High voltage SCR hookup

[stany]

Hey guys,
I'm a new member to this forum, so I have no idea if this topic has arisen before.
I have a project coming up where I need to hold off about 3000 volts using an SCR. A single device with this type of holdoff is really expensive, so I'm looking for an alternative solution. Years ago, I read (or heard of) someone putting two SCR's in series to achieve higher voltage rating. I believe it can be done, but I'm not really sure how to trigger two SCR's at the same time (isolation transformer, precise timing and Di/Dt and all that good stuff). Anyone out there with any experience in this application??? Thanks, Stan

 

[pauseflash]

I know one of the radars I work on, uses 2 SCR boards, with 8 SCR's each on them to hold off about 10,000. Off my head, I cant just ramble out the way it works, but I'll look through the schematics and the theory book tommorow at work and try to get you a more definitive answer.

 

[budavarsa]

what about current?
ain't it hard to cut off scr? will you design a cut off circuit?
will you use controller?

 

[ante1]

From what I have seen they use pulse transformers to trig from a lower voltage to achieve simultaneous response.

 

[AN920]

Look at this AP-note
http://www.eetchina.com/ARTICLES/2000DEC/2000DEC12_AMD_AN5.PDF

This gives you the basic idea, but there are other problem issues not mentioned here.

 

[AN920]

Let's look at a typical hookup using two 2N4444's in series.
http://pdf1.alldatasheet.co.kr/datasheet-pdf/view/87010/MOTOROLA/2N4444.html
Each has a blocking capability of 700V. We want to use it in a 1000V application. We connect it as shown in the first diagram, but find that the load turns on when we connect the power, even with no trigger applied.

The problem is the leakage current that may vary from 10uA at a cold junction temperature to 2mA for a hot junction. In the diagram I used a leakage for 10uA for the one and about 20uA for the other. We can see that D1 with the lowest leakage will have more than its rated 700V over it. This will cause voltage breakdown of D1, followed by breakdown of D2.

We need to create a additional leakage path greater that the worst leakage spec in order to balance the voltages. If we take 3mA to be very safe, the voltages balance out but we have to live with the high leakage current through the load, which may be a problem for the application. Also note that we will be generating a fair amount of heat in the balancing resistors as shown on the wattmeter.