Summary: Trying to optimize efficiency of my DC generator. Losing charge from sharp points.

want to check before I go to town with beeswax or other, because once I apply that to things like the foil I won’t be able to safely remove it. If it’s actually conductive, then I will have totally ruined my machine… anyone have experience with this?

Yes, I have some experience operating quite old electronics that used beeswax as an insulator/packaging though not your particular device. Beeswax was a readily available product before plastics but had several shortcomings:

• low melting point
• impurities and varying water content
• sticky surface that attracted lint

I never experimented with beeswax properties, per se, such as resistance and conductivity despite keeping bees at college. I do recommend using a less volatile substance on your generator.

Non-conductive paint or certain adhesives, collectively referred to as dope in early DIY 20th Century texts, might serve you better. I suggest testing trial substances on a mockup of your working device then comparing effects.

As beeswax has become pricey, consider other biological substances as insulators such as various art gums (from plants) and shellacs (from insects), if you want to preserve an old-timey solution. More modern solutions include acrylics with or without additives.

 

[Baluncore]

Like the top and bottom edge of the Leyden jar foil.

That requires you bend a metal tube into a torus that will sit outside, on the end edge of the foil. It only needs to contact at one point.
It could be made from any convenient material, then painted with conductive paint, or covered with gold leaf which looks good and is very low cost.

 

[jg21]

Thanks for all the good input. In future builds I’ll try to incorporate more of those features from the get go (torus, round everything) but for this one I’ve rounded a fair amount of sharp points so it’s a good start.

Did some experimenting and found that my main source of lack-luster sparking was poor electrode design. I did not account for polarity swapping and so my big and small electrodes were only good 1/2 the time. On top of that the sizes of them were not optimal even when polarity was correct… I have ordered a few more size variations of balls to try out. Been reading up on breakdown voltage for different geometries because I’d like to calculate the optimal radii.

Anyone know the reason behind why the large and small balls must match the electron dense and sparse sides respectively?

 

[Baluncore]

Anyone know the reason behind why the large and small balls must match the electron dense and sparse sides respectively?

The emission of electrons requires less potential than the emission of metal ions.
The flow of electrons is reduced by increasing the radius, so reducing the curvature of the conductive surface, since the voltage gradient near the surface is reduced.
https://en.wikipedia.org/wiki/Thermionic_emission