homemade hv diode? / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

homemade hv diode?

Move Thread

LAN_403

haulinick

Sat May 12 2007, 12:51AM

Registered Member #619 Joined: Sat Mar 31 2007, 05:26AM
Location: San Jose, CA
Posts: 16

I need to make a diode to rectify a 10k nst (30ma), but i don't wanna spend a bunch of money or bother ordering anything. I do have 10 1000v diodes. Can I use them and if so would i put them in parallel or in series?
I'm guessing series. But if so, wouldn't the first diode in the line just get all the voltage? I am so confused.

Billybobjoe

Sat May 12 2007, 01:41AM

Registered Member #396 Joined: Wed Apr 19 2006, 12:55AM
Location: Pittsburgh, PA
Posts: 176

You would put them in series, but with only 10 that is very risky. Ideally one should use more than the bare minimum, especially when using semiconductors at high voltages. Your setup will yield a 10kv half wave rectifier, but considering your NST is rated at 10kv, and voltage is not always spot on 10kv, your diodes probably will not last long.

GreySoul

Sat May 12 2007, 01:55AM

Registered Member #546 Joined: Fri Feb 23 2007, 11:43PM
Location: Albuquerque, NM
Posts: 239

heh.... funny you should ask. I just did the same thing.

I got 500 1N4007's from ebay for like $8.

Here's what Steve Ward, and many others, done learned me:

You need your diode stack to resist the full peake voltage of your NST. use the formula:
(Vpk = 1.414xRMS) or (Vpeak= Vrms * SQRT(2)) to be exact.

So you really need more like 14,140 volts for your diode, plus you want factor 70-80%, so a 14kv diode doesn't like being pushed past about 9800v...

Just to estimate and make everything very robust I'd make your string out of 14-16 1000v diodes per leg of a rectifier. So for a fullwave you'd need like 60ish diodes. for a simple halfwave you'd just need 16-18

Since most NSTs are center tapped you can probably make a simple doubler off each leg with like 40 diodes.

anyways... hope that sounds right.

thanks Steve and taylor, you guys rule.

-Doug

edit: I made as full wave 10kv 1A rectifier with 60 1n4007 diodes.. I use a 12/30 NST with a variac and a stop for 10kv.

pic is clicky for larger image.

I haven't tested it yet... not sure why not. I think I'm waiting to pot it in my vacuum chamber. yeah..that's it.

-Doug

edit:

ok I tested it.... seems to hold up at 14kv (estimated) no problem.

I forgot to hook up my current limiting resistor... I have no idea what kinda current I was drawing on my poor HeNe tube... not more than 30ma tho. It's a 40mw M-G tube... after about 3 minutes the anode was too hot to touch :\ The rectifier is the same as the one above, and as you can see it has no problem powering my little laser off an NST.

Dr. Slack

Sat May 12 2007, 07:10AM

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

If you put commodity diodes (that is not specifically matched or intended to be seriesed) in series, then you want to derate at least 50%, that is don't design for more than 500v across any 1N4007.

When you consider the various things that can unbalance the string and make one diode see much more than its fair share of voltage, unequal reverse leakage current, unequal junction capacitance, and worst of all unequal charge storage time that makes them switch off at slightly different times, you may feel you want to derate them more.

You could add complexity by shunting each diode with a C and an R to swamp variations in the first two, but that's more expensive and more work than using a few more diodes with more derating.

Diodes intended to be used in series (and the individual die that are series stacked in multi-kV rated diodes) are avalanche rated as well as being well matched, so the first one to switch off doesn't die while it waits for the others to catch up.

I've got 200 off 1N5408 as well. I made a 10kV rated diode with 20 of them. Clipped the leads short and soldered all in series. I "potted" them in hot-melt by puttign a gob of glue on each solder joint, filling the space to about the same diameter as the bodies. Then I used heatshrink tube, heating carefully and moving very slowly from bottom to top, trying to allow air bubbles and excess glue to rise as I moved up. The resulting assembly gained a lot of mechanical strength from the glue and tube. It's not failed yet.

r3d

Sat May 12 2007, 05:12PM

Registered Member #532 Joined: Sat Feb 17 2007, 05:00PM
Location:
Posts: 11

You could get 10 24kV 3mA (like BY713 or BY8420) rated diodes for around 7.5$ so i don't see why bother with 4007 or 5408!

Dr. Dark Current

Sat May 12 2007, 05:26PM

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

if you want only half-wave rectification, then you must keep in mind that the diode must be rated 2*sqrt(2)*NSTvoltage, this means at least 30kV for a 10kV nst, but preferrably more (I would suggest 40kV)!
Edit> but i would not suggest half-wave rectification since it is no good for a transformer

cbfull

Tue May 15 2007, 12:00PM

Registered Member #187 Joined: Thu Feb 16 2006, 02:54PM
Location: Central Ohio
Posts: 140

jmartis wrote ...

if you want only half-wave rectification, then you must keep in mind that the diode must be rated 2*sqrt(2)*NSTvoltage, this means at least 30kV for a 10kV nst, but preferrably more (I would suggest 40kV)!
Edit> but i would not suggest half-wave rectification since it is no good for a transformer

I think that the additional factor of "2" only applies to a capacitive load. If there is no capacitor the formula is...

Non-capacitive load rating formula:
= Vrms x âˆš2

Capacitive load rating formula:
= Vrms x 2âˆš2

This has been discussed (debated is a better word) at length before. Some people were even upset at the idea!

Steve Ward

Tue May 15 2007, 05:22PM

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

I think that the additional factor of "2" only applies to a capacitive load. If there is no capacitor the formula is...

And i agree. If dealing with purely resistive loads, then you dont need to double the rating. But the most common application for rectifiers is to charge up capacitors, so this is why most people dont realize you can cut corners for resistive loads.

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