Some time back (like year or year and half) my stepbro build a new PC for our friend Ivan. It feature reasonably fast AMD64 (at the time, of course) and good graphic card as well, as reasonable mainboard. Now recently Ivan complained that the maching is not turning on anymore, so it come here with it and we soon discover a bit bulged up one cap in the PSU. And since replacing the PSU with my Antec 480W NEO recapped one fix the problem and machine became reliable again, we know the problem is in the PSU.
My stepbro ended up bying another one and I was looking forward - when I get the caps and time - to recap this one to see how well it can work recapped - with slightly bumped capacitance in some places
The overal PSU design is not all that bad, but if they use as imput caps just pair of 560uF 200V caps - I feel that they cheated a lot with customers. 400W PSU deserve at least 2x 680uF 200V, if not 2x 820uF 200V. But surprisingly, that is not the worst thing this PSU show me inside.
Fuhjyyu caps! Now that is a red alarm, since they are the most inferior crap caps and they already destroyed once my DFI LP B mobo. We told Ivan how lucky he are, not to get damaged the mainbaord, but he did not seem appreciate his luck much.
Given the high number of Fuhjyyu caps he probably should thank his luck way, way more.
The other Fulltec caps are pretty bad ones as well...
So I started to replace the caps with quality Samxon ones. Even GC grade was used for the PSU:
At the end filtering caps I used the time proven capacitance upgrace from 1000uF 10V Fulltec to 3300uF Samxon GD.
On the imputs I used the proven, tested Samxon EP 820uF 200V as reasonable capacitancy bump to cope better with the 400W ratings.
To fully recap this PSU you need:
Eurocase PSU 400W
----------------------------
2x 560uF 200V d22.5 - 820uF Samxon EP 200V d22.5
1x 2200uF 16V d10 - 2200uF Samxon GD 16V d10
3x 1000uF 16V d10 - 3300uF Samxon GD 6.3V d10
3x 470uF 16V - 470uF Samxon GC 16V d8
1x 1000uF 10V - 1000uF Samxon GC 10V d8
1x 330uF 10V - 330uF Samxon GD 16V d8
1x 47uF 50V - 2x 47uF Samxon GF 50V d6.3
1x 22uF 50V
3x 1uF 50V - 3x 10uF Samxon GF 50V d5
Results exceed my expectations. First at all, there is NO NOISE at all from the PSU, it is dead-silent. No usual low intensity hi-frequency whispering, nothing. Perhaps because there is the big coil on the primary side? Neverless, the ripple results is shocking.
3.3V line ripple -1mV
5V line ripple - 1,5mV
12V line ripple 2,4mV
Please note that these results are the highest ripple spikes measured after the scope stabilized in few seconds. Usualy it is even lower... You can see for yourself, measuring 5V line:
I would recommend, hoewever, taking these measurments with grain of salt. The scope is measuring at the lowest end of it's sensitivity and I somewhat did not exactly trust the +/- 0.5% + 10 digits accuracy when measuring at the 4V AC range and quite close to the actual resolution of the scope itself - 0.001V (
http://www.atcweb.com/tpi440.htm ) ...
Neverless these results are very, very good. Probably best I ever reach. So, no more SMD ceramic caps experiments in PSUs
Also I found interesting thing. The Fuhjyyu caps are Fuhjyyu from one side, but VENT from another!
Hence I think that there is no difference between Fuhjyyu and VENT label and every caps marked with either one of them, should be treated as the utmost crap of caps - Fuhjyyu caps. Kill them with fire!
