fluorescent camping lanterns, inverters, and regulation for DC converters.

[Opcom]

Needing a small inverter transformer for a laser power supply, also maybe being a little lazy, I have been going through old photoflash units, etc, looking for one that could take a doubler. They are just too small and the best one takes a whole minute to charge a 39uF cap to 900V.

The storage is 15J, so, I dug into some of those two-tube fluorescent lanterns. Those have two F6T5 lamps, so they ought to be 12 Watts total. The lamps are also in series so the voltage should be in the range.

I got a surprise pulling them apart (non destructively). It's not clear from the outside which were US made and which were Chinese. When opened, the difference is huge. The US made one has a ferrite transformer core about 1.25x1", and 1/4" thick and two 1A transistors in push pull. It has three windings as well with the center tapped primary being many small stranded wires, and it has an RFC in series with the battery. The Chinese ones have an iron laminated core transformer about the size of a 500mW transistor radio output transformer. They also use just one transistor and have one tapped winding. They all work and the US one is brighter. I haven't made any measurements yet but I usually don't see such cheaply made junk as these ones not made in the USA. I would rather pay 3x for a good one. It's just too bad globally that things are not made the way they should be. These are old and it's not likely any are still made in the USA. The good one is a Ray-o-vac Workhorse WH8.

I hope the USA made one can charge it in 1-2 seconds and I hate to part out the lantern so the tests will be done before ruining it.
Might need a soft start on the cap so the oscillation can start, but that goes for all self excited inverters.

The big deal might be regulating the voltage somehow. It has to be efficiently done, so when the cap is at 800-900V, the inverter has to shut off. If a high-megohm divider is used with a neon lamp, that could trigger a shutdown and give some hysteresis. It would be nice to figure out a way to cut off the osc. and as the voltage decays via the resistor, it will occasionally burp and bring the voltage back up. I'll have to draw out the schematic.

Anyway it's a good education in hackery for building stuff around here. In the last few years I've built a couple of 12V-powered switching supplies for my tube gear, one for DC filament voltage on 6146's and another for grid bias.

[Opcom]

Came across this ancient topic and to close it with an answer in case others need a low current HV source, the small 120V 100W 60Hz inverter of the sort that plugs into the cigar lighter socket was able to charge up the capacitor easily when a voltage multiplier was used. These have always about the same peak voltage with the duty cycle of the 'modified sine wave' varying with load. This is much better working than photoflash units and hacked battery type fluorescent lamps. Some inverters need close to 12V, will shut down under 10.5 or over 14. So, no compact 9V batteries.. although two could be put in series and a buck converter used to make a regulated 13V. When the big cap is empty, the inverter will be called upon for a surge of about 800mA peak or 70W.