Provisional 25 kV lower brush power supply


I decided to build the 25kV brush power supply by using an automotive ignition coil. The main consideration was the supply must have no semiconductors in the drive circuit, as I was unable to devise a fully convincing scheme to shield vulnerable semiconductors from intense high voltage pulsed electrostatic fields.  Hence, I chose to use a very conventional design based on the old-time (1960s) automotive ignition systems. These have proved to be very reliable in practice.

When I initially tested the van de Graaff, the lower brushes were connected to ground.  The voltage produced at the upper terminal was purely triboelectric in origin (vinyl belt on a PVC roller), and I thought I could impove the output (about 80 microamps) by using a sprayed charge scheme. The apparatus below supplied 25 kV (or more) at 100 to 200 pulses per second.  Unfortunately, when it was connected to the lower brushes, the tests showed no difference in current output. Hence, for now I have decided not to formalize this scheme.

(Afterword:  This problem was never adequately investigated. An ignition coil can produce a monopolar pulse but the inclusion of a capacitor across the breaker points will produce a ringing waveform.  This will give a robust spark for a spark plug and is fine for automotive use.  The ringing itself would not be a problem in this application if the waveform was always positive, but the setup as shown would produce a zero crossing waveform and would not be suitable here. However, a string of series-connected high-voltage diodes was found in the project box a year later.   A test procedure was never written, and my memory of these trials is too fuzzy to be reliable. I hope to reinvestigate this problem sometime in the future. (Afternote: the diodes were later tested with a 9V battery, a 100 ohm resistor, and a voltmeter. Tests  showed all the diodes were "open" and non-functional. They were NTE517 microwave oven diodes (15kV @ 500ma). These are probably not suitable for fast switching applications. )


Cam was made from 0.625 square bar stock. It looks like a big bolt with a square head. The points are Honda CA102 CT200 CB92 CA95.  The ignition points capacitor was obtained locally.  The cam is driven by a variable speed electric drill. A wood block serves as a bearing and mount.  A 12V battery and a "universal"  ignition coil, (ACCEL 8140 Super Stack) make up the rest of a very conventional "brute force" 25kV pulsed power supply.

An alternative I might investigate involves coating the lower rollers with Teflon tape (or Teflon roll  covers) because its position in the triboelectric series indicates it should improve the output. As shown in the table below by Alpha Lab, Inc.,  the nC/J values for vinyl, PVC,  and Teflon are -75, -100, and -190, respectively. The present output (80 microamps) is based on the difference between -75 and -100 (a difference of only 25). With a Teflon coated roller the numbers would be -75 and -190, a difference of  115; this implies a factor of 4 improvement , and possibly a current output as high as 300 microamps. This would significantly improve the repetition rate--something crucial for Project Whitefire experiments. Inspection of the table shows that other combinations could go as high as a factor of 10 (800 microamps!).  (The problem is in locating materials, and maintaining reliable performance. )   See below.

Another thing the needs testing is the position of the brushes on the lower belt rollers. Should they be along the bottom of the belt, or should they be where the belt separates from the roller? The latter seems to be more logical (but later tests showed  no significant difference in performance).




http://www.trifield.com/content/tribo-electric-series/  :


Teflon sleeves for rollers are used by various industries and are available commercially. Here is an example screen shot of the sleeves and supporting information.


http://www.americandurafilm.com/film-distribution/heat-shrinkable-rollcovers-saint-gobain/




http://www.fluoron.com/salesbook/FEP.pdf
 http://americandurafilm.com/data-sheets/rollcover-data.pdf  

Pricing (June 2015) for a minimum quantity was quoted as:
2” x 3’ - $60.70 each
2.5” x 3’ – $73.90 each
However, the overall minimum for an order was $250.
_____

http://www.holscoteurope.com/en/fep-heat-shrinkable-sleeves-and-teflon-heat-shrinkable-sleeves/  


teflon roller covers, heat shrinkable sleeves

_____

http://www.fluoron.com/fep.htm  http://www.fluoron.com/salesbook/FEP.pdf  "roll covers"





This is a test of Teflon tape wrap. The "glue" is uncured polyurethane foam dissolved in a small amount of acetone. It is smeared (use vinyl gloves!) across the entire roller and then the roller is wrapped with 3/4" wide Teflon tape. The direction of the wrap is such that the torque will tend to tighten the tape. The lathe is unplugged and is only  used as a  convenient fixture.

html 4/2015