Home p103.html

 High Voltage Generator Subsystem Details  

(This subsystem headed by Colin Joye)
  • 12/1/01S: We hooked up the high voltage circuit to a 100W lamp (for current limiting) and turned up the power. We did generate a corona discharge, approximately 2cm in height. It looks like a palm tree now with spark-like flames pointing in random directions, rather than one coherent candle-flame like flame. We are trying to understand why, but one theory is that the becuase we lowered the frequency by a factor of 10 when switching over from tube to FET, and the air looks like a capacitor to ground, the air reactance increased by a factor of ~10.
  • 6/2/01: After having constructed two different coils, coil#1 wound with 20ga magnet wire and rather tight winding spacing; and coil#2 wound with 16ga. magnet wire and losser winding spacing, we tested these coils on our HP microwave tester and found some interesting and very significant results. First, the resistance at resonance Rac, was much higher for coil#2 (370k ohms) than coil#1 (62k ohms), and resulted in a much stronger flame with much less wasted power dissipation from the vacuum tube-driven high voltage generator. The Rac in effect hinders the Q-factor from reaching infinity. The Q-factor arises due to the coil being in self-resonance. We will be designing several coils and testing them in order to find out how to maximize the Rac and minimize the Rdc, which prevents the coil from receiving its excitation current. I plan to obtain the thickest solid copper wire I can in order to drive down Rdc and hopefully raise Rac.
These maximum impedances for the coils are not reactive, so this is just leading to wasted power. The circuit model below shows a simplified circuit of what the inductor possibly looks like at resonance. The resistance in series with the inductor is what we are trying to minimize in order to get a higher Q-factor.

Circuit model of inductor
Output Coils
  • ---------- Inductor Measurements -----------
  • Inductor #1 noticeably works best.
  • Inductor #1 plot: Xicon 100uH, 1A
    RMax= 62kohms, Xmax= 52kohms, f0= 3.17-3.18MHz.
  • Inductor #2 plot: Radio Shack 100uH, 2A
    RMax= 110kohms, Xmax= 35kohms, f0= 8.7-9.1MHz.
  • Inductor #3 plot: [unknown] 100uH, 1.2A
    RMax= 5.8kohms, Xmax= 3.7kohms, f0= 2.8MHz.
    log:
  • 11/2/01: Try building a lower frequency coil for ~3MHz to avoid all bandwidth problems and make driving the FET easier.
  • 10/01: Testing TI gate driver devices, the TPS2828 and TPS2829. Phenominal specs designed to drive FETs. Cheaper, smaller and better than what we could possibly design.
  • 9/24/01M: Performed gate drive simulations using ADS with 2n2222 model.
  • 9/22/01S: Got tube-drive high voltage generator to produce corona using coil#6 at ~300vdc. The coil didn't seem to heat up too much. The corona was not self-standing (required a screwdriver near by). Tried a Radio Shack 2A 100uH inductor in place of the Xicon 1A 100uH, but it severly hindered the flame. Two Xicons in series or parallel didn't have much of an effect over just one Xicon.
  • 9/20/01R: Started simulation of gate drive circuit using HP's Advanced Developement System (ADS) as the simulator.
  • 9/18/01: Received parts for tube-driven oscillator that will be used to pre-test the coil and for other research (and to have as backup in case the FET doesn't operate).

Top