Need a snubbing diode with "Peak Forward Surge Current Rating" (greater than 400A) (for capacitive discharge)

Thread Starter

Hamlet

Joined Jun 10, 2015
448
I've done parametric searches at manufaturers websites and no option for searching "Peak Forward Surge Current Rating."

Diodes such as the 6A10 or 10A10 have ratings of 400A, but I wanted something better. I can not find anything close to this,
even with more exotic SiC or Schottky offerings. I've used 6A10, and 10A10, from different mfgs., but all blow eventually.

I am experimenting with high energy coils, with periodic capacitor discharges >1second. Voltages exceeding 800v. ("Electric Fence.")

As designed, it works on the bench, but not in the field. When the output is unloaded, I only get a few thousand "cycles"
before my diodes blow, and take out the triacs. On the bench, or fence, with a spark-gap load, it works forever. I can't employ a spark
gap while it is on a fence however (unloaded output) because then it would be a transmitter, and the hams & FCC would eventually come knocking. On the output, I am considering using a kv-rated capacitor, with suitable resistor, to soften the "blow"...

1612806845358.png
 

Thread Starter

Hamlet

Joined Jun 10, 2015
448
I would be surprised if such a thing existed. Why can't they be uses in parallel?
I've tried up to 4 diodes in parallel. They still blow. Maybe someone over at one of the can-crushing/tesla-coil forums have a trick. I'm dumping much energy in a short period, and the kick-back is tremendous if the secondary doesn't have anywhere to go...
 

Papabravo

Joined Feb 24, 2006
16,162
I've tried up to 4 diodes in parallel. They still blow. Maybe someone over at one of the can-crushing/tesla-coil forums have a trick. I'm dumping much energy in a short period, and the kick-back is tremendous if the secondary doesn't have anywhere to go...
How about diode connected FETs
 

jpanhalt

Joined Jan 18, 2008
11,088
For a fast discharge, very high pulse current diode, have you considered SCR's?

I used a T500108005AQ (https://www.galco.com/buy/Powerex/T500108005AQ) in a capacitor discharge welder. I don't fully understand your application, but you might use something consider that if you can wait until shut-off (zero crossing).

Opps: link doesn't work anymore. Let me look for a good link. Done
 

BobaMosfet

Joined Jul 1, 2009
1,776
@Hamlet In reviewing the data sheet for the 10A10:

1612822079476.png

10A10 has _10A_ for DC, and _600A for AC with NON-REPETITIVE surges no longer than 8.3ms. So that things can thermally cool- otherwise you're frying the component.

When you dump current quickly, you could be dumping 2 or 3 KiloAmps if it has nothing to resist it. How do you know how much current you're dumping? How have you measured it?
 

Thread Starter

Hamlet

Joined Jun 10, 2015
448
@Hamlet In reviewing the data sheet for the 10A10:

View attachment 229904

10A10 has _10A_ for DC, and _600A for AC with NON-REPETITIVE surges no longer than 8.3ms. So that things can thermally cool- otherwise you're frying the component.

When you dump current quickly, you could be dumping 2 or 3 KiloAmps if it has nothing to resist it. How do you know how much current you're dumping? How have you measured it?
With my thumb, haha.

The two (2), 2uF caps, and toroid inductor are there to slow the rush of energy, to save the life of the silicon.
It's not enough to blunt the dv/dt. It dosen't help that my transformer is almost electrically perfect. When there is a load, the energy has somewhere to go, and when the secondary is open circuit, everything just bounces back, hard.
 

metermannd

Joined Oct 25, 2020
191
What about something that would present just enough of a load so the load has somewhere to go until something makes contact with the fence?

Like a neon lamp or high-ohm resistor (several 22 meg resistors chained together) or somesuch...
 

MrAl

Joined Jun 17, 2014
8,251
I've done parametric searches at manufaturers websites and no option for searching "Peak Forward Surge Current Rating."

Diodes such as the 6A10 or 10A10 have ratings of 400A, but I wanted something better. I can not find anything close to this,
even with more exotic SiC or Schottky offerings. I've used 6A10, and 10A10, from different mfgs., but all blow eventually.

I am experimenting with high energy coils, with periodic capacitor discharges >1second. Voltages exceeding 800v. ("Electric Fence.")

As designed, it works on the bench, but not in the field. When the output is unloaded, I only get a few thousand "cycles"
before my diodes blow, and take out the triacs. On the bench, or fence, with a spark-gap load, it works forever. I can't employ a spark
gap while it is on a fence however (unloaded output) because then it would be a transmitter, and the hams & FCC would eventually come knocking. On the output, I am considering using a kv-rated capacitor, with suitable resistor, to soften the "blow"...

View attachment 229886
Hello,

Why are you playing with such low current diodes when you know you have a very high current you have to work with?
The 70HF80 diode is rated at 70 amps, 1000A peak, 350A repetitive surge, and i didnt even have to look far.
There are much higher rated diodes than that too so i am not sure what the problem is. Is it that you have very low space requirements? Cost?
When i worked in the power industry we used diodes like this and much higher on a somewhat regular basis. It is hard to blow them out usually something else goes first :)
These stud mount diodes need a heatsink also if you pulse them repeatedly.
This diode VS-VSUD360CW40 has 360 amp rating and almost 3000 amps single surge pulse rating.
The voltage may be too low at 400v but you can look for another diode with higher voltage rating.
So you get the picture: look around for a higher rated diode and it should solve the diode problem.
Also watch out for the reverse voltage peak due to open circuit or oscillations.
 
Last edited:

Thread Starter

Hamlet

Joined Jun 10, 2015
448
Hello,

Why are you playing with such low current diodes when you know you have a very high current you have to work with?
The 70HF80 diode is rated at 70 amps, 1000A peak, 350A repetitive surge, and i didnt even have to look far.
There are much higher rated diodes than that too so i am not sure what the problem is. Is it that you have very low space requirements? Cost?
When i worked in the power industry we used diodes like this and much higher on a somewhat regular basis. It is hard to blow them out usually something else goes first :)
These stud mount diodes need a heatsink also if you pulse them repeatedly.
This diode VS-VSUD360CW40 has 360 amp rating and almost 3000 amps single surge pulse rating.
The voltage may be too low at 400v but you can look for another diode with higher voltage rating.
So you get the picture: look around for a higher rated diode and it should solve the diode problem.
Also watch out for the reverse voltage peak due to open circuit or oscillations.
Thanks! No, I have no space constraints. I'm just living in the weeds, and haven't had a proper education or industry exposure to know what is available out there... I suppose I expected something available in a R6 package, or TO-220, TO-247, or something like that.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
448
What about something that would present just enough of a load so the load has somewhere to go until something makes contact with the fence?

Like a neon lamp or high-ohm resistor (several 22 meg resistors chained together) or somesuch...
I've blown big resistors just for fun. Small neon lamps just cry under that abuse. I have tried a 4foot florecent bulb to ground, and it seemed to help stretch the lifespan of the diodes/triacs.
 

MrAl

Joined Jun 17, 2014
8,251
Thanks! No, I have no space constraints. I'm just living in the weeds, and haven't had a proper education or industry exposure to know what is available out there... I suppose I expected something available in a R6 package, or TO-220, TO-247, or something like that.
Well it would be nice to find a TO-220 package that can handle the current and the power heating, but i think that package size is too limited. The TO-220 has less than 1 square inch surface area, and with a forward voltage at peak current that could be more than 1 volt, at 400 amps that is 400 watts easy. With that kind of power the surface area for a temperature rise of around 100 C would be more than 200 square inches, and so that little 1 square inch is only 1/2 percent of the required surface area and we would have to expect that 1 square inch could conduct heat fast enough to spread out the heat to 200 (or more) square inches. I seriously doubt that would work that's why bigger packages are used and stud mounts and so called chassis mount packages ... they can spread the heat out better.

And yeah some of them are more expensive too they can be much more expensive, but hey if you gotta do it you gotta do it :)
Maybe you can figure it this way...
If you get 100 'cycles' out of a 6 amp diode, imagine what you could get out of a 60 amp diode with similar other ratings and larger package. I would say it just has to be a lot more.

Some of these diodes are seriously tough characters too. I've seen over voltages blow out big gigantic parallel electrolytic capacitors that shoot their tops off like a cannon and bend thick copper buss bars while the big diodes dont even develop a runny nose :)
 

Thread Starter

Hamlet

Joined Jun 10, 2015
448
Well it would be nice to find a TO-220 package that can handle the current and the power heating, but i think that package size is too limited. The TO-220 has less than 1 square inch surface area, and with a forward voltage at peak current that could be more than 1 volt, at 400 amps that is 400 watts easy. With that kind of power the surface area for a temperature rise of around 100 C would be more than 200 square inches, and so that little 1 square inch is only 1/2 percent of the required surface area and we would have to expect that 1 square inch could conduct heat fast enough to spread out the heat to 200 (or more) square inches. I seriously doubt that would work that's why bigger packages are used and stud mounts and so called chassis mount packages ... they can spread the heat out better.

And yeah some of them are more expensive too they can be much more expensive, but hey if you gotta do it you gotta do it :)
Maybe you can figure it this way...
If you get 100 'cycles' out of a 6 amp diode, imagine what you could get out of a 60 amp diode with similar other ratings and larger package. I would say it just has to be a lot more.

Some of these diodes are seriously tough characters too. I've seen over voltages blow out big gigantic parallel electrolytic capacitors that shoot their tops off like a cannon and bend thick copper buss bars while the big diodes dont even develop a runny nose :)
It may be the solution I'm looking for. Remember, the diode only has to snub the current very briefly, and only every one or two seconds (1Hz). That's why I was thinking TO-220 would suffice, if rated high enough for surge.
 

MrAl

Joined Jun 17, 2014
8,251
It may be the solution I'm looking for. Remember, the diode only has to snub the current very briefly, and only every one or two seconds (1Hz). That's why I was thinking TO-220 would suffice, if rated high enough for surge.
Yes well experimenting with that tells us the story, and that seems to be that it isnt enough.
But did you try a higher rated TO-220 package anyway? I am pretty sure i had one with two diodes in it that were each rated for 30 amps. They were the Schottky type. Since there were two in the package and wire with their cathodes connected, it should be easy to wire them in parallel. Not that parallel is great but it's better than just one diode :)

Did you by any chance check the peak reverse voltage the diode might see? Sometimes stray inductances can really get ya when there are high currents involved.
 
Top