Testing of Full Cylinder, Dual Dielectric e-field Thruster 1 - Failed, arced at 8kV

October 25, 2003 - First test failed, arcing at 8kV

The test setup

View of everything. The power supply is on the left side of the table. The long horizontal black rod is the HV probe.	From left to right on the table is the control unit for the digital scale, ammeter, oscilloscope and the control unit for the power supply. Note that the control unit for the digital scale is completely enclosed in fine wire mesh (window fly screen) which is grounded.
The pan for the digital scale has a polyethylene cover on it followed by a grounded fine wire mesh. The scale is also sitting on another grounded fine wire mesh. Note that the two meshes are not in physical contact.	The thruster sitting on a glass tower which is sitting on the scale.
A view showing the wires sticking out horizontally so as to minimize any lift effect that the wires might produce. Note that two tests were done, one with the thruster as shown and the other with the thruster flipped 180 degrees such that the side that is "up" in the picture below was "down" instead.

The test - failed due to arcing inside the thruster

As the scope reading clearly shows below, there was arcing inside the thruster meaning that there is one or more air cavities either inside the thruster tube or at one end. A sanity check was done by replacing the thruster with a gravitator. The gravitator permitted voltage to increase to as much as 50kV with no arcing showing on the scope.

At all voltages, a weight gain was showing on the scale. This was true regardless of which side of the thruster was "up". The scale did show an unbalanced condition as long as the weight gain was showing. The weight gain increased as the voltage increased.

No matter how high the voltage was turned up (though it was never turned up very high) the scope showed that the voltage at the probe did not exceed 8kV until it arced. This is the spike on the scope picture below. Voltage then built up again as shown to the right of center. Note that as voltage was turned up, the 8kV maximum still held but the spike was larger.

At lowest power supply setting the ammeter read around 30-40 microamps. Note that the current was probably present mostly during the arc as the meter was making a rapid fire clicking sound. The needle appeared steady probably because the internal mechanism did not have time to move before an arc occured again. Note that as voltage was turned up, the amp reading also increased.

At Zoltan's suggestion I did some tests to eliminate the weight gain that was showing up on the scale by improving the sheilding on the scale. The following elimination tests were done:

1. The scale unit that the thruster sits on was completely wrapped in aluminum foil as shown in the pictures below. First this was tested without shielding the cable too. The weight gain was the same.
2. A simple idea came to mind. I removed the glass jar that the thruster was sitting on and instead put a milk crate upside down on top of the round table that the scale unit is sitting on. Note that the milk crate was NOT sitting on the scale. I then put the thruster on top of the crate. The weight gain was the same even though there was no way for the thruster to affect the weight on the scale.
3. I removed the milk crate and put the glass jar back on the scale. I then replaced the thruster with the gravitator (which does not arc). There was no weight gain. The scale remained at 0.0 grams. This is a very good sign as it indicates that when we do get the thruster to not arc, it will likely have no affect on the scale.
4. I put away the gravitator and put the thruster back in place. I then shielded the cable so that now the entire digital scale -- the control unit on the table, the cable, and the scale unit that the thruster sits on -- was all grounded. There was no weight gain. The scale remainged at 0.0 grams
5. To prove that step 4 above was not caused by the fact that the scale was completely wrapped in aluminum foil and that the foil was somehow physically preventing a weight change from showing, I put a single playing card on the thruster. The scale immediately showed a weight gain of 1.3 grams.

The completely sheilded digital scale after step 4 above.	A closer view of the scale unit completely wrapped in aluminum foil.
Putting the single playing card on the thruster showed a weight gain of 1.3 grams. See step 5 above.

The conclusion is that the weight gain showing up on the scale's control unit's LCD panel was due to the arcing in the thruster affecting the electronics in the digital scale and not due to any actual weight gain in the thruster. Step 3 above gives hope that not a lot of grounded sheilding will be needed on the scale once we do get the thruster to not arc. However, a force should have been measured despite the arcing at 8kV. This indicates that this device will not produce thrust.

October 26-November 2, 2003 - Attempts at fixing problems and balance test

I dug cut off the ends of the wax and then dug out the interior of the thruster. After a few hours of digging I found an easier way of removing the wax from inside the thruster by heating two metal skewers (long metal sticks) and pushing them into the wax, thereby melting it out.

Wax had gotten into one of the plaster endcaps and broken it so I replaced it with a balsa wood one instead (see picture below).

I then put an endcap of wax on one end and gradually filled the interior of the thruster (see picture below). After letting each pouring harden, I tested it to see if it arced inside the new wax. Instead of arcing inside the wax, it would arc in a section not yet filled with wax. I would then add a little more wax and repeat the testing. After filling it about 75% of the way, the wax started arcing internally. So I removed a bit of wax and tested again. I continued doing this and it kept arcing internally until I'd dug it all out again! At that point, I ran out of time.

My guess is that during the initial repouring I had the thruster interior very well cleaned out so it arced easily in some places where there was no wax instead of arcing in any cavities in the parts that I had already poured wax. During the gradual pouring, the interior walls of the thruster were gradually getting coated with wax until finally the areas that were not completely filled with wax would no longer arc and instead arcing deeper down would start to occur.

I did a test with a triple beam balance to see if there was any weight change (see picture below). Note that it arced during this test as before. There was no weight change. And yet if I gently put a single playing card on the thruster while it was on the balance, the balance easily showed a weight increase.

The new wooden endcap.	Gradually pouring wax.
The triple beam balance test.