I have a circuit that I am trying to troubleshoot.

It is supposed to output 1400 Volts DC.

So there is an AC input that I measured with a high voltage probe and a simple multimeter, at 2000Volts.

This AC output is connected to a diode bridge . The diodes keep sparking when I power up the system and smoke comes out of it.

The diodes are rated at 1600 Volts. Should the diodes rating be lower than the AC voltage that you input them with?

Could there be a problem with the power transfomer? Is it possible that a faulty power transformer gives out more voltage than it normally should?

 

The diode voltage rating should be larger than the peak voltage of the AC input. For 2000VAC you should have diodes rated for a least 3000V

 

Thanks for the reply. Apparently the hi voltage probe was faulty. Took the transformer to a pro, and he measured it at 1000 Volts AC.

So apparently something is wrong with the circuit.

Can a diode test ok with the diode tester and still be faulty?

 

Thanks for the reply. Apparently the hi voltage probe was faulty. Took the transformer to a pro, and he measured it at 1000 Volts AC.

So apparently something is wrong with the circuit.

Can a diode test ok with the diode tester and still be faulty?

It's possible for the diode to test ok with a diode tester but have a low reverse breakdown voltage which would fail in the high voltage circuit. But that is not very likely. Generally when diodes fail, they fail short.

 

I am missing something?

The dc out out is supposed to be 1400VDC therefore you need 1000VAC at the bridge input.

Took the transformer to a pro, and he measured it at 1000 Volts AC.

So apparently something is wrong with the circuit

I presume the transfo output is 1000VAC, however, what was his input to the transfo and is it the same input voltage as what you had before you took it somewhere for testing?

1600 volt diodes on 1000vac(1400 peak) is pretty close.

1. Whats the number and manufacturer of your diodes?
2. Where are you getting them?
3. Are you replacing these diodes after they smoke?
4. What is the current your expecting?
5. Is this a homemade circuit?
6. Are you sure of the proper orientation of the diodes.
7. What's the input voltage on the transfo?

 

I am missing something?
I presume the transfo output is 1000VAC, however, what was his input to the transfo and is it the same input voltage as what you had before you took it somewhere for testing?

The input voltage was coming from 220V in both cases.



1. Whats the number and manufacturer of your diodes?

4x P513 , 1600Volts. Diode Bridge. made out of those 4.

2. Where are you getting them?

Local Electronics Shop-littlediode.de

3. Are you replacing these diodes after they smoke?

No I didn't actually since they tested ok with the multimeter. But maybe this test is not enough?

4. What is the current your expecting?

No clue from the schematics.


5. Is this a homemade circuit?

No it is a scientific instrument Power Supply PCB.

6. Are you sure of the proper orientation of the diodes.

YES

7. What's the input voltage on the transfo?

2x110Volt AC is the input voltage

Thanks for the response

 

The p513 are a basic 1 amp diode that can take 1600 volts. Don't appear to be any thing special apart from the voltage.

Can you post the schematic? If the load is drawing more current than allowed then one needs to disconnect the load to see if the sparking still occurs, however, sparking is not healthy and may have already damaged those diodes.

Your first post you mentioned the diodes sparking and smoke coming out of the system. Are the diodes sparking between/across each other or just themselves along the body and where is the smoke coming from?

The diode test from a meter only applies a couple volts and can trick you in a high voltage scenario. Depending on the output voltage of the meter, I have found flaky diodes by checking the resistance both forward and back out of the circuit against a known good diode.

Diodes are cheap and if I had new ones in my hands, well, I would replace all four and try the smoke test again, first with the load disconnected and then with the load.

 

any ideas?