Chaining Small Diodes to Make Big Diodes

[rwbohring]

I am cutting corners again. I'm building a single stage rectifier 15kv in 30kv out. I don't have the budget for HV diodes so I connect 20 1000v diodes end to end (7ft long) for each diode in my rectifier.

 

Seems to work. I haven't measured with a divider and a voltmeter but the gap discharge measures about 1.75cm. A little wider than my calculated Paschen estimates, possibly because of a Corona effect.

 

Anything I should be conscious of or worried about as I continue?

 

I plan on adding more stages and I'm curious if I'll get more waste in later stages using this diode method?

[studiot]

Yes the method is fine and used to be used to generate HV for things like cathode ray tubes.

 

The only thing to watch is to make sure that the chain is routed so later stages are distant from earlier ones so there is never a significant field due to proximity.

They are often laid out in a zigzag pattern on the board to save space. I have also seen 3D helixes constructed on three boards radially arranged at 120^o inside a plastic tube to keep the 30kV away from ground

 

Also make sure that all joints are smoothed over, and there are no sharp elbows in the wiring.

[Carrock]

One other point.

The reverse leakage current can vary significantly from diode to diode.

You could get most of the reverse voltage appearing across one very low leakage diode, exceeding its PIV.

You should check the spec of your diodes to see if they can survive this.

Normal practice is to connect a high value resistor rated at 1000V across each diode, passing just enough current to swamp the diodes' reverse leakage.

[studiot]

One other point.

The reverse leakage current can vary significantly from diode to diode.

You could get most of the reverse voltage appearing across one very low leakage diode, exceeding its PIV.

You should check the spec of your diodes to see if they can survive this.

Normal practice is to connect a high value resistor rated at 1000V across each diode, passing just enough current to swamp the diodes' reverse leakage.

 

Good point +1

This is better practice but not always done.