The thread about hFE got me thinking about seemingly simple things that I don't understand.

Years ago I acquired a bushel of MDA2500 .
25amp 50piv rated.

Most every one tested @ 1000 volts with almost no reverse current. Less than 10uA.
It's been years, so I don't remember if that was per diode or thru bridge.

Since then I've used them on innumerable hobby projects. Some even mains powered.

Never a failure.
Anything critical I still use a proper piv rated component.

My question then is:
Are diodes made specifically 50piv?

Or are they produced in a batch and tested for labeling and sale?

What is the difference for a hobbyist, between A 400piv rated and a 50piv tested @ over 400volts?

Aging, heat, breakdown with high current? Failure to recover from a spike?

 

when they make a batch of rectifiers, they test them and the ones with the highest voltage tested are labled as that, it used to be that way anyway. when they had a big order for 50 volt ones, they used what they had. years ago, I decidec to stop stocking 1n4001, through 1n4006, only 1n4007, cut down on inventory, and the circuits didnt even know they were 1000 volt instead of 50 or 200. they might have improved gtheir techniques so they actually make 50 volt diodes by now, but it dosnt seem like it.
also, did you ever wonder why so many transistors were sequentially numbered too? same thing, they made a batch and labled them as what they test as.sometimes the standard batch came out with higher frequency or gain, and they got the numbers that coresponded.

 

Would you conclude that testing reverse leakage, at a voltage higher than will be needed, is a valid test to classify these?

I once thought it was, and sold some at swap meets. Listing test results.

That is until chastised be electronic engineers. Who of course wouldn't explain.

Since then I've thought maybe the 50piv rating was designed. That testing higher couldn't change the safe rating.

 

In Shottky diodes, the PIV has a lot to do with the forward voltage per amp. Maybe 50 years ago there was a significant yield per batch in 1KV silicon rectifiers. Apparently, not so much now. When 1N4007 became the same price per hundred as 1N4001, I quit buying anything except 1N4007 (ECG125).

I expect that it is valid to test for reverse voltage at some tiny number of microamps. That's how the big boys do it. I can't seem to feel a need to do that when I'm getting KV rectifiers at $5/100.

 

I'm in opposite situation.

I don't see the need to buy 400 volt bridges when I still have a shoe box full of 50 volt versions that may be fine at much higher levels.

Sometimes they were all I had. Other times I just forgot. (the rating)

I have some of them in very voltage inappropriate situations over the years.

Unless I hear different, I'm going to treat them as they test.

I knew they were ok, but I don't know what I don't know!

 

A real test result is worth a dozen theories.

 

I either learned something in the last 20 years or my memory is failing.

As I don't trust my memory I tested a few more.
Honestly, don't remember how I did it before. It's obvious now to test each diode and rate at the minimum.

First few tests were over 1000 volts as I remembered but one diode out of the bridge was 800.

After a few more I found minimums as low as 200.

 

Use a datasheet to find the test current. 5 ua for 1N4004 at 25 C.
Set up a KI generator at 5 ua and read the voltage supported by the diode.

 

My set up is a 2kv dc supply (2ma max) on a variac and uA meter in series.

Small region between slight meter movement and full scale.
Well 50v or so, but I'm not splitting hairs.

I only test to 1Kv because that's the limit of my handy bench meter.

 

when they make a batch of rectifiers, they test them and the ones with the highest voltage tested are labled as that, it used to be that way anyway. when they had a big order for 50 volt ones, they used what they had. years ago, I decidec to stop stocking 1n4001, through 1n4006, only 1n4007, ...
...

Agreed! I only stock 1N4007 (1000v rated), and you can buy bags of 100 or 500 quantity etc for a handful of dollars. I think from memory last time I got a bag of 500 1N4007 for $7 or so.

No need to mess with 1n4001 or 4004 ever again, I just throw them out.