Codegen 350XA supply

[milesy]

HI Guys

hoping to get a bit of guidence here...i have the above PSU which stopped working a while back.

Purple Wire has 5v
Green Wire sits at 4.3 volts before startup

With only the 24way ATX plug connected i can press the start button and the PS starts giving 5.38v on the red and 3.3V on the orange, however on the yellow wire i only get 11.0V

If i plug in the 4 way (yellow + black) to the MB the yellow wire drops to 5.27v and the red wire INCREASES to 6.72V......its not the MB as another supply works fine.....any ideas where i start ??

Adrian

[c_hegge]

Re: Codegen 350XA supply

Junk the PSU. Codegen are junk, and should NEVER be purchased, sold or used in PCs.

EDIT: and this is why - http://www.youtube.com/watch?v=xYB9WkiZ9SQ

[milesy]

Re: Codegen 350XA supply

fair enough but im sort of curious as to what would cause the fault im seeing....just the 12v wrong?

[c_hegge]

Re: Codegen 350XA supply

Could be bad caps. When they go bad, it can cause problems with the feedback circuit, which can cause all kinds of unusual behavior, as well as the obvious high ripple.

[milesy]

Re: Codegen 350XA supply

ok thanks.....so any idea which particular caps i would be looking at.....im pretty sure the large 470u input caps are ok and one of the others look expanded at all?

[rogfanther]

Re: Codegen 350XA supply

Well, you really should replace all of the caps of the secondary . But as c_hegge said, the psu is not worth it. It is ok if you are doing it to learn, but do not trust some piece of hardware you like to it.

[milesy]

Re: Codegen 350XA supply

hope you guys can bare with me here.....i have had a really good look at all the caps (visually) and i cannot see any that even look suspect - no expanded etc et......given that myonly problem with this supply is the 12v rail would i be just looking at any caps that are connected to the 12v rail or is likely to be something to do with the switching controller???

i realise its a bit hit and miss without being able to see the supply but if you can offer me any help please do

[c_hegge]

Re: Codegen 350XA supply

Knowing Explodegen (Codegen), the caps are probably HEC brand (which stands for Horrible Electrolytic Capacitor). Like any junk cap, they can go bad without visibly showing it, but yes, it is also possible that the problem is elsewhere.

[selldoor]

Re: Codegen 350XA supply

i realise its a bit hit and miss without being able to see the supply

How about some clear well focussed no flash pictures then?

[milesy]

Re: Codegen 350XA supply

yep photos i can do..........


IMGP9561.jpg by Adrian Miles Photography, on Flickr


IMGP9562.jpg by Adrian Miles Photography, on Flickr


IMGP9563.jpg by Adrian Miles Photography, on Flickr

plus i think i found the schematic of the supply........


codegen_atx_300w_m by Adrian Miles Photography, on Flickr

[milesy]

Re: Codegen 350XA supply

anyone?

[c_hegge]

Re: Codegen 350XA supply

As I suspected, it has HEC Caps. Try replacing them

[rievax_60]

Re: Codegen 350XA supply

HI Guys

hoping to get a bit of guidence here...i have the above PSU which stopped working a while back.

Purple Wire has 5v
Green Wire sits at 4.3 volts before startup

With only the 24way ATX plug connected i can press the start button and the PS starts giving 5.38v on the red and 3.3V on the orange, however on the yellow wire i only get 11.0V

If i plug in the 4 way (yellow + black) to the MB the yellow wire drops to 5.27v and the red wire INCREASES to 6.72V......its not the MB as another supply works fine.....any ideas where i start ??

Adrian

There might be one phase missing at the 12v rectifier.
Check for any bad solder connections from the transformer secondary to the 12v diodes, like solder cracks at the diode pins.
Also check for any open diodes

[milesy]

Re: Codegen 350XA supply

thanks guys..........well as i didnt have any replacement caps i thought i would investigate the open diode caper.......it made sense given that the 12v under load was only reading 6v - and if one of the diodes was open the 12v output cap would only get half the amount of power rather than both +ve and -ve pules.....

so i removed the heatsink that contains two diode packs (S20C40C) and 3 discreet diodes (photo below) unfortunatley all seems to be ok with this lot of components...removing this does allow me to see a little better though and there is still no obviously bad caps.

if there are bad caps - am i likely to be able to test them somehow???


IMGP9564.jpg by Adrian Miles Photography, on Flickr

[c_hegge]

Re: Codegen 350XA supply

Diodes on a bracket!!

Replace the two diode-bracket with a proper rectifier. one of them could be toast

[milesy]

Re: Codegen 350XA supply

You think the diodes could be faulty under load even though they test ok with a DMM??

[rievax_60]

Re: Codegen 350XA supply

I have seen axial diodes go intermittent open, very rare though. You might pick it by a fluctuating meter reading while the diode is being gently flexed.
Cracked solder joints can be difficult to spot. Maybe you should check some more like at the transformer pins and the bracket the diodes are soldered to and just try the power supply again.
There should be continuity between the pads that the diode anodes solder to. The path is through the secondary winding.

[milesy]

Re: Codegen 350XA supply

Thanks I will recheck..am I right in saying the symptoms definitely seem to be only getting half of the cycle through to the 12v line?

[rievax_60]

Re: Codegen 350XA supply

Thanks I will recheck..am I right in saying the symptoms definitely seem to be only getting half of the cycle through to the 12v line?

Yes, I can't think of any other possible reason.

[c_hegge]

Re: Codegen 350XA supply

Thanks I will recheck..am I right in saying the symptoms definitely seem to be only getting half of the cycle through to the 12v line?

Sounds about right to me too.