DIY studio strobe

Fairchild has a application note on small IGBT controlled strobes.

Steve

If you know about MOSFETs and bipolar transistors, you and think of an IGBT as an N channel MOSFET feeding a PNP transistor.

You need to apply a pulse to the gate of the IGBT that is about 10V tall and lasts for as long as you want the current to flow. The gate of the IGBT must be moved from near zero to the near 10V and back quickly. Holding it high and low is the easy part.

The IGBT must be in the path of the current through the tube. You can't control the tube via the trigger electrode once it is fired. The plasma inside the tube is a short circuit. You need to open one end connection of the tube by turning off the IGBT in that path.

Thank you both for your comments! Is an IGBT the only solution to achieve my goal? What other alternative do I have, should I want to reach currents higher than say 150A?

--
Not trying to be difficult, but are you sure you understand the comments
stating that in order to stop the discharge through the tube you must
interpose a switch between either the high or the low end of the tube
and what that end is connected to?

JF

One of my early patents...

3,496,411 Timing Circuit for a Flash Camera (PDF on my website)

One of those rare and amazing situations where an engineer in another department at Motorola came by and asked a lot of questions. He went away and tried my suggestions... they worked... and he told the patent attorney I was the inventor :-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

--
I've always liked Motorola.

ISTM they've always been pretty straight-up kind of folks and your
experience adds credence to my belief.

Do you remember when they used to make capacitors?

JF

I didn't know that. I was only at Motorola from 1962-1970. (Though I did do later chip designs for both Motorola and ON-Semi... as a contractor.)) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

I have heard of GTO ("Gate TurnOff") SCRs being used in applications like these. I have also heard of MOSFETs and IGBTs being used for this.

One fault of IGBTs that I have heard in a bit of older literature (mid

1980's or a few years older) is that they, "in my words", are "latched-conductive" when current exceeds some amount.

MOSFETs have a different fault - their "on resistance" varying directly with temperature, to extent of being roughly or even over double that of Tj=25 C (which the nominal figure usually is) at the highest Tj at which they are rated to operate continuously.

- Don Klipstein ( snipped-for-privacy@misty.com)

Ever heard of MCT's? MOS triggered SCRs.

Apparently important for big power, but I've never seen one for sale anywhere (and come on, I can buy a megawatt SCR from Digikey right now).

All the manufacturers are saying that's fixed now. Collector characteristics often go up to 300A+ on a 50A device. Then again, I get concerned when they *don't*.

Makes me wonder about this 1200V 50A IGBT module I have... it's kind of old, date code something in the 80s. Not sure I can trust it for actual duty, not in my hands... (Not that I have any 480V three phase in the house anyway!)

Plus being approximately resistive (and rising at higher currents), MOSFETs are quite poorly suited to a pulsed power application like this. In fact, the deliciously low Vce(sat) of an IGBT, even at high peak currents, makes them particularly well suited to this application.

What else is there to reflect on... BJTs? Imagine the darlington you'd need to switch a flash tube. ;-)

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

Wouldn't it be easier to fire an SCR in parallel to the tube to eat up the remaining charge? No floating grounds, can use the normal ignition transformer. ciao Ban Apricale, Italy

That is a possibility, but I may adapt the design to be power-pack based, so that I would like spare electrons! Using a 2 SCR design, the storage capacitor would be emptied completely each time. Thank you anyway for highlighting this relevant point!

Of course I do, but as you may have understood electronics is not a field I am heavily trained with and thus I call for readers'know-how. The concept is crystal clear, yet I have no perfect idea about the practical limit of the discussed devices.

You do have other options.

(1) If you know before hand how much power you want in this pulse, you can connect only enough storage capacitance to give that flash. This could be done by having IGBTs, SCRs or bipolars switching in and out the capacitors before the shot is taken.

(2) You can provide a second discharge path for the capacitor that takes away some of the energy and doesn't let it go towards the flash.

(3) You can use some sort of semiconductor device to block the current in the tube. It doesn't need to fully block it It could have a capacitor across it so that it doesn't have to take the full current as it first turns off.

(4) You can put a mechanical obstruction in front of the flash to block some of the light.

e

... or short out the tube, but good luck with that.

See...

Patent 3,496,411, "Timing Circuit for a Flash Camera"

on my website. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

On Jan 29, 11:11=A0pm, "Tim Williams" wrote: [... photoflash ...]

need

To reduce losses, you wouldn't want a darlington's extra drop. I am thinking of something more like this:

+3V ------------+------- ! ! !/ e Q1 ! logic -/\/\---! PNP ! !\ c !Q2 ! !/ Q3 +-----!NPN ----+-------+-Tube ! !\e !/ ! ! \ +--/\/\-! =3D=3D=3D C1 [R] High voltage / ! !\e ! ! resistor \ \ ---+-------+-GND ! / GND \ ! GND

If Q3 needs to pass 150A and has a HFE well above 30, Q2 only needs to take perhaps 5A from the +3V rail.

Q1 would only need to pass perhaps 100mA. This puts the Base drive of Q1 well within the range of 5V 74HCxx logic.

C1 keeps the voltage on Q3 from rising until the base charge has been removed. The high voltage resistor discharges C1 between shots.

$0$5384$ snipped-for-privacy@news.free.fr...

e

ve

nts

e

You need to put a resistor in series with that SCR. Resistors rated for that sort of duty can be a bit hard to come by. The power in that pulse can be very high even though it is very short. For a home project, I would most likely make my own resistor out of copper wire. Copper is soft and thermally conductive enough that the heat pulse doesn't cause a lot of stress.

On a sunny day (Sat, 30 Jan 2010 08:20:47 -0800 (PST)) it happened MooseFET wrote in :

I think the normal procedure is to have a small inductor in series with the flash bulb, to lengthen the flash, the lifetime of the capacitor, and to reduce the maximum current...

HV charge current >--------- L1 ------------------------------- | | | | { } Flash bulb | === | | --- C1 | / early discharge switch. | / fire switch | | | | GND -----------------------------------------------------------

[about turning OFF a discharge lamp ] e

Exactly correct. If you put a small choke in series with the tube, the SCR needn't conduct high currents for long, either, to quench the arc while leaving the power supply full of electrons. SCRs, however, have a dI/dt limit (they can't fire quickly) so you don't want to use inexpensive phase-control devices for this. OptoSCRs have the best dI/dt performance.

The series switch idea, while good, has risetime limits (if you want one microsecond control, the turnoff will put hundreds of V/us on the switch). Miller effect makes this a problem.