Power Supplies

[mixedgas]

Unless some portion of the PSU sensing and control system is grounded on the output side, generally there is no issue reversing the polarity. Only folks who would know are the folks that engineered and configured the cap charger. Some times that is mentioned in the manual, some times it is not.

I like dedicated cap chargers as opposed to HV psus. Cap Chargers generally have a circuit that allows faster, more optimal, charging with far less wasted energy by changing the switching frequency in the PSU as the cap charges.

I would switch to series injection triggering if I were you. External field triggering is so out of date. Some times it is needed because of lamp cooling issues etc. Series injection allows two series lamps to run off a given cap bank, as well.

Without pictures and seeing the system, its difficult to comment outside of generalities.

Steve

[planters]

Unless some portion of the PSU sensing and control system is grounded on the output side, generally there is no issue reversing the polarity. Only folks who would know are the folks that engineered and configured the cap charger. Some times that is mentioned in the manual, some times it is not.

Panasonic. Reassigned microwave oven, inverter power supplies. The current is controlled by the duty cycle of the PWM %. I think this will be much better than the big linear system I am using now. I just have to get a 220hz variable PW control circuit running.

I would switch to series injection triggering if I were you. External field triggering is so out of date. Some times it is needed because of lamp cooling issues etc. Series injection allows two series lamps to run off a given cap bank, as well.

I actually moved away from series injection to what seems an anachronistic triggering system. I elected to use the external triggering so as to make the lamps themselves the HV, HC switch. This reduces the rather significant inductance that the secondary coil of a series transformer introduces. I also decided to operate each lamp with its own capacitor bank to avoid the increased impedance associated with operating the lamps in series. This impedance could be countered by going to very high discharge voltages and therefore low capacitance (maintaining the LC product), but we're talking 10's of kV's @ kw's for the power supply. Also, the interesting thing about these lamps is that a single lamp will spontaneously break down at about 2,200V, but two lamps in series will break down at about 2,800V. So, with multiple lamps in series, the switch would have to be a large, triggered SG or Thyratron that could hold back nearly the full discharge voltage and survive well over 10 kA. That gets expensive.

After reviewing the literature, where much of the criticism of external triggering seems to come from is jitter and poor reliability. However, I think I've discovered an interesting phenomenon. Pulse rise time and duration is the result of the increasing density of ionized charge carriers vs the decreasing drive voltage of the capacitor bank as the pulse progresses. As the external trigger pulse becomes more and more vigorous, the main pulse rise and duration decreases. This makes sense because the trigger itself introduces charge carriers and the more of these that are present before the capacitor voltage drops, the higher the average current. This is a lot like the pre-pulse technique without the simmer drive. Eight joules at 55kV is a REALLY HIGH trigger energy and I am considering increasing it further. The jitter is low enough that the light output from one lamp has the same duration as from four lamps and I have yet to see a dropped pulse. I have a laser threshold at about 120J and above 1200J I'm now struggling with damage to the cavity optics. Suca pRoblem? It's like being too thin or too rich. In the words of Jeremy Clarkson "more powwwwwwer" I miss him, but I digress.

I'm happy with the discharge circuit with the exception that the power supply needs to be more efficient to allow a higher rep rate.

[mixedgas]

One end of the Uwave suppy is probably grounded. Do you have a model number, some of the Magnetron Driver schematics are on line.
Commercial cap chargers tend to be very efficient and faster then a raw voltage source, just something to think about.

Steve

[planters]

Commercial cap chargers tend to be very efficient and faster then a raw voltage source, just something to think about.

I don't have a line on any cap chargers that might be close to the need. 5,000V @3-6 kW. The power supplies I'm looking at are the Panasonic #220943086526.

[lightlinked]

is there a spec sheet for the tubes? how are they wired? 2 parallel? do you need them to all go off exactly at once?

[planters]

These were custom made. Four tubes are wired in parallel. Each tube is attached to its own 10kV, 50uF capacitor. They have to discharge within a usec or two of each other otherwise the envelope of the light pulse will be widened.