Home made HV diode 50KV, 30A peak - complete protocol / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Home made HV diode 50KV, 30A peak - complete protocol

previous 1 2 3 4 next

Move Thread

LAN_403

Weston

Fri Feb 22 2008, 02:34AM

Registered Member #1316 Joined: Thu Feb 14 2008, 03:35AM
Location: Cambridge, MA
Posts: 365

where are you buying this many diodes? at my local radio shack it is $2 for three.

CompWiz

Fri Feb 22 2008, 03:01AM

Registered Member #1038 Joined: Mon Oct 01 2007, 08:02PM
Location:
Posts: 96

there cheap almost anywhere online

At mouser:
1: $0.05
10: $0.04
100: $0.03
250: $0.02
500: $0.017
1,000: $0.015
2,000: $0.013

...

Fri Feb 22 2008, 03:06AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

you can get them from a real electronics supplier for about a penny, or on ebay for even less.

In any case, it is true that without balancing resistors/capacitors you will run into trouble sooner or later, but really if all you want to do is rectify HV these things should work fine.

I have made similar strings (but without oil, I never really had a problem with insulation with long diode strings) using sf1600 (1.6kv/1a avalanche rated, with a recovery time of a few ns) or the more common uf4007's (similar ratings save for a 1kv max voltage and no avalanche duty), for rectifying ac flybacks and the like, with which I have had very good luck in the past just as long as they are well over designed.

I think the reason you can get away using strings like this for low impedance sources (at least anything that isn't a 20kj cap bank...) if the voltage gets too high for one of the diodes it will avalanche (all diodes do this at a voltage somewhat higher than their rated voltage, only some of them are designed to be able to handle high amounts of currents in this condition, but all of them do do it), which will probably cause a chain reaction that avalanches the whole string, which will probably short out the supply for one cycle, but after the voltage drops back down to 0v the diodes reset and you don't really notice it anymore.

Dr. Dark Current

Fri Feb 22 2008, 07:50AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Maybe I'm mistaken but don't most diodes balance themselves by of the avalanche effect?

Proud Mary

Fri Feb 22 2008, 11:04AM

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

The things is that Mates' string of 1N4007s works, is not expensive, and can be copied by other enthusiasts without much difficulty. A job well done!

Tesladownunder

Sun Feb 24 2008, 12:21PM

Registered Member #10 Joined: Thu Feb 02 2006, 09:45AM
Location: Bunbury, Australia
Posts: 1424

Chris wrote ...

I was using a 7.5kV 60mA NST, and had four strings arranged as a full wave voltage doubler, thereby outputting about 21kV peak. If I remember right each string had 35 1N4007 or similar under Exxon Univolt dielectric oil and balanced by 1Mohm resistors. One of the strings failed once the caps were charged to 18-20kV, just shy of the target voltage

Several points.
1) When you charge a cap you need a diode capable of double the DC cap voltage, because for part of the cycle the opposite voltage is applied, hence MOT diodes are rated at 10kV for a 2KVAC supply.

2) I also suspect that resistor strings are a BAD idea to balance diodes. To understand why, consider that a half watt resistor has a 500V continuous rating and 700V peak or something like that. If only one resistor goes open that diode blows and puts more stress on the others.

3) Full wave voltage doubler on half an NST? Doesn't this mean your big storage cap was floating with respect to earth. Or if it was earthed, there may well have been an insulation breakdown somewhere.

4) It is just possible that your failure of your diodes was secondary not primary. Consider a short circuit at a different site. The discharge will ring and on the first voltage reversal your caps will be subjuct to a fast pulse of reverse voltage. So if you were charging 20kV, you suddenly get the equivalent of 40kV with no limiting resistor. and there go your diodes as plenty of ringing kJ comes back at them.

5) I am beefing my pulse caps up from the 5kJ limited supply to one that can use the full 16kJ. My testing will be "around the corner" with remote voltage metering and remote triggering but with no high voltage path away from the unit. I am scared of a few things in HV. A loaded bomb of a pulse cap is one of them. I might have to take it to a firing range to use it.

6) Long strings work fine, but over rate them and never have a situation where failure will be catastrophic. I have a 60k 100W resistor in series.

7) In regard to Mates Ignition coil rectification. If you are trying to rectify an interrupted ignition coil then the 1N4007 string will fail as the pulses are a lot faster than the 1N4007's can handle. And for the reasons above, your "50kV" diode will only be able to rectify 25kV into a capacitative load. it should be fine for an NST but the oil is overkill. I just use the PVC flexible tube (like on my 80kV CW multiplier with ?8 strings of ?12 x 2.5kV diodes.

Here is my 300 string of 1N4007's. Works fine for a voltage tripler for 30kVAC denatal transformer to give a 5 inch spark.

TDU

Mates

Sun Feb 24 2008, 06:02PM

Registered Member #1025 Joined: Sun Sept 23 2007, 07:53PM
Location: Czech Rep.
Posts: 566

Tesladownunder wrote ...

7) In regard to Mates Ignition coil rectification. If you are trying to rectify an interrupted ignition coil then the 1N4007 string will fail as the pulses are a lot faster than the 1N4007's can handle. And for the reasons above, your "50kV" diode will only be able to rectify 25kV into a capacitative load. it should be fine for an NST but the oil is overkill. I just use the PVC flexible tube (like on my 80kV CW multiplier with ?8 strings of ?12 x 2.5kV diodes.

TDU, this is not a theoretical thread. The diode string which is presented in this thread works fine for ignition coil rectifying. I do not understand how you can claim that these diodes can not handle 6KHz AC while see the opposite. I can rectify the iggy output, connect a single cap in parallel and make 5-6cm long discharges... Cheers

Kizmo

Sun Feb 24 2008, 07:34PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

I think this is mostly because component makers have to give their products quite 'low' specs that they will survive thousands and thousands of hours with certain voltage/current values and those might be little bit lower than what they can take in relatively short runs.

Proud Mary

Sun Feb 24 2008, 09:32PM

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

Mates wrote ...
I do not understand how you can claim that these diodes can not handle 6KHz AC while see the opposite. I can rectify the iggy output, connect a single cap in parallel and make 5-6cm long discharges... Cheers

Yes. The reverse recovery time of 1N4007 is about 3us, and the forward recovery time just a few nanoseconds.

Tesladownunder

Sun Feb 24 2008, 11:33PM

Registered Member #10 Joined: Thu Feb 02 2006, 09:45AM
Location: Bunbury, Australia
Posts: 1424

Mates wrote ...

Tesladownunder wrote ...

7) ....If you are trying to rectify an interrupted ignition coil ......

I missed the 6kHz which means you are not running an interrupted ignition coil but are using solid state. Never the less it will not be a sine wave and the rate of change will still be a lot higher than 6kHz would indicate. An applied square wave is made up of successively higher frequencies - which is what Fourier transforms are about. You will have components of that going to 30KHz and higher, I would suggest. Reverse recovery of 3uS means that it takes a half cycle of 150kHz to recover (by which time the voltage has reversed). Or 1/5 of a half cycle at 30kHz during which time it may be shorting a capacitor.
I have done experiments with an interrupted twin ignition coils (using SIDAC's) getting 4 inch sparks and I can't rectify that spike using 1N4007's.
If your spark initiates at 5-6cm WITH a capacitor then this will be around 50kV - 60kV depending on your electrodes and you are over double the voltage rating of your diode. Hence you would need rating of absolute minimum 100 - 120kV.
If on the other hand, you are talking about a spark that can be drawn out to 5-6cm than the voltage is a lot lower and you may just be within specifications.
And theoretical thread or not, people blow up diodes in rectifiers and multipliers because they think it is ony the output voltage they need to rate it at. Sometimes this can have serious implications as in Chris's cap bank explosion (for what ever reason).

So does your spark initiate at 5cm or get drawn out to that?

TDU

previous 1 2 3 4 next

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