Copper Vapor laser switchmode PSU opinions

[DrMario]

Alright, since I have couple high-ampere IGBTs lying around, along with PSU-related chips, I figured I would ask here for the opinion on high-voltage step-up SMPS for the eventual Copper Vapor laser I may assemble at some point.

So, which output driver stages would be preferable? I do have few options, as I haven't started on KiCAD yet. 1. Magnetic compressor IGBT switcher. 2. DC RF generator (transmitted via coaxial high voltage cable, to the hollow cathode electrode inside Copper Vapor laser). I may not use heater, instead using Helium (or Neon) discharge to the advantage while the laser tube is to be put in the ceramic casing to trap the heat, keeping the Copper halide hot enough to be separated by high voltage pulse.

Any opinions and ideas are welcome as I am figuring out how I like it assembled.

 

[DrMario]

Also, I think I should add few more things. The PSU is planned to be of programmable variety, thus adjustable for any metal halide lasers. I also will construct the Copper laser tube from scratch eventually also.

The PSU will take in 12 Volts at 20 - 30 Amps, at least to be determined when I get around to build one. The SMPS controller may be an ARM microcontroller running a modified version of FreeRTOS for inverter system monitoring and laser behavior monitoring (current consumption, light output and temperature), while tailoring transformer voltage to the laser tube being run. I may as well as go with either H-bridge, ZVS, or flyback for the main transformer to obtain the desired voltage. Still exploring the options.

I would want to hear what the experienced SMPS designers has to say.

 

[Sigurthr]

Are you planning on running it all off of the ~360W psu input? IIRC CVL's need close to a kW of heat input just to maintain lasing temp. I remember seeing one where it was heated via a natural gas mains.

 

[DrMario]

I am going to run it off PC ATX power supply which is a-plentiful. Also, some Copper halides doesn't require high temperature, FYI, and the one that you mention being heated by Natural Gas is of Copper Iodide which requires 500 F to melt in the buffer gas. Chloride and/or Bromide disassociate at lower temperature easily obtainable by Helium discharge which can get hot.

Plus, 12 Volts input's ampere consumption will trace with high voltage, meaning if I wanna 1 or 2 Amps of 10kV, that may be 50 - 70 Amps of 12 V, and I will have to shove that kind of power if necessary (the worst case scenario may be that I might have to build A123 26650 Li : FePO4 battry pack so it can provide this kind of power).

 

[Cyparagon]

... if I wanna 1 or 2 Amps of 10kV, that may be 50 - 70 Amps of 12 V...

Uh... would you mind doing the math on that for us? I don't think I follow.

 

[DrMario]

The ampere consumption is to be determined at the time I start firing the PSU up. Yet, sometime PSU consumption math is not "one size fits all", due to several important factors like primary transformer winding impedances (both Ohm and Henry altogether at given AC frequency which will be pretty high), IGBT switching loss (Rds-on - resistance of Source - Drain at power-on), and finally the operating frequency.

I may also need to either find or make the high frequency ammeter (DMM doesn't cut it unless there are capacitors at DC power filtration block immediately before the IGBT switching stage), so I would be able to see how much amps the PSU is drawing as I increase ampere from secondary winding (which will be just long Copper strip wrapping around the transformer bobbin like some high efficiency inverters) - the output ampere would be monitored by microcontroller making sure everything's good.