help, why does my 20kV diode keep burning up

Thread Starter

genekuli

Joined Oct 21, 2018
22
input is 15kV 15ma (as is on attached picture) going into capacitor bank
diode is a Single Phase high-voltage diode rectifier silicon reactor
Model: PRHVP 2A-20 (HVP 20 HVP2A-20) 2A20KV
Size: 18 x 19 x 75mm (Height * Width * Length )
Operating Temperature: -40 ° C ~ + 175 ° C
 

Attachments

Alec_t

Joined Sep 17, 2013
11,752
What series resistance is there to limit the initial surge current? A capacitor bank behaves like a short-circuit at power-up. The transformer 30mA rating is probably the continuous current rating, so it may be able to supply a lot more current for short periods.
 

Thread Starter

genekuli

Joined Oct 21, 2018
22
thank you, i know nothing (about circuitry) but the diode is rated to 2A so that should never be overloaded, right?

oh and there are 4 of the diodes

like this,
but different diodes than is on the plan
 

Attachments

AlbertHall

Joined Jun 4, 2014
10,572
The peak voltage of the 15kV input, which is the reverse voltage the diodes will see, is over 21kV and if you are drawing less than 30mA then the transformer voltage will higher than the 15kV specification.
 

MisterBill2

Joined Jan 23, 2018
7,320
Post #7 is indeed correct, the peak voltage is probably a lot more than the RMS given. In addition, there can be some very large switching spikes when turning the transformer on and off. Such as 1.5 times the listed voltage.
 

shortbus

Joined Sep 30, 2009
8,056
If you read the description of the bridge rectifier and resistor shown in the schematic you will have a better understanding of why this is a load of c@p. This is one of those magnetic pulsers for quack medicine (I think) that has shown up recently. The so called Dr. Bob Beck Protocol.
 

DickCappels

Joined Aug 21, 2008
6,656
The diodes shown in the schematic are 75 kV strings of diodes, not 20kV diodes.

Another thing to consider is whether you have them connected correctly.
 

DickCappels

Joined Aug 21, 2008
6,656
That reminds me: The voltages you are using are not the kind of people use when learning electronics. You can easily be killed or maybe just maimed by accidentally coming into contact with the voltages in your circuit. At a minimum you should have a bleeder resistor across your capacitor and wait until the voltage bleeds down before touching anything. One place where we worked on high voltage, high current power supplies for smoke stack electrostatic precipitator we used a grounded hook at the end of a long (1 to 1.5 meter) insulated handle to short out various points in the circuit for safety purposes.

When running stuff like that, always keep one hand in your back pocket to reduce the chance of electric shock current passing through your heart.
 

jpanhalt

Joined Jan 18, 2008
10,525
When running stuff like that, always keep one hand in your back pocket to reduce the chance of electric shock current passing through your heart.
Hi Dick,
That is an often repeated claim. I could find no evidence to support that pathway theory based on experiments in dogs. I believe the real reason is that it reduces the chance of touching a live lead(s), at least in theory.
 

MisterBill2

Joined Jan 23, 2018
7,320
The circuit diagram mentions a string of 75 diodes, 1N4007. so the 20 KV prv diodes are no match. They are intended to be used in power supplies of 2000 to 3000 volts, where they provide an adequate safety margin. So use a half wave crectifier with all 4 diodes in series and they probably will not fail. But understand that you will have a cap charged to 20,000 volts and that can easily jump 2 inches to your hand and deliver one really bad shock that may leave you quite DEAD. And with a nasty spark burn as well.
And by the way, that magnetic pulse can be very uncomfortable, as I discovered in an MRI machine several years back. It made my bruises really hurt. The magnetism jerked on all of the iron in the pooled blood clots.
So that quack treatment can cause a hemorage inside your brain, very painful until you die.
 

Thread Starter

genekuli

Joined Oct 21, 2018
22
The diodes shown in the schematic are 75 kV strings of diodes, not 20kV diodes.

Another thing to consider is whether you have them connected correctly.
thank you, the thing is that the setup, as is, has worked for many hours, as is, so the way it is, works, but it just keeps eventually frying the one diode (when i replace it).
so considering that, would you think i could just have the diodes doubled up, in parallel, so that the one that keeps on frying would be 2 in parallel instead of just one?

That reminds me: The voltages you are using are not the kind of people use when learning electronics. You can easily be killed or maybe just maimed by accidentally coming into contact with the voltages in your circuit. At a minimum you should have a bleeder resistor across your capacitor and wait until the voltage bleeds down before touching anything. One place where we worked on high voltage, high current power supplies for smoke stack electrostatic precipitator we used a grounded hook at the end of a long (1 to 1.5 meter) insulated handle to short out various points in the circuit for safety purposes.

When running stuff like that, always keep one hand in your back pocket to reduce the chance of electric shock current passing through your heart.
absolutely, i do, you would probably laugh at how meticulously careful i am when touching this thing; i unplug, i discharge (x10) the cap bank, i touch every metal with a ground wire and i never touch the discharge output terminals
 

MisterBill2

Joined Jan 23, 2018
7,320
thank you, the thing is that the setup, as is, has worked for many hours, as is, so the way it is, works, but it just keeps eventually frying the one diode (when i replace it).
so considering that, would you think i could just have the diodes doubled up, in parallel, so that the one that keeps on frying would be 2 in parallel instead of just one?
It is far more likely that it is an over-voltage problem than over current. And if it is only one diode in the bridge, possibly it is too close to something conductive. And please consider the half-wave circuit with all four in series. It should charge to the same voltage but it may take a few seconds longer.
 

Thread Starter

genekuli

Joined Oct 21, 2018
22
The circuit diagram mentions a string of 75 diodes, 1N4007. so the 20 KV prv diodes are no match. They are intended to be used in power supplies of 2000 to 3000 volts, where they provide an adequate safety margin. So use a half wave crectifier with all 4 diodes in series and they probably will not fail. But understand that you will have a cap charged to 20,000 volts and that can easily jump 2 inches to your hand and deliver one really bad shock that may leave you quite DEAD. And with a nasty spark burn as well.
And by the way, that magnetic pulse can be very uncomfortable, as I discovered in an MRI machine several years back. It made my bruises really hurt. The magnetism jerked on all of the iron in the pooled blood clots.
So that quack treatment can cause a hemorage inside your brain, very painful until you die.
absolutely, i do, you would probably laugh at how meticulously careful i am when touching this thing; i unplug, i discharge (x10) the cap bank, i touch every metal with a ground wire and i never touch the discharge output terminals

thank you, the thing is that the setup, as is, has worked for many hours, as is, so the way it is, works, but it just keeps eventually frying the one diode (when i replace it).

so considering that, would you think i could just have the diodes doubled up, in parallel, so that the one that keeps on frying would be 2 in parallel instead of just one?

It is far more likely that it is an over-voltage problem than over current. And if it is only one diode in the bridge, possibly it is too close to something conductive. And please consider the half-wave circuit with all four in series. It should charge to the same voltage but it may take a few seconds longer.
the discharge rate frequency i want is 3Hz

i made sure it is away from all metal already, but it still fried
 

Thread Starter

genekuli

Joined Oct 21, 2018
22
the setup, as is, has worked for many hours, as is, so the way it is, works, but it just keeps eventually frying the one diode (when i replace it).

so considering that, would you think i could just have the diodes doubled up, in parallel, so that the one that keeps on frying would be 2 in parallel instead of just one?

If you add any diodes I would put them in series. You can also try adding a series resistor, 100 ohms, 1 watt
right.
in series to cope with the voltage,
right

right.
in series to cope with the voltage,
right
the diode is simply short-circuiting, because of the excess voltage and the arch then fries it, correct?
 

MisterBill2

Joined Jan 23, 2018
7,320
When the Peak Reverse Voltage limit is exceeded many diodes fail in breakdown, where suddenly lots of current flows in bot directions. That excess power dissipation causes the burn-up.
 

Thread Starter

genekuli

Joined Oct 21, 2018
22
interesting that it is only that one diode that has an issue (it is away from conductors) and is in mid air for cooling, and it fries a bit immediately then it still works for hours when i reduce the load by reducing the spark gap distance

When the Peak Reverse Voltage limit is exceeded many diodes fail in breakdown, where suddenly lots of current flows in bot directions. That excess power dissipation causes the burn-up.
so does that mean that parallel doubling up of the diodes would be the way instead of in series?

so how do i increase the capacity of the rectifier to withstand the excessive Peak Reverse Voltage (if that is the issue)? doubling diodes in series?
 

DickCappels

Joined Aug 21, 2008
6,656
I hope you realize that the design you have chosen was not well designed, rather that something thrown together to get an effect.

It get around failures of what is apparently rectifiers that are not rated at a high enough voltage, you put more identicle junctions in series.

In the case of a stack of 1N4007's to make sure that the voltage is shared equally among the diodes in a string you might want to put 47 pf across each diode. This forms a capacitive voltage divider that helps overcome (or as we sometimes say "swamp out") the variations in individual capacitance across the diode string.
 
Top