Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Shorted diode in PC ATX power supply

Status
Not open for further replies.

Willen

Well-Known Member
I have three ATX power supplies manufractured by 'Perfect' and all are damaged. I found a diode connected to a switching transformer and that was shorted I replaced it with another low power diode and got ATX running. But I am sure the diode will not not survive with load any more. I tested on other two ATX and found same diode shorted. I think this model 'ATX-450W' has some design problems (at least choosing the diode). The shorted diode was HER207 (2A, 800V). To get a permanent solve maybe I need 3A diode of same family.

The circuit attached here is not exactly the ATX of PC, it's pretty similar SMPS and diode is in exactly in same place as I have. Or see this and see there D4, my shorted diode is there. Probably huge spike (Back EMF?) from transformer is breaking the diode again and again in the model. What could be the reason of short- higher voltage or higher current?

I have searched and found HER207 is high efficient diode, I don't know what type of high efficient it is than common silicone diode. I have no such high efficient diode with 3A (even 2A) so thinking alternatives. If the reason of short is 'High current' then could I use 1N5402 (low voltage 3A) etc or parallel of 1N4007 or BY39 (3A 800V)? (I am in doubt for parallel diode). Or do I need faster diode than 4000 or 5400 series? I saw some datasheets and cannot found the speed (frequency?) of diode rated there.
 

Attachments

  • Shorted diode.GIF
    Shorted diode.GIF
    11 KB · Views: 722
Last edited:
You need a faster rectifier, 'slow' ones like 1N5402 and 1N4007 will go S/C VERY quickly and are completely useless for SM applications.

Basically you're looking for a fast recovery Schottky rectifier - there are thousands of different ones suitable for you.

I can't even remember which ones I stock at work :D - basically I have them in two drawers, one labelled 'big fast diodes' and one labelled 'small fast diodes' - the exact diodes vary depending what I buy.
 
hi Willen,
Look at this option.
E
 

Attachments

  • uf5400.pdf
    72.7 KB · Views: 269
Hi, Thank you Nigel and erricgibbs,
My first curiosity is- may I know any more details about my problem? (like what was the diode there for? What is the reason of shorting etc)

2nd curiosity is- may I know about 'fast recovery' mean? I found in datasheet that the fast diode has very low junction capacitance around few pf and slower one has few hundreds of pf. I did not find more than this. So how to know the fast or slow diode?
 
So how to know the fast or slow diode?
Look at reverse recovery time.

There is very little average current in the diode. I usually use a 1A diode. The peak current is large.
uf5400 is 50 volt 3A. Use the 800-1000V version.
Also:
uf4007 is 1A. Probably should be 1000 or 800 volt.
With out seeing the thing work I don't know what the voltage on the diode is.
Both the current and voltage will be high in the first 100mS. or When the load is high.
----edited-----
If you really want to know what voltage is on the diode. (use a meter) Measure the voltage across R1C2 and add Vcc. OR Measure from the IC's ground to the output of the diode. The problem is that at startup for a very short time the voltage will be higher. It takes a storage o-scope to see that.
 
Last edited:
Hi again,
Luckily I found RL207 fast recovery diodes in a near city. But I am sure it will also burn out after few weeks or months. I think the design around the diode is poor. All SMPS of the same model has problem on the diode and burns frequently.

According the example schematic at #1 post of mine, the diode seems like a protection diode (same as a diode across a relay coil). Sorry I wrote '1uF' in the schematic, it's actually 0.1uF, 400V . IC 'UC3843A' creates 500KHz and the MOSFET drives the transformer. So the 0.1uF capacitor acting there as a 3 ohms series resistors? (capacitive reactance of 0.1uF+500KHz is around 3 ohms.) And then a 22k 1watt resistor is a bleeder resistor for the capacitor?

I guess I can modify the R1 or C1 to solve frequent diode burning problem. How you think? (I didn't get more powerful diode than HER207 so thinking like this way)
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top