grounding for a marx generator / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

grounding for a marx generator

1 2 next

Move Thread

LAN_403

PanosB_GR

Thu Nov 06 2008, 06:22PM

Registered Member #1680 Joined: Fri Sept 05 2008, 04:19PM
Location: Greece
Posts: 43

I was just wondering, what considerations should be taken for the grounding system of a marx bank and the device under test.

I have seen that the high voltage ground is usually a fat copper bar that connects the marx bank ground, the test object and the voltage divider. But i was wondering if someone should put any metal bars into the soil, to be sure that the ground voltage does not rise to very high levels, and thus killing any measuring equipment, like oscilloscopes.

I have some 40mm wide and 4mm thick copper bar for ground, but i am planning to place copper rods into wet soil aswell.

Regards!

Proud Mary

Thu Nov 06 2008, 09:08PM

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

Hi Panos,

the impedance of the ground needed depends upon the size and speed of the charge to be dumped. The bigger and faster the Marx, the better the earth/ground needs to be. Where the Marx is very fast, the length of the earth line may become an issue, with reflected pulses forming high voltage nodes along the earth line.

Remember that the ground impedance will depend upon your local soil conductivity. Dry sandy soil will have a much higher resistance than moist, closely packed earth.

You can reduce ground impedance by the use of multiple earth rods, or by burying long lengths of wire mesh or mat in the ground. Local conductivity may also be improved by mixing salt (such as cheap road salt) into the earth around the ground rods, and keeping the salted ground moist by watering.

PanosB_GR

Fri Nov 07 2008, 04:28PM

Registered Member #1680 Joined: Fri Sept 05 2008, 04:19PM
Location: Greece
Posts: 43

I am designing a marx bank with 50kV-100nF/stage. This is going to be a plain air insulated marx-bank (non triggered spark gaps), so it will have a lot of parasitics, but i hope that i can get good risetimes (<100nsec maybe?). Number of stages is not yet known, but i hope to reach 6 stages. I also have found a neat 50kV/5mA charging unit. It has the high voltage terminal inside it, so i need some coaxial means to get the voltage out. I will try with some solid ptfe coaxial cable like the RG-58 or the RG-218 or something like that.

This sounds odd, but a coaxial with 1cm of solid teflon insulation should withstand the voltage. The ends just need to have something like 30cm clearance from the core to the shield.

I will try to make a resistive divider out of high voltage resistors, which will be about 6Meg. The divider will be the initial marx bank load, to test the synchronization of the bank.

My concern about the grounding system is not to burn the oscilloscope, that I will hook up there to measure the waveform!

If things go well, i will post some pics :)

Steve Ward

Mon Nov 10 2008, 03:00AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

You need to understand what grounding really is doing. Your marx doesnt give a crap that you are connecting it to soil, what it cares about is its capacitance to its environment, and that that environment is firmly connected to its "0V/ground" point. I would suggest you build a faraday cage and contain the entire system within that cage, then grounding becomes simply a safety measure for humans that incidentally touch the cage. You basically want to ensure a proper return path for the marx discharge, and keep in mind capacitively coupled currents due to the extreme dV/dt.

I would suggest NOT hooking up any test equipment that you wish to keep in working order to the marx generator. If you are asking questions about grounding, then you do not have a good enough understanding to make test equipment work reliably around these machines.

Also, using big fat copper strap for grounding that connects to resistive soil is basically pointless in most cases. A truly good ground would have a large sheet of conductor embedded in the earth below the device you are grounding to. Resistive soil approximates this

Mathias

Mon Nov 10 2008, 01:06PM

Registered Member #1381 Joined: Fri Mar 07 2008, 05:24PM
Location: Hungary
Posts: 74

Well now , at the lab we have a 6 stage marx generator (50kV 36nF maxwell discharge caps per stage)
A "good grounding sceme" with some more precautions implemented : (Like the lab is on the basment level , separate "ground" for everything thats connected to the marx and/or its return path , and spec line filltering ,etc)
We managed to make an environment where we operate a 200kV marx while the peak current in the system is something like 20kA (T/2 140-170ns pulse) and still we are able to make measurement with our delicate equipment (cause of the great EM Shildings theres just only a few hundred mV! of noise 3-4meters away from the system)

There's also a bad example:
Around 2000 at some german university , the first time they fired the thing, all the stuff on the floor "Fried" , when the pulse passed trough the wiring of the building :P

By the way what are you planing to accomplish with a marx big as the one you are building ?
Just asking cause , if charged to more then 200kV the output can flash over to the case of the box in air
So for 300kV , you will need to build it in a "huge" box if used in air ,or pressurize the closed box with N2/SF6 , or just put it under oil :)

Best of luck with your project

Proud Mary

Mon Nov 10 2008, 03:38PM

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

PanosB_GR wrote ...
so i need some coaxial means to get the voltage out. I will try with some solid ptfe coaxial cable like the RG-58 or the RG-218 or something like that.

This sounds odd, but a coaxial with 1cm of solid teflon insulation should withstand the voltage. The ends just need to have something like 30cm clearance from the core to the shield.

It is possible to foresee circumstances in which pulse reflection in a transmission line output could create nodes of 2V, which in your case would be 600kV - not an easy voltage to contain!

PanosB_GR

Mon Nov 10 2008, 03:55PM

Registered Member #1680 Joined: Fri Sept 05 2008, 04:19PM
Location: Greece
Posts: 43

This marx is aimed for driving a microwave tube. The return of the discharge current is indeed what i thought really mattered in the first place, and probably the high voltage "ground" should be kept floating, only connecting the marx reference node, the tube chassis and the dummy high voltage load/divider.

I can't enclose the bank into a cage, it will just be standing in the middle of the room.
If I make some metal shield for the measuring equipment, and place it in there, running on car batteries, will that be safe for the equipment?

PanosB_GR

Mon Nov 10 2008, 04:02PM

Registered Member #1680 Joined: Fri Sept 05 2008, 04:19PM
Location: Greece
Posts: 43

Harry wrote ...

It is possible to foresee circumstances in which pulse reflection in a transmission line output could create nodes of 2V, which in your case would be 600kV - not an easy voltage to contain!

If big charging resistors (megohms) are placed between the supply and the marx bank, why should something like pulse reflection even happen?

Whatever happens in the marx bank leaves the charging unit relatively unaffected.

Dr. Slack

Mon Nov 10 2008, 05:02PM

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

One of the easiest ways to get confused about what is happening when several circuits (marx, oscope) are connected is to start thinking that, having called any given node "ground", the electrons give a toss. When you draw an earth symbol on a diagram, or write "0v" against a wire, all it's doing is relfecting your choice. It may be that if that node is actually connected to some remote large conductors (like the earth), then parasitic capacitances will exist, and draw charging current, but that's pretty much all.

The return of the discharge current is indeed what i thought really mattered in the first place,

absolutely!

When the marx is just driving a load, it's quite easy to see where the current is flowing, and where the rapidly changing current will induce significant voltages across the miniscule inductances you have left lying around. When you come to connect a scope, you have to choose where to reference the measurement to. One feasible way is to consider the marx as effectively floating, if you connect it to its power supply via two very large resistors, or even disconnect before firing. Then you can connect the scope reference voltage and the chassis (potetially two terminals called more specific names than "ground"), to the bottom end of your marx load and to the bottom end of your voltage divider.

For general safety, you may also have your scope connected to the soil via a huge inductance.

Proud Mary

Mon Nov 10 2008, 05:16PM

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

PanosB_GR wrote ...

I don't quite follow your reasoning here. When the Marx is "erected," then all the capacitors are connected in series to form a single capacitor, which then starts to discharge into your proposed coaxial transmission line forming a pulse.

There is a good deal of literature on Marx generators with coaxial transmission lines which you will easily find on Google.

1 2 next

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