• Getting there

    YSPACE Labs • 11/21/2023 at 23:54 • 0 comments

    I was able to get a real inductively coupled ring discharge in an E27 neon globe from Aliexpress https://www.aliexpress.us/item/3256804202033036.html ("Rose" option). The gas in this bulb is very easy to ionize, so getting a toroid is comparatively easy. Interestingly, the circuit will stop oscillations, then start again and repeat when making a ring discharge in this bulb. I think the ring puts a lot of load on the circuit, which causes oscillations to stop, and since it's unstable, oscillations start again. Adjusting the bias potentiometer actually adjusts how long the oscillator stays on (with more bias meaning longer on time up to the point where the mosfet always turns on and won't oscillate). The ZVS during the on periods is not very good, but it doesn't cause much heating since the circuit only draws 6-14W avg in this mode (could be more than 100W in pulses though). There is actually another mode where a toroid doesn't form in the gas, but instead some sort of diffuse inductively coupled plasma discharge forms. This can be triggered by lowering the power past a certain point. The oscillator will be continuously on and in very good ZVS. 

    Also copied description from a post elsewhere:

    Everything is driven from a 10MHz class E oscillator. It seems like the key to ICP is having more inductance on the work coil and less capacitance on the tank capacitor to get higher voltages on the coil-tank cap node. If ICP doesn't work, but there is oscillation and HV, decreasing gate capacitance to increase drive voltage helped. Also probably helps even if the circuit fails to oscillate. A lower resistance fet will greatly increase power output. I also killed a SiC JFET cascode because of bad zvs (at only 15v in though lol), but the IRFP250 Si mosfet has been running fine even in kind of bad zvs and processing over 100W of power.

    It seems like getting good ICP comes down to a few things:

    having high enough voltage on the tank cap (which means bigger inductor and less capacitance)

    having a snug fit around the gas tube (someone on Discord has suggested using a coil around the equator since it worked well for creating the toroid in a proper xenon flask for him)

    Having high enough oscillator power (which can be increased by decreasing fet resistance or increasing gate drive voltage (meaning lowering capacitance from gate to ground))

    This also worked (with a different work coil) for creating a small ICP "lamp" (extremely inefficient since 50W was going into the circuit for a small amount of light (most was being turned into heat in the plasma and work coil)). 

    I think getting a plasma toroid in a plasma globe is possible, but only experimentation will tell.