DRSSTC efficiency experiments

I just got hold of a nice wattmeter that reads 0-1250w at 240V AC. Since it is the old dynamometer type with air-cored coils and no electronics, I expect it to read fine on highly distorted line current, and not be bothered by RF from a Tesla coil.

*Wattmeter explained

If you're under 40, you've probably never seen a dynamometer wattmeter. I hadn't until I got this one off eBay for 12. It's a simple mechanism consisting of two coils. The load current flows through one coil with a few turns of heavy wire, and the line voltage is placed across the other coil, which has hundreds of turns of thin wire and a series resistor. The magnetic force developed between the two coils moves the pointer against a spring.


Wattmeter under test. You multiply the reading by 5 when running off 240v

Once I had checked it using a true RMS multimeter and a resistive load, I hooked it up to measure the input power to my DRSSTC. I decided to try generating the same size of spark using various breakrates and burst lengths, and see how efficient each was.


The "lab" ready for experiments

Experimental procedure

For my first experiment, I used breakrates of 50, 100, 150 and 200Hz. I kept the burst length constant at 150us, and adjusted the DC link voltage so the spark was 24" at each breakrate. When I say it "was 24 inches" I mean that it hit a target 24" away from the end of the breakout point about once every 5 seconds. The target is the sheet of tinfoil hanging over the back of a chair in the picture above.

I noted the DC link voltage and wattmeter reading for each breakrate. At higher breakrates, it seemed to strike either not at all, or very frequently. The wattmeter needle kicked up with each strike to the target so I had to "average by eye".

For the next experiment, I used the same four breakrates. This time, I held the DC link voltage constant at 250V, and adjusted the burst length to keep the spark at 24".

Results


Table of results


Results plotted by bang energy


Results plotted by power

Here is the MS Excel spreadsheet (Sorry, *nix users! Hope you have OpenOffice :P)

Analysis and discussion

I think this demonstrates the following trends:

Short bursts of high peak power seem to grow sparks more efficiently than long bursts of low peak power (but the same energy). It was already known, for instance, that a 150us burst from a DRSSTC would make a more efficient spark than a 10ms one from a plain SSTC, but this is the first demonstration that, say, a 100us burst works better than a 150us one.

The difference between short and long bursts is hardly noticeable at low breakrates. But streamers made from short bursts seem to coalesce (reach their full length) at a higher breakrate. This may explain the "gas burner" effect on spark-gap coils, which have shorter and more powerful bursts compared to DRSSTCs (but see below)

Evidence for the above: See graph 2. The pink trace levels off at 150bps whereas the blue one had already levelled off at 100bps. The point where the graph levels off suggests that coalescence is complete and increasing the breakrate any further won't buy you bigger sparks.

The results don't say anything definite about the corollary: that short high-power bursts are LESS efficient at promoting streamer growth at LOW breakrates. There is a trend in this direction but it is smaller than the experimental errors. So the explanation for the gas burner effect is not very well supported.

These trends are in general agreement with Steve Ward's DRSSTC work: he found that using a lower impedance tank circuit to draw higher peak power from the inverter, and shortening the burst length to maintain the same bang energy as before, gave longer sparks.

Back to Tesla page