I need an HV fast sqaure wave generator. / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

I need an HV fast sqaure wave generator.

first previous 2 3 4 5 next last

Move Thread

LAN_403

Proud Mary

Thu Oct 20 2011, 09:53PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Patrick wrote ...

Proud Mary wrote ...

If you Google 2N5551 avalanche pulse you'll at once find a number of papers detailing the use of this very inexpensive HV transistor in simple nanosecond and sub-nanosecond pulse generators with outputs up to 3kV from series strings.

Physical layout is the ultimate determinant of speed in the low nanosecond regime.

There is also a useful application note for the more expensive ZTX415 avalanche transistor giving other fast pulse generator circuits with PFNs which you could use with 2N5551:

Im thinking of using this circuit.

I haven't encountered the difficulties suggested by others, and have just used an ordinary 3kV variable bench PSU, and 555 as trigger source, as described here:

Patrick

Thu Oct 20 2011, 11:12PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Is this a good choice? 3.10 USD for 100 pcs.

I was thinking the TO-92 case type.

Please reply soon people, ive got to put in a purchase order in to mouser,

teravolt

Fri Oct 21 2011, 02:19AM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

we have used theas at work for a fiduciary generator's. I would go with a combination of figure 13 and 14. you could replace the PFN with a peace of coax. use a pcb with a ground plane and keep your leads as short as pausible.

Patrick

Fri Oct 21 2011, 04:12AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

teravolt wrote ...

we have used theas at work for a fiduciary generator's. I would go with a combination of figure 13 and 14. you could replace the PFN with a peace of coax. use a pcb with a ground plane and keep your leads as short as pausible.

Ok im looking at the details now...

Proud Mary

Fri Oct 21 2011, 08:31AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Patrick wrote ...

Is this a good choice? 3.10 USD for 100 pcs.

I was thinking the TO-92 case type.

Please reply soon people, ive got to put in a purchase order in to mouser,

As good as any, Patrick, a fair price, and plenty of spares.

Pinky's Brain

Fri Oct 21 2011, 08:55PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

Patrick wrote ...
what devicewould you use then ?

Not so much device, I suggest an entirely different circuit ... having thought about it I don't think existing half bridge drivers would work well ... maybe something like this :

HVblock

Or to describe it in words, have a square wave generator, create the complement of the square wave too, delay the rising edges of both with a L-C-D circuit (for break before make), clean the edges up with Schmitt triggers, use totem poles to drive several GDTs to drive several MOSFETs in series stacks forming a half bridge.

PS. as I said before this circuit is not suited for driving low impedance loads, any mismatch in turn on of the MOSFETs on a leg will result in avalanching ... but the HV divider is very high impedance limiting the avalanche energy, so it doesn't matter here.

Weston

Sun Oct 23 2011, 04:31AM

Registered Member #1316 Joined: Thu Feb 14 2008, 03:35AM
Location: Cambridge, MA
Posts: 365

I took the idea of using a delay line from the papers on avalanche transistors and implemented it on my original setup.

Lacking a very long (and expensive) section of coax cable, I used a single section from a long bundle of cat5 cable. As cat5 cable has a impedance of 100 ohms, I had to change the terminating resistor. Here is a new scope shot into a 100 ohm load. Current transformer is 100ma/v.

The pulse is probably too short for your application. But its all a matter of finding a long section of cat5 cable.

Pinky's Brain

Sun Oct 23 2011, 10:35AM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

Looking around, making a GDT with 100 kHz (for the actual switching) to 10 MHz (for the edge transition speed) bandwidth is pretty much on the edge what is technologically possible ... you would probably need a slightly more complex gate drive which is edge rather than level triggered so you can use a low winding count HF pulse transformer :/

Something like this.

PS. getting a 1/100000/2 second flat top pulse out of a cable PFN is not going to happen :) That's over a 1000 meter of cable.

PPS. not to mention that the amount of energy when putting it into an impedance matched load at a couple of kV is going to be huge with twisted pair or coax ... you're going to be burning 100s of kW.

Patrick

Sun Oct 23 2011, 10:17PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Im also wondering what the output of a flyback looks like on an oscope.

Pinky's Brain

Sun Oct 23 2011, 10:27PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

Ignoring all the parasitics for a moment you can treat it as an inductor with a current going through it (the primary current at switch off divided by the turn ratio). The voltage output depends on whatever is behind the flyback.

first previous 2 3 4 5 next last

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