> > To fire the lamp 20 times per second one needs a trigger circuit > > without a charge capacitor > > I don't see it. The typical disposable flash trigger cap is something > like 0.022 uF charged through a 1M resistor, so it ought to charge > just fine in less than 1/20 second as long as the flash tube doesn't > completely dump the HV reservoir cap during the first flash. The rc circuit wastes energy and the trigger element is not a thyristor, it's a fet, so there is no point in having an rc circuit in the first place, the trigger coil is fed directly from +Vb. Also, some trigger circuits can apply more than one pulse per triggering to ensure that a hot lamp is triggered promptly. A thyristor or RC circuit cannot do that easily. > wasting the rest of the charge.) I can't comment on the early units, > but like others have said, I believe IGBTs are commonly used in modern > units. (flash controller chips frequently include an IGBT driver, > whatever that means.) It is possible that IGBTs are used in very new units but the book I was quoting, and my experience, goes back to the 1980s when mosfets and bipolar transistors were used. At the time IGBTs were onion sized and expensive. > > Current control cannot be used with a plasma arc tube, the rise time > > is on the order of a few nanoseconds > > That doesn't match my observation of cheap flash tubes (discharging > them through various other components...) A TRIGGERED flash tube may The purpose of the flash is to achieve a certain light intensity and color temperature. This is not achieved if the current is limited, and the lamp maker specifies a very specific Vb voltage range for operation. The pre-flash used by cameras adjusts the exposure and other parameters so it is useful for it to use the full current but short pulses. None of the circuits I know (dozens) use current control, many use voltage control in a few steps and all use bottom switching of the flash tube. I.e. the current is set by setting Vb for a given lamp, and never changed during firing. > If anyone finds a simple circuit for a flash capable of the "blip blip > blip FLASH" sort of action, I'd like to see it, too! Connecting the tube directly after a rectified and lightly filtered output of a 300Vdc output 50Watt, primary current (!) limited, smpsu should work. Continuous triggering can be achieved using a continuous RF source (mini ionizer or tesla coil). High pressure arc lights (excepting for the additional periodic polarity reversing circuit) and welding equipment work the same, I'm sure a schematic could be found. A 60Hz stroboscope scheme I saw used a neon bulb to fire a flash tube using a thyristor every 2nd half wave of the mains ac. The xenon tube had its own voltage doubler or tripler rectifier and a largish charging resistor for the pulse capacitor. This is not efficient but it works. The total power into the lamp must not exceed its ratings, derated for continous operation. Eg a 50Ws tube can be run at 5 Watts at 20Hz probably. For 20Hz and 1msec pulses and 300V Vb that's 20 1msec 5A pulses per second, each consisting of a trigger pulse and an off pulse. Both can be controlled by the same single output (derive the trigger pulse from the rising edge, and use the high level to keep the cathode switch on, then it turns off at the end of the pulse). With this a garden variety mosfet with Rdson of about 3 ohms can be used as a cathode switch, at negligible losses. Replacing the trigger circuit with a microprocessor would probably do what you need. My work was tied more to specialized flash systems and less to stroboscopes. Anyway many modern quality flashes can be programmed into strobo mode, typically 3 to 10 flashes. Here is a (semi?) professional flashes datasheet that has strobo mode: http://www.metz.de/fileadmin/fm-dam/Download/Homepage_Englisch\ /Photo_Electronic/Bedienungsanleitung_GB/mecablitz_GB/mecablitz_40_bis_45\ /mecablitz_40_MZ_3i_GB.pdf (paste all on 1 line without the trailing '\'s) (search for strobo inside the pdf). There are others, but Mecablitzes are one of the best imho. Note that the technical data table clearly says that the flash duration goes down with diminishing light output (p.30), from 1/200 down to a respectable 1/26000, in direct proportion with the light output. Also note that the technical data is clear and complete and not in chinglish. Peter -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist