http://www.highvoltageconnection.com/high-voltage-switches.html

These work very well...i don't know the specifics of your application...but this company makes a great series of hv switches and switching hv mosfet solutions. hope this helps .

I have industry experience in ESD testing and have built test equipment for EMC immunity compliance for computer equipment. Happy to discuss your application and typcial practices in industry.

Normally, a swtich is not used at these voltages for ESD testing. I can explain further if interested.. If you really need a switch for your application, however, try:

http://www.gigavac.com/ they make switches up to 70kV.

We would be happy to discuss a potential solution. Please contact me at [email protected] Our website is www.divtecs.com.

Thanks for Your pointers to these switch manufacturers! It seems like the number of companies making HV compoents are fewer than they used to be, some components are getting rather hard to find.

My problem is that most semiconductor solutions seem to have poor di/dt at higher voltages and gas discharge tubes have their issues as well. Primarily in order to try to improve the initial current waveform, I did build a mechanical gas pressure driven switch.It works better, but I still would like to improve repeatability of the current waveform. The setup is for variable energy tests around the working points of some existing standards in order to ascertain margins during design work. So it is not exactly a standard setup. The initial current can be a little over 100A. My present solution essentially is a mechanically driven spherical electrode spark gap. It is configured as a SPDT switch so that it is also used to charge the test capacitor.

Two mechanically driven spherical electrode spark gap may be solution but if you add third electrode triggering to this( as we do in case of impulse generator) the system will be more deterministic and repeatable with lesser margin. The third electrode trigger can be done by solid state or relay operated circuit with proper isolation.

Buy small rating semiconductors switches, and connect them in series so that you can get High Voltage Switch. For example, if use want to make 10 kV switch, you need 6 numbers of 3.3 kV switch (3.3 + 3.3 + 3.3 + 3.3 + 3.3 + 3.3 = 10kV, for safe we are using half of the rating of the switch). To avoid voltage imbalance, you can use snubber circuit (load side balancing) or Active Gate Control (Gate side balancing).

Let me know if you need more information or you can refer ABB website/IEEE Papers for Series connected IGBTs.

@Hansjoerg Contact bouncing is no problem for these switches at higher voltages since the discharge event is well over before mechanical contact is made. For lower voltages and consequently shorter discharge distances I have seen the contact bounce problem for longer discharge times. I have limited experience with more modern solid state switches, and will certainly look in to what is available. Thank You all for the links! What I have previously used in the past for other purposes is puck type thyristors from Asea. They are good for rather high currents but to slow for many purposes. And thyrathron and similar older gas discharge type devices used for magnetrons and similar purposes.

@Arijit I believe my main potential improvement to be decreased jitter in di/dt. Absolute time jitter is not important. My risetime is inductance limited already. What You describe seems similar to the ignition electrode in gas discharge switches and xenon flashlamps. For several reasons in this particular circuit, I believe an air spark gap switch to probably be the best solution. Can You give me any pointers to practical designs and is it possible to design such a triggered switch for varying voltages? And do You think di/dt jitter would be improved?

Erik, if the current ESD standards do not work very well for you, please look at this publication to definde your present testing needs and the solutions will be obvious.

http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=148232

Eric, the switch can be designed for break down at variable voltage and I think if designed properly di/dt jitter will be improved. Please find attached a sketch of a possible design if you find it suitable and need further clarification you are free to write to me

We have used in the past, a Vacuum-Gap discharge unit made by EEG. Data is difficult to find on this product, but they seem to be "relatively" reliable. We use them in a crowbar circuit to protect Inductive Output Tubes in broadcast amplifiers.

You used IGNITRON as a high voltage switch. and if its not suitable then used solid state crowbar. I am sure it will work better.

@Arijit That seems to be a very interesting solution for a low side switch. In my setup, the potential on the switch varies, and I measure the current waveform during each discharge, which makes the ignition electrode difficult to implement. It would certainly have to be on the high side of the circuit. Clever and simple idea!

@John I remember using EEG units in the past. They seem hard to find these days. Some old copying machines used to contain such units, and they can have excellent risetimes and peak currents compared to solid state solutions. The units I have seen have been for lower voltages than my present requirement. Some of these kryotron type units now are restricted in trade for security reasons.

I've used EG&G triggered spark gaps for years in radar transmitters at around 40 KV. The problem is that they work on a limited voltage range. The air triggered gap has the same problem, but at least you can adjust the gap. Air gaps also keep the riff-raff out of the testing area (after the first few shots). Or, try a transformer fed by a PFN, fired by a lower voltage SCR bank. You can sharpen the pulse with a passive gap on the output.

Oh yeah, I used to use Ignitrons in my transmitters. They worked great, but the Greenies confiscated them all (full of mercury).

Thanks for Your comments, James! Those are among the reasons for which I decided to use an air gap in this my most recent contraption. Plus that it has to be a high side switch and work over a wide voltage range. I have an old stash of goodies that I have carefully guarded, but it seems that a simple pneumatic air gap gives very good rise times. My problem is that I would like to reduce jitter, eg.make the spark behave more uniformly. I believe this may be easier at higher currents, peak current is just 50A to 100A in my case. A secondary adjustable airgap seem like a very good idea. Just being able to use better shapes for the electrodes may make the difference.

Thanks

I have used the spark gap on the discharge side with good effect for testing devices. The initial design for the spark gap had just a set screw for adjusting the gap but during troubleshooting the gap moved resulting in a much larger gap than needed and a rather loud discharge the scope captured it at near 100Kv. The second design had a screw thread for adjustment with a nylon set screw to lock the position the output voltage was adjust ed by increasing the gap for higher voltage and decreased for lower voltage. I have attached a picture similar to the spark gap I used.

Thanks for the picture Brian!

Actually it looks somewhat similar to the switch I ended up using. I used slightly flattened spheres about 50mm diameter for the electrodes and 16 mm copper tubing for conductors to minimize corona etc. The parts were assembled by silver soldering. The charge was held in a low inductance variable capacitor monitored by a electrostatic voltmeter. I got best results (in my case defined as lowest jitter on the current waveform) when moving the spheres to close the gap by a pneumatic system. Slightly flattening the spheres does not affect corona perceptibly and seems to reduce jitter.