ASTROD Circuit Construction
3 saltwater capacitors
20 feet GTO-15 (rated for 15000V, also known as Sign and Ignition Wire.)
9500-12000V 30ma neon transformer
Rife-Bare style tube (80/20 argon/neon or 100% argon) with internal
1 or 2 120V 3-3.5" diameter boxer fans (optional)
3x3 metal brackets, 2 holes in each leg
1/4 zinc coated (standard) nuts and bolts
12ga 1/4" ring lugs
10 feet 10-12ga magnet wire (or insulated solid core)
Paddle lug disconnects for connecting to tube if desired.
I once used Monster cable or double heat shrunk grounding braid for wiring,
but these required utmost attention to not placing wires of different potentials
near one another. Using GTO-15 makes the placement much easier,
especially if the device will be mounted in a cooler or other confined space.
Mounting the thing in a cooler is highly recommended to both cut down on the
noise and the ozone. This is not a good kind of ozone to breath a lot,
being produced by a spark gap on cheap metal.
Neon transformers are available from neon parts supplies. Call around neon sign shops and ask who their supplier is. I use a 9500V 30ma pull-string transformer and it works well. It cost about $US40. Increasing voltage to 12000V and current to 60ma probably works better, but changes the characteristics of the circuit, and I have not experimented with this to know the ramifications and modifications necessary to use it. 60ma transformers are quite a bit heavier and more expensive than the 30ma type, which is already quite heavy. Beer sign transformers are typically of adequate size and can sometimes be found used for cheap.
The tube used must have internal electrodes. A standard BRG tube is a 28" long 80/20 argon/neon 10mm pressure bubble tube. This seemed a bit long for the circuit (increased length made the beam weak). I had one made which was 24" long with a small bubble which was still 10mm pressure and 80/20. The tube
can get quite hot during use if a large spark gap is used. I made a tube rack to house the tube and mounted a 120V boxer fan in one end to cool it. This allows use for over 10 minutes at a time without the tube getting excessively hot, but I will eventually mount a second fan on the other end to provide longer run times.
Once the saltwater capacitors are made, construction can commence. Attach a foot or two of
GTO-15 to each bare wire coming out of capacitors.
Homemade inductors protect the transformer from the RF produced in the circuit. To make these, wrap 10ga magnet wire (epoxy insulated) 16-18 times around 1" PVC or other 1" form. Plain solid core insulated wire can also be used. Wrap 12ga insulated 20-22 times. Allow a few inches on either end for connections.
Solder the leads from the transformer onto one side of each of the inductors. Again, double heat shrink all bare connections or wires. To make connections easier, Home Depot sells little copper tubes called Splice Connectors (about 10 per box for $2). The wires are slipped into either side and the connector is crimped then soldered. I left the PVC forms on the inductors and drilled holes through the PVC to mount to a board and hold the inductors in place.
They must also be well spaced to prevent arcing.
To make the spark gap, I used metal brackets. These are L-shaped and have two 1/4" holes in either leg. I mounted these to a board approx 6" apart and facing each other. One hole on the extended leg is used to mount the lugs and the other is used to mount the bolts used for a spark gap.
In the first setup I tried, 4" bolts were placed in the top holes of the brackets so the heads were facing each other and exactly aligned. The nuts which hold them could be adjusted so that the gap between the bolt heads could be made from 1/32" to 1/2" or so apart. Wing nuts were used to facillitate easy adjustment.
I also tried carriage bolts and round head nuts so the spark would leap from the rounded heads. This required much cleaning of the heads, though, since the tiny area became black with carbon quickly and they did not work as well. Using the larger flat heads of bolts worked better since the increased area did not need to be cleaned as often.
Then I worked out a method that required the least cleaning of all and is the one I am currently
using in my own device. I found 5" long by 1/4" bolts that had threads on only the last inch. I
cut off the bolt heads so there was <5" long smooth rods with screw threads on an inch or so.
I drilled out the top hole of one bracket so it was elongated into a 1" (by 1/4") oval. I mounted this bracket on a 1/2" stack of washers to shim it up above the opposing bracket. Then I mounted the rods so they were parallel for approx 4". The oval hole allows one rod to be adjusted up and down so the gap can be changed. The parallel portion of the rods is used for the gap and allows a
larger area to be used, so does not need cleaning often, and assuming the rods are not perfectly parallel, a slightly varying gap to perhaps increase the bandwidth. The sparks will occur mostly on the area that is closest together but will occasionally leap on the area that are slightly further apart, especially as the closer area becomes blackened.
To clean the gap bolts, sandpaper is used. With the flat and carriage bolt setups, this had to be done approx every 10 minutes of use or when it begins sparking less. With the parallel bolt setup, this increases to
hours of use.
The next device I built was for a friend and in this one I did not use rods
as the spark gap. I used the short 90 degree brackets (wide style -
2") with the upward legs near each other. This provides an even
larger potential sparking area and even less cleaning. The brackets can be
mounted 3/8" apart and the top of them bent slightly to make a shorter gap.
The tube can be soldered to the leads, but I used paddle lug connectors to facillitate trying different tubes. Since the tube burns quite
hot when using a large spark gap, I mounted the tube in a rack made of 1x2" struts connected to solid 3/4" 5x5" ends. In one end I mounted a 3" 120V boxer fan to blow on the tube. This allows use for about 15 minutes at a time, which is fine since typically only 10 minute sessions are used. When it gets too hot, the glass of the tube turns blue. With a fan mounted in both ends, it could probably run over a half hour or even continuously based on tube temperature.
For a permanent setup, I used a Rubbermaid cooler which cost about $15. It was the only type I found which was tall enough to house the wine bottles. The $15-20 Coleman nor any other cooler I found in this range does not.
On the bottom right inside is mounted the transformer on top of 3/4" board. The drain hole assembly in the cooler unscrews to reveal an approx 1" diameter hole through which the plug of the transformer can pass when it is reamed out a bit. On this side, a power strip was mounted outside the cooler which provides outlets for the transformer and fan(s). This allows the device to be powered on and off by using the power strip switch, instead of the transformer pull string.
The spark gap and inductors are mounted on the same board and on the front right inside the cooler with the spark gap near the top for easy adjustment.
On the left side of the cooler are the wine bottle capacitors. They are staggered to allow maximum distance between them. A thin board was mounted in the cooler to separate the bottles from the rest of the circuit. Then, filler foam spray was sprayed into that section to secure the bottles. I erroneously used the expanding type which grew to cover the bottles almost to the top overnight, but they are now very secure, even upside
down. The non-expanding type is preferred since it is much more predictable.
Using the foam to mount the bottles is probably the most secure method since they are glued in, but this would not be good if one of the capacitors had to be replaced (which I would not foresee since they appear quite reliable.) With great effort, a wine bottle could be pulled out of the foam, but the most "breakdown friendly" method would be to cut holes in hard block foam to accept the bottles and then place a three-hole board over them to secure.
Holes were drilled through the ends of the top of the cooler to pass the tube leads through. This allows the top to open and the leads to ride along. Ensure all wires are of adequate length to accomodate the top being open when everything's connected. To be safe, zip tie wires of equal potential to each other and away from wires of unequal potential.