This machine is another variation of the Holtz machine, invented in 1873 [p85]. It consists (schematic)of a rotating disk of insulating material (originally glass) that rotates behind two inductors plates, one at the top and other at the bottom. These inductors are charged by combs collecting charge from the front of the disk, before it passes under the charge collectors. The charge collectors collect charge from the front of the disk, at the sides of the machine. At the back side of the disk, there is a vertical neutralizer bar with combs, opposite to the inductors in the front. In old books it is possible to find two versions. One with the inductors being vertical metal bars with wood cylinders close to the disk (invented by Weinhold with the objective of avoiding polarity reversals (1887) [p14]) and another (Leyser (1873) ) with paper inductors behind an insulating plate.
In January, 1998, I built a machine that could be mounted in both versions to investigate this machine. The tests showed that the version with paper inductors, at least as I initially built it, is not reliable. It is difficult to excite it, and once excited it works for awhile and suddenly loses excitation. Apparently, the mechanism that causes polarity reversal in the Holtz machine doesn't work there. The version with bar inductors works better, but results in a weak machine due to the absence of insulation other than air between the inductors and the disk. At least, it doesn't revert polarity nor stops working while the disk is turning. Another view. For this version I used a simpler driving system.
I tried also to mount the paper inductors in a different position, with the charging combs taking charge after the disk passes under the main charge collectors instead of before. This configuration is equivalent to a modified Holtz machine (see my comments about Holtz machines) with the output being taken at the inner side of the rotating disk and the normal output being used as neutralizer. The machine works evidently better in this configuration, maybe because of better insulation, but still loses excitation after some time instead of reversing polarity.
The next experiment was to add metal sectors to the disks and brushes instead of combs in the neutralizers and the inductors. I applied 16 aluminum sectors to each side of the rotating disk, in Wimshurst style, and made brushes using single thin nickel-chromes wires. The inductors were also replaced by aluminum plates, with brushes fixed to them by round wood blocks glued to their surfaces. I found that this solves the loss of excitation problem, converting it into polarity reversals at each few seconds. I tried two versions: one charging the inductors by brushes located before the charge collectors, as in the original design, and another with the brushes located after the charge collectors, as above. Both versions showed the same performance. The machine self-excites easily, but the maximum voltage obtained, is low, producing at most 2 cm sparks between the 2.5 cm terminal balls, going to 4-5 cm with a small ball fixed to the positive terminal. This is a poor performance for the 30 cm disk. Without the sectors the spark length was about twice greater. The short-circuit current reaches 7 uA, and polarity reversals occur at each few seconds.
Finally, I returned the machine to a sectorless version, making a new acrylic disk a bit larger (31 cm), mounting again combs in the neutralizer circuit, but keeping the metallic inductor plates and brushes touching (short brushes close to the disk surface are enough) the rotating disk to charge them. I returned, at least initially, also to the simple driving system, because more force is required to turn the machine in this configuration. This experiment was successful, with the machine showing periodical polarity reversals instead of losing excitation, and producing much higher voltage and current (20 uA) than the previous attempts. With a steel ball fixed to one of the terminal balls, separated from it by a 0.5 cm plastic tube, the machine produces easily 10.5 cm sparks even with 85% of relative humidity in the air, reverting polarity after each long spark, much as my Holtz machine does. The machine requires external excitation to start, what is easily done by the application of some charge to one of the inductor plates. It keeps the excitation if stopped for some time, and works even when one of the terminals or one of the inductor plates is grounded. This is a curious characteristic, consequence of the absence of grounded connections at the side of the rotating disk where the output circuit and the inductor plate charger brushes are.This machine, however, shows rather inconsistent performance, one day producing long sparks, and almost nothing in another. I could trace one reason for this behavior as corona leaks developing at the borders of the fixed inductors, and fixed partially the problem by covering the borders with self-adhesive plastic foil. This resulted in significant increase of the average spark length, that now reaches 10 cm with only the normal terminal balls.
Last update: 22 February 1999
Antonio Carlos M. de Queiroz
Return to Electrostatic Machines