Marx Gen proposal / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Marx Gen proposal

1 2 3 4 next

Move Thread

LAN_403

Dr. Shark

Thu Feb 09 2006, 11:40AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Its kind of sad seeing the High Voltage board completely empty like this, so I guess some europeans need to fill it up while the guys on the other side of the pond still sleep

I recently bought a bunch of 30 doorknobs on eBay, with a rating of 30kV 590pF or so. What better use for these than a Marx? I was thinking that in order to keep the losses down, it would be a good idea to use as few stages as possible, with as high a stange voltage as possible. So I am planning of seriesing 3 caps for each stage, for a stage voltage of 100kV or so (yeah, these are though little caps, they can easily stand that!). This seems more feasible to me than having a huge tower of more than 10 stages and all the associated hassle.
For charging resistors (dont laugh!) I was thinking about drinking straws filled with water and termindated by a 1/2" brass sphere at each end. This gives me a resistance of som 100kOhm, which should be OK. Spark gaps and all other connections are going to be made with the same brass spheres, to keep corona losses down. I hear that this is important, I have never been near 100kV myself

Charging the thing up will probably be the hardest part, and I am planning to build a 5 stage CW running off of 15kV for that. I've got some 20 foil caps rated at 10kV and a bunch of 100 diodes with 15kV standoff which I could use for that. corona is going to be a much bigger problem here than in the Marx, since it is going to run CW, so I'll definitely have to pot that in some way. Parafin wax would be nice, but I hear that it is prone to form voids, so I might have to use messy oil, bah!

This thing will probably run off a halfbridge chopping mains at 50kHz or so into some jet to be devised homemade HV transformer, or a monitor flyback if it can stand the abuse.

So I already spend the money for all the parts, but I have not started actually working on the thing yet, so if there are any fatal flaws (and I am sure there are many) please let me know before I put too much work into this

Other comments are also welcome of course. I'll post a few pictures soon and keep you up to date on the progress, if I make any

psychop

Thu Feb 09 2006, 12:21PM

Registered Member #83 Joined: Thu Feb 09 2006, 12:15PM
Location: Warsaw, Poland
Posts: 5

sounds good, i am not sure about the straws as resistors, though. I've given several
tries to water resistors and sooner or later they ended being choked up with the air
bubbles forming inside.

Supposedly you can prevent this by filling it with compatible salt solution (ie copper electrodes, copper sulphate solution) but i haven't tried that (got some commercial
HV resistors instead).

Dr. Shark

Thu Feb 09 2006, 12:43PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Yeah, that is a good point. Actually I was hoping to get some input from the x-ray community, plenty of experience with these ultra-high voltages.
Commercial resistors that stand off 100kV are rare though, and stringing up smaller ones is inevitably going to lead to corona trouble.
I have got two ideas I'd like to try out to get the water resistors working:
1) fill the straw with some porous material (e.g. cotton wool) to act like a wick and make sure there is contact to both electrodes, even when some water is lost
2) have a wire hanging from the top electrode into the water by a few cm

both ways, it would not be nessesary to seal the thing, which probably makes live a lot easier. Have you played around with these possibilities?

HV Enthusiast

Thu Feb 09 2006, 01:06PM

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

wrote ...

I was thinking that in order to keep the losses down, it would be a good idea to use as few stages as possible, with as high a stange voltage as possible.

I think your losses will be much greater (at least during charging) with 100kV than 30kV per stage - mostly due to coronal and leakage losses.

wrote ...

This seems more feasible to me than having a huge tower of more than 10 stages and all the associated hassle.

If if you use fewer stages (say even 5 at 100kv per stage), your tower height is proportional to the maximum voltage, not number of stages. 5 stages at 100kv per stage will be about the same height as 15 stages of 30kv per stage, and likely higher since the stage voltage is now 100kV vs. 30kV.

Also, although the 30kV looks feasible on paper, 100kv is a HUGE jump over 30kv.

1. Generating 100kV is quite difficult - coronal losses can typically exceed your source current, unless you are using big caps in your CW and have a good source to drive it with.

2. 100kV voltage stand-off vs. 30kV voltage stand-off

3. Components must be rated at 100kV vs. 30kV.

If i was building the marx generator you proposed, i would do keep it to maybe 10-15 stages, but parallel up those caps to get a bigger bang and keep each stage at 30kV.

Anyways, whichever way you go, good luck and be safe!

Dr. Shark

Thu Feb 09 2006, 02:10PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Hey Dan, thanks for your input! Actually you got me into the idea of having only few stages when you were talking about some serious Marxes on the old board. With losses I was of course referring to the discharge, but I am not even sure about that. The inductance loop will be about the same, and a larger number of shorter sparkgaps probably dissipates the same power.
But surely you must have had a reason when you said that for a serious pulse, the number of stages should not be above fife?

Well, paralleling caps sounds like the best option to get high peak power without too much trouble, especially since my caps are really tiny compared to what comercial Marxes have anyway, and that is probably my limiting factor.

HV Enthusiast

Thu Feb 09 2006, 03:13PM

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

I still have a marx generator project i'm working on for EMP generation that i'd like to get to sometime. My design was basically using about 6-10 capacitors per stage (40kV doorknobs) parallel, and about 4-5 stages.

If you are looking for really big arcs, the bipolar arrangements i've seen some people do look promising.

the_anomaly

Thu Feb 09 2006, 03:30PM

Registered Member #19 Joined: Thu Feb 02 2006, 03:19PM
Location: Jacksonville, FL
Posts: 168

This Marx idea might work if you want to use water resistors. The idea is you have two tubes on the sides which are the resistors and the frame of the unit. These tubes are filled with salt water and are made out of plastic. This could eliminate all the smaller resistors but the tough part might be keeping the capacitors equally spaced or they might over/undercharge?

Dr. Shark

Thu Feb 09 2006, 03:46PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

I actually had the same idea, and I do not see the problem you point out about the charging voltage. All caps see the full supply voltage, dont they?
The reason why I abandoned this idea is simply that I think it is not very practical. I would have to use deionized water to achive a high enough resistance, as even with drinking straws the resistance is a bit on the low side. But with deionized water I would have to keep things very clean, and thats not something that I am good at. Well, if I had a couple of thin acrylic tubes which could be used for this I would probably try it.

Marko

Thu Feb 09 2006, 06:20PM

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

100kV will be surely too much for 3 30kV doorknobs.
4 will be on edge and 5 safe enoguh, but its obivously impractical.
Rather power single knobs with some 15kV, this idea is much better.

Water resistors seem messy to me, has anybody tried to make resistors of thin vulcanized rubber ?

I tried but resistance gets too low (for marx) with thickness I used, maybe different rubbers would give different resistances.
It would be cool and simple high power resistor, just tracks of rubber and capacitors connected to them ?!

Dr. Slack

Fri Feb 10 2006, 08:40AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

My idea for marx water resistors (never got round to trying it so this comes with a "talking out of my @r$ warning"). All dimensions are examples, scale to suit

Make a long tube consisting of 1" copper, 10" pushed-on PVC, 1" copper ... and so on until you have as many bits of copper as stages in your marx. Tie it into a vertical helix, connecting each copper section into the marx stage, looping the plastic on from stage to stage. Take some time to make sure that the tube has a nice consistent upward gradient. Now fill the entire tube with the fluid of your choice. With copper electrodes, very weak copper sulphate would be a good choice. For filling, connect the two bottom ends together with another piece of plastic, any you have effectively one giant U-tube. The initial fill will not be bubble-free, but you can flush through easily, and see bubbles through the clear plastic. An alternative way to fill bubble-free would be to dip the bottom tube into the solution and suck from the top (if you have enough suck, and don't swallow too much CuSO4, LD50 for mammals is in the order of 10g for our weight).

The advantages I see to this arrangement are as follows. The fact that both ends of the tube are accessible means that you can flush the fluid through all the resistors in a single operation. The cylindrical geometry is right for corona. The system is vented, so any heating or gassing will result in spitting out fluid (into a spherical resevoir at the top), not a burst. You could leave the bottom plastic tube in place, increasing its resistance by clamping in several places or closing plastic taps so as not to load the charger too much, and then refilling/flushing becomes extremely easy. Any bubbles generated during operation should move up to the top by bouyancy.

I'm sure this must have been done somewhere?

1 2 3 4 next

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