Switching HV with IGBT's? / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Switching HV with IGBT's?

1 2 3 next

Move Thread

LAN_403

codesuidae

Wed Mar 22 2006, 03:10AM

Registered Member #337 Joined: Wed Mar 22 2006, 02:52AM
Location:
Posts: 5

Greetings,

I'm investigating the possability of switching about 15kV with IGBT's for the purpose of powering CO2 lasers.

I've seen articles about using many IGBT's in series to handle higher voltages. I'm hoping that very cheap (~$1) devices can be used for this purpose. Only small currents need to be switched but it is also necessary to run the switch at several kHz for PWM control of the laser power.

Is this idea practical?
Thanks

Desmogod

Wed Mar 22 2006, 06:29AM

Registered Member #139 Joined: Sat Feb 11 2006, 11:01AM
Location: Perth, Western Australia
Posts: 358

I think the problems with parralleling semi's is that there can be a propogation delay between them.
i.e. 30nS delay between all of them lets one open fully before the others, letting the majority of the current through it.
I'm sure I have this wrong, or marginally correct. (Someone will give you the FULL reasons and workarounds I'm sure!)

ragnar

Wed Mar 22 2006, 08:48AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Yup, Jim is totally wrong, not even marginally correct. He meant to say "seriesing" and "voltage" =P

TDU has worked with four hockey-puck SCRs in series... he might have some valuable words from his experience. Turn-on time is critical, but how many kHz? Can't be that hard.. =)

Isolation of the gate drive will be.. 'interesting'.

dan

Wed Mar 22 2006, 08:50AM

Registered Member #223 Joined: Mon Feb 20 2006, 06:42PM
Location: Ottawa, Ontario, Canada
Posts: 125

It may be possible but they all need to turn on at the same time which isn't that easy. But why do you want to go through the trouble of interrupting the power on the HV side. I mean you could just interrupt the power to the HV supply to vary the power that is fed to the laser. Or even better power the tube with a rectified high frequency transformer and adjust the PWM on the drive circuitry to vary the power to the laser.

Sulaiman

Wed Mar 22 2006, 10:14AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

If you do make a hv string of IGBTs to handle 15 kV, e.g. using 1200V IGBTs;
-you will need say 15 'stages', each stage with an IGBT and across it a 1000 V zener/tvs
(e.g. I would use 3 x 350 V tvs because I have them)
That will ensure that no single IGBT 'sees' more voltage than it should.

Then you will have to work out how to drive the gate-emitter of each;
insulating for at least 1 kV per stage.
One way is to wind 15 x 1:1 toroidal transformers, add gate and damping resistors,
put them in series and drive one IGBT from each secondary.
This way each transformer only needs 1 kV insulation/isolation, (easy)
instead of a 'master' drive transformer requiring 15 kV isolation overall.(difficult)

If the current is very small and the frequency low, you could use a CRT as a valve!

Personally I'd desparately try to work out how to do the modulation at lower voltage.

Steve Conner

Wed Mar 22 2006, 11:32AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I once saw a stack of series'd IGBTs designed for switching pulses at 10kV. It was about 2ft tall, the drive circuitry looked pretty complex, and the guy running it complained that it blew IGBTs fairly often. (there were neon lamps on the stack so you could see which storey the bad IGBT was on.)

For an application like this, a better way would be to build a high voltage switched-mode power supply and control the power level by PWMing the low voltage side, like a few other posters have suggested. The Thor CCPS by Marco DeNicolai (or the simplified version of it by Steve Ward) would make a great CO2 laser driver.

Is this for a CNC laser cutting application?

HV Enthusiast

Wed Mar 22 2006, 12:52PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

This is actually quite common. We do it all the time, although we use MOSFETs, for grid modulators and cathode switchers on TWT tubes. Since you are not switching lots of current, then MOSFETs should be more than adequate for your application.

Marko

Wed Mar 22 2006, 01:26PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

On higher frequencies there are big problems, but it can be done. There was a trial of SSTC that would operate with higher input voltages by Steve Ward, but very soon after topology was ashamed by DRSSTCs.

For CO2 laser switching can be simply done on LV side and then voltage stepped up, you need serious ferrite transformer but it is far better idea than stacking semi's.

codesuidae

Wed Mar 22 2006, 04:05PM

Registered Member #337 Joined: Wed Mar 22 2006, 02:52AM
Location:
Posts: 5

It sounds like while IGBT's can handle it, they aren't necessary because I only need a very small current (on the order of a few tens of mA).

What considerations are there with driving the gate on series MOSFETs for this purpose?

How critical is getting them all turned on at the same time, and how does one go about getting that precise syncronization?

As an alternative to switching the 15kV I would consider switching the input side, but I'm don't yet know what the output requirements are. It seems that this would involve something a bit fancier than switching line AC to the primary. Is it practical to use a NST or MOT, or do I want to build a transformer from scratch? I have flyback cores from three 15" monitors that might be useful for that.

The idea is to use PWM to control the power output to allow for cutting of a variety of materials. I haven't worked with cutting lasers before, so I'm still learning the details that I'll need to know to set up a system.

My goal is to set up a CNC laser cutter on the cheap while learning about laser operation, control and optics. The project is more about learning than it is about a need for an effective machine, so that will determine how much I buy and how much I build.

Thanks for the input!

EEYORE

Wed Mar 22 2006, 04:46PM

Registered Member #99 Joined: Thu Feb 09 2006, 06:10PM
Location: florida, usa
Posts: 637

Couldnt you just use a variac to control power in to the laser?You mentioned a NST or MOT...Go for a NST and run it off a variac.

Matt

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob