Marx generator questions / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Marx generator questions

1 2 3 next

Move Thread

LAN_403

genious 7

Sat May 29 2010, 11:34PM

Registered Member #2887 Joined: Sat May 29 2010, 11:10PM
Location: Panama City, Panama
Posts: 107

I am going to build a Marx generator using these capacitors, and a hv flyback generator (Which I have already made and tested). I've read that marx generators can generate a lot of EMI and such.

1)What is the best way to avoid pulses going back into mains?
2)Is this issue a problem with 5-10 stages at around 15KV?
3) Any ideas for controlled triggering of the generator?
4) Any other foreseeable problem?

Approximate number of stages will be between 5 - 10. I'd prefer if an independent RF ground isnÂ´t required. The ballasting resistor is around 6M (Made of six 1/2W 1M resistors), charing resistors will be 1M (3 330k resistors in series).

PS: sorry for any grammatical error, English isn't my first language

Proud Mary

Sun May 30 2010, 08:11AM

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

Michael Chen wrote ...

1)What is the best way to avoid pulses going back into mains?
2)Is this issue a problem with 5-10 stages at around 15KV?
3) Any ideas for controlled triggering of the generator?
4) Any other foreseeable problem?

Approximate number of stages will be between 5 - 10. I'd prefer if an independent RF ground isnÂ´t required. The ballasting resistor is around 6M (Made of six 1/2W 1M resistors), charing resistors will be 1M (3 330k resistors in series).

Hello Michael!

1) I disconnect the Marx from its supply by means of a vacuum switch the instant before firing the trigger pulse in order to remove this problem, but you could use an simple mechnical switch at atmospheric pressure if you had no choice.

2) High voltage spikes can be a serious problem if reflected back into the supply. They will often damage semiconductor bench power supplies, sometimes beyond economic repair. Even without any direct connection, the local field gradient around the Marx at the moment of firing may be sufficient to destroy nearby CMOS circuits, and so on, so put your expensive instruments far away.

3) The easiest triggered gap would be a three terminal device - two heavy electrodes for the main discharge, with a trigger needle off centre between them.

4) The loud bang when the Marx fires will sound like a gun shot and may disturb your neighbours. Protect your ears!

Good Luck! 88888

Myke

Sun May 30 2010, 01:09PM

Registered Member #540 Joined: Mon Feb 19 2007, 07:49PM
Location: MIT
Posts: 969

Proud Mary wrote ...

2) High voltage spikes can be a serious problem if reflected back into the supply. They will often damage semiconductor bench power supplies, sometimes beyond economic repair. Even without any direct connection, the local field gradient around the Marx at the moment of firing may be sufficient to destroy nearby CMOS circuits, and so on, so put your expensive instruments far away.

Isn't it the rise time of Marx Generators that cause the problems of induced voltages?

mikeselectricstuff

Sun May 30 2010, 01:10PM

Registered Member #311 Joined: Sun Mar 12 2006, 08:28PM
Location:
Posts: 253

You may want to go a bit higher than 15kv to get consistent gap operation (and bigger sparks of course!). For triggering, try a mechanical arrangement or HV vacuum relay if you can find one. Whetever the trigger mechanism, use a non-conductive switching method ( e.g. air operated hand button, or pull-cord/plastic rod ) as holding anything metal near a marx is a bad idea..
IME resistors are the biggest problem - see if you can find solid carbon ones, as film types (even HV rated ones) tend to degrade quickly

genious 7

Sun May 30 2010, 03:13PM

Registered Member #2887 Joined: Sat May 29 2010, 11:10PM
Location: Panama City, Panama
Posts: 107

High voltage spikes can be a serious problem if reflected back into the supply. They will often damage semiconductor bench power supplies

The supply (for the flyback generator) is a common transformer (with appropriate current ratings), so power supply damage isn't going to happen (I guess).

I disconnect the Marx from its supply by means of a vacuum switch the instant before firing the trigger pulse in order to remove this problem

Would using a relay to disconnect the fly back on the low voltage side work? I just want to avoid noise going back to mains. The flyback is of the new ones, with an internal diode, so that could prevent capacitors from discharging through it. Would it work?

Proud Mary

Sun May 30 2010, 05:19PM

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

Michael Chen wrote ...

High voltage spikes can be a serious problem if reflected back into the supply. They will often damage semiconductor bench power supplies

The supply (for the flyback generator) is a common transformer (with appropriate current ratings), so power supply damage isn't going to happen (I guess).

I disconnect the Marx from its supply by means of a vacuum switch the instant before firing the trigger pulse in order to remove this problem

The point is to completely isolate the Marx generator from anything else at all in the instant before firing. Large high voltage spikes go through the world doing much as they please unless we take great care to keep them in their place.

I would also suggest you avoid discharging the Marx without a load, as this places great stress on all its components. If you discharge the Marx into a transmission line, much of the pulse can quite possibly be reflected back into its source where it will try to cause trouble.

Not even a god can create mountains without valleys, so your Marx discharge won't be a simple positive pulse - it will be followed by a big negative swing that will try its best to jump your charging resistor and eat the tiny diodes in your LOPT for breakfast given half a chance, before seeing where else it can do mischief.

The resistors you have selected will create such a long time-constant that your discharge will drag on like a state funeral, and instead of being a quick up-and-down pulse full of meaty goodness, will look like a mole hill after a rain storm. I use 500R non-inductive wire-wound resistors in my own small apparatus, but many think even 500R too high, and will use 50R to create pulses with nanosecond rise times, so compromise based around the needs of the experiment is always necessary. 1M looks very high to me, but with a flea powered LOPT delivering 1 or 2 mA of charging voltage, I suppose it may be necessary.

I should point out that my methods are probably more conservative - cautious - than many on this forum, because I do not have the money to keep replacing damaged parts and equipment caused by a gung-ho attitude to design and construction.

My own small Marx has only 4 stages to produce an erected capacitance 0.65uF charged to a nominal 40kV, giving a usable energy of about 500 Joules. For spark gaps, I used M14 fine-thread stainless steel dome nuts so that the gaps can be easily set with a feeler gauge by turning the bolts. For charging, I use an 8kV transformer (a 4kV-0-4kV 180mA neon sign transformer with the centre tap removed) and a bridge rectifier made from 4 x 30kV 1A diodes from China. The lowest gap has a trigger needle powered by a one shot HV pulse circuit based around a small lawn mower ignition coil. Immediately before firing (~100 millisec) the Marx is automatically isolated from the power supply by a vacuum relay for the reasons I have given above. I have so far fired 23 shots with this Marx over the last two years, and it has held up well, so my time and money was not wasted!

Anyway, I will now leave it to others to tell you things more to your liking! Good Luck, Michael!

plazmatron

Wed Jun 02 2010, 09:40PM

Registered Member #1134 Joined: Tue Nov 20 2007, 04:39PM
Location: Bonnie Scotland
Posts: 351

Marx Generators can be a royal PITA with respect to kickback.
My largest 300kV marx would always trip the main breakers in the house when it fired. The problem was resolved by running the Flyback supply off of lead acid batteries, and providing its own ground. It all depends on how big you want to make it!

Resistors are the biggest hurdle/limitation when designing marx generators.

Steer clear of carbon composition, and metal oxide resistors. Whilst it is possible to insulate these for high voltage use, you will find somewhere down the line that they fail internally. Some fail short, some fail open, most fail by permanently changing resistance!

The best bet is either wirewound or copper sulphate resistors.
I have been recently experimenting with water/copper sulphate resistors, (since they are inexpensive) and have found them to be extremely robust, and almost indestructible. I have even started using larger ones as ballast resistors on x-ray power supplies.

Les

Proud Mary

Wed Jun 02 2010, 11:50PM

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

plazmatron wrote ...

I have been recently experimenting with water/copper sulphate resistors, (since they are inexpensive) and have found them to be extremely robust, and almost indestructible. I have even started using larger ones as ballast resistors on x-ray power supplies.

Coo that's good, Les...

I completely agree with you about the Marx resistors - I think nichrome wound in the non-inductive way (folded in half and wound outward from the centre) on mica cards is good, and is used in several professional designs described in the literature.

plazmatron

Thu Jun 03 2010, 12:14AM

Registered Member #1134 Joined: Tue Nov 20 2007, 04:39PM
Location: Bonnie Scotland
Posts: 351

I had actually considered using relatively low resistance wirewounds as inductors for a marx gen, but it turned out that water cost a lot less

I have tried choke charging before, but you do need quite large air core inductors for it to work well, and I ended up with a cumbersome Marx gen to say the least.

The only real drawback with water/sulphate resistors, is that resistance changes with temperature, so if your resistor is dissipating quite a few watts, the resistance change as a result needs to be taken into account. (OR you increase the SHC, ie build a physically larger resistor).

For pulsed power applications like in a Marx gen, water resistors can be built fairly small (mine measure 30mm x 6mm), and since they are made from copper and clear tubing, they actually look quite tidy as well.

Les

genious 7

Thu Jun 03 2010, 01:15AM

Registered Member #2887 Joined: Sat May 29 2010, 11:10PM
Location: Panama City, Panama
Posts: 107

Thanks for the suggestions. I'll try to build a relay this weekend out of acrylic to disconnect the Marx from the supply before firing.

Resistors are the biggest hurdle/limitation when designing marx generators.

For the moment, Ill use several common resistors in series in order to do a proof of concept. If it works, I'll try to make a water resistor.

1 2 3 next

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