# HV Power supply

Discussion in 'The Projects Forum' started by eugesd, Jul 31, 2013.

1. ### eugesd Thread Starter New Member

Jul 31, 2013
5
0
I'm working off a design to build a +/- 2kV power supply

I'm using this as reference, I've simulated it in TINA and wasn't able to find the N Chan MOSFET SPICE model to correctly simulate it, so I used a generic one.

http://m.eet.com/media/1134953/15459-figure_1.pdf

This the original document, it contains some errors that I seemed to have fixed.
http://www.edn.com/design/analog/4329148/High-voltage-amplifier-uses-simplified-circuit

Could someone explain how I could ramp up this designed to handle 2kV, I only need about 1mA of current and it needs to be able to jump voltages in about 100μS. It would be great if I could put in a 5V signal and see 1000V, then put in a 10V signal and see 2kV.

Thanks!

2. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
A power supply and a power amplifier are different things.

Jump from what voltage to what voltage? That defines slew rate of the output voltage.

3. ### eugesd Thread Starter New Member

Jul 31, 2013
5
0
Oh yes! Sorry about that, I'm very concentrated on this that I forgot about that subtly. I definitely mean an amplifier.

That document I shared with you is basically what I want to accomplish. I'm just trying to increase the swing of an op amp by adding a gain stage to it using mosfets, however the design I found is only capable of 1800 v p-p and I need that to about 4000v p-p. I'm not sure what the slew rate works out to be, but basically I need to be able to get from -2000V to 2000V in 100μs.

Excuse me for being a bit unclear but I absolutely appreciate any help. Thanks!

4. ### LDC3 Active Member

Apr 27, 2013
920
160
Let's see:
4000V in 0.0001 sec is 40MV/s.

5. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
40V / us is a pretty severe requirement. Theoretically possible, but a lot faster than most op amps can do. Looks like this might have to be a discrete design.

6. ### eugesd Thread Starter New Member

Jul 31, 2013
5
0
I could sacrifice some speed, however it would be great to have it there.

Any words on how I could improve that circuit above to handle the 2kV?

7. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
I hesitate to assist in a project that could kill you. I really don't like the design at all where there are so many devices series stacked. During transients you may not get symmetric sharing of voltage and that would put too much voltage across individual components. Every one of the FETs has capacitance from source to gate, gate to source and they do cause voltages during transients. That design looks unreliable to me.

8. ### eugesd Thread Starter New Member

Jul 31, 2013
5
0
Currently, this work is conceptual and I will be simulating in a SPICE program. Once I get something down and tested thoroughly, I'll take the proper precaution to avoid an electrical safety issues.

Would you be able to suggest a reliable design that could get the job done?

Thanks for the help!

9. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
Not really. But FYI: the worst shock I ever got in my career was from a 3kV power supply for a traveling wave tube (military contract). I thought I was being safe. Knocked me across the room. I was lucky I wasn't killed because I took the shock across the chest and it could have stopped my heart. I still have an electrical heart arrythmia that is permanent.

10. ### gootee Senior Member

Apr 24, 2007
447
50
40 V/us is not very fast. There are op amps that will go at 2500 V/us. e.g. LT1818 and LT1819.

I don't know if these will help. But there are some interesting "op amp booster" circuits in the following application notes that sound similar to what you are trying to do. Maybe they will at least give you some ideas, and possibly some tips about doing it as safely as possible.

http://cds.linear.com/docs/en/application-note/an18f.pdf

http://www.ti.com/lit/an/snoa600b/snoa600b.pdf

(I like the one where they put a McIntosh tube amp inside of an op amp's feedback loop.)

11. ### JoeJester AAC Fanatic!

Apr 26, 2005
3,387
1,211
I got shocked by a capacitor charged to 20kV once. Thankfully, it only travelled from where my finger was pointing (3 inches from the capacitor) and about four inches up from my wrist that was on ground. I did have a technician secure the transmitter, but I learned that the technician didn't know the protocol to secure the transmitter. In fact I turned to the other technician and asked him if he saw where that blue arc originated ... which was the point I wanted him to measure after ensuring the capacitor was de-energized. Then I had a heart-to-heart with the mis-informed technician.

The worst shock I got was working on a AN/URC-77 transceiver. I hadn't reviewed the technical manual and the Ep was 900V and up to 300 mA for normal operations. It hung me up for about one minute or so, till I let out a MFer and pulled myself off. I later found out that the transceiver used switched Escreen to begin transmission. I was upset. I didn't read the manual because a year before some CDR saw me reading and my Chief came to me and told me not to read. Well, after that episode, when that same CDR wanted me to talk about it at a training session, I gave him a piece of my mind that he would never forget. I also directed everyone who ever worked for me to read the safety and theory of operation for every piece of equipment they were going to energize and if some officer says anything, I would take care of it. I was pretty brazen and mouthy as a PO2 ... and won more times then lost when dealing with officers.

I would find it difficult to assist someone who has demonstrated little experience with a high voltage circuit.

on edit ...

Last edited: Aug 3, 2013
12. ### gootee Senior Member

Apr 24, 2007
447
50
Yes. My first lab manual's first page indicated that 0.1 to 0.2 Amps through your heart will make it fibrillate, such that even if the current is removed, your heart will not restart normal operation on its own and you will die (unless a defibrillator is handy, and someone is there to use it, and it works). A higher current amplitude will generally "clamp" the heart muscle, and if the current is removed the heart will start beating normally again, on its own.

When my Dad was in the army, in WWII, in the Philippines, he was sweeping up between tents at a staging camp, after a storm, and picked up a downed power line with his left hand. Only one other guy was nearby, on his cot a few rows of tents away, and he came to check when he heard my Dad's whistling stop abruptly. Dad's left boot sole was melted and on fire. The guy took the broom and knocked the wire out of his hand. He was back on duty a couple of days later, with new boots.

13. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
507
There was a great episode of Married With Children where Bud Bundy was trying to fix a light and told his sister Kelly to go down into the basement and turn the circuit breakers OFF.

He gets fried by the electricity and walks up to her and says:

"Hey, Kelly... how do you spell OFF?"

She replies: "Uhhhh.... I think it's O something......."

14. ### eugesd Thread Starter New Member

Jul 31, 2013
5
0
Thanks.

I suppose I already have an idea on how to boost the amp, if you look at my original post. I'm just caught up on how I could increase the voltage while keeping the consistency of its function.

As for your ethical dilemma, I think not helping me would put me a greater danger. However, please rest assured that I have other people working with me that would prevent me from doing such a thing.

It looks like I might not have a chance to actually put it together anytime soon, but it would be a good experience to learn how to do it.

Thank you all