UsiCase 330U-SCE

[ml2]

I made a post about my now apparently dead MSI K8 Neo4 motherboard and it was suggested to me that I post about the PSU here first with some pictures.

Basically, I turned on the PC and logged on, left the room, came back a few minutes later and the PC was dead. The system had been stable without any hint of what was coming.

The PSU in question is a Usicase 330U-SCE.

I opened the PSU up and found one cap with a bulged top. I noticed another one tonight starting to bulge (very slight). I can see no scorch or burn marks anywhere.

The caps are Fultec, Fuhjyyu and Vent. The failed two are a Fultec and a Vent, both down where all the pigtails join the board.

Where I've gotten a bit lost is in the behavior of the PSU after the failure. When the PSU is not plugged into the motherboard shorting the ps-on pin to ground turns the PSU on as it should. When the PSU is plugged into the motherboard (both 24 pin and 4 pin connector) shorting ps-on to ground causes the ps-on pin to initially go down to near zero and then shoot up to 14 volts. Would this behavior be indictive of failed caps or something more sinister?

I guess what we're trying to find out is did the PSU go bad all of a sudden and kill the motherboard in the process, or perhaps the PSU generally ok (just a couple of bad caps) and the motherboard is mostly at fault.

I guess the PSU might not be worth saving in the grand scheme of things but if I was able to fix it and the mobo to get another month or so out of them while I build a replacement PC that would be good.

Tim

[stj]

Re: UsiCase 330U-SCE

you have to keep those well mounted or they blow away in the wind!

the bad news is it's junk,
the good news is it wont cost mutch to replace all the caps!

[ChaosLegionnaire]

Re: UsiCase 330U-SCE

wtf it has only one pi coil for filtering the secondary side? really? dont u wanna split that in half for the other rails as well?! /end sarcastic post

[ml2]

Re: UsiCase 330U-SCE

At the request of another on my mobo thread I measured the PSU again.


PSU in standby:

12v = 0.4v (remaining bad cap in PSU?)
5v = 0v
3.3v = 0v
-5v = 0v
-12v = 15.1v **
5vsb = 5.19v


No load
12v = 12.06v
5v = 5.26v
3.3v = 3.429v
-5v = -4.94v
-12v = -11.39v
5vsb = 5.19v


2 HDD load
12v = 12.06v
5v = 5.26v
3.3v = 3.429v
-5v = -4.94v
-12v = -11.7v
5vsb = 5.19v

This shows something I missed before (bad me), the -12v line is doing polarity reversal. With the PSU in standby it's at positive 15v but when the PSU is turned on it goes negative (I think I have that the right way around, red DMM lead into blue, black dmm lead into black lead)

What would cause that? I guess the PSU is really junk now which is no problem, it cost me nothing 6 or so years ago.

Now I have to research a good value PSU.

[Dan81]

Re: UsiCase 330U-SCE

At the request of another on my mobo thread I measured the PSU again.


PSU in standby:

12v = 0.4v (remaining bad cap in PSU?)
5v = 0v
3.3v = 0v
-5v = 0v
-12v = 15.1v **
5vsb = 5.19v


No load
12v = 12.06v
5v = 5.26v
3.3v = 3.429v
-5v = -4.94v
-12v = -11.39v
5vsb = 5.19v


2 HDD load
12v = 12.06v
5v = 5.26v
3.3v = 3.429v
-5v = -4.94v
-12v = -11.7v
5vsb = 5.19v

This shows something I missed before (bad me), the -12v line is doing polarity reversal. With the PSU in standby it's at positive 15v but when the PSU is turned on it goes negative (I think I have that the right way around, red DMM lead into blue, black dmm lead into black lead)

What would cause that? I guess the PSU is really junk now which is no problem, it cost me nothing 6 or so years ago.

Now I have to research a good value PSU.

I know you may have heard this before but you could try and boot your K8N Neo4 using another power supply. I know this sounds quite stupid,but I had this happen with a Deer PSU (which was actually built pretty good except the sillicon,but that's going to be a whole another story - bye bye TT2164 transistors!)and using a modded Sun Pro (testing and such) actually worked and the MB booted. Check that your PSU has a -5v rail. I know this really sounds dumb but it seems that even though it's 939 it could still need -5v. (please people don't tell me about the -5v rail being removed story,I know it )

[momaka]

Re: UsiCase 330U-SCE

As much as I like fixing cheap PSUs, this is one that I wouldn't - at least not for powering a computer anymore. The PSU above has a tiny output toroid, not enough filter caps on the output, and worst of all, the 3.3V rail appears to be derived directly from the main transformer rather than regulated by a mag-amp or a linear regulator (this is pretty evident with the high readings of 3.4V for the 3.3V rail).

So all in all, I suggest not to use this PSU for powering a computer. It might be okay for running small DC motors, light bulbs, and other non-sensitive projects. Or you can use it for parts. Although this is a cheap PSU, at least it appears to have all of the input line filtering: two common-mode chokes, two X2-class safety caps between Line and Neutral, and four Y2-class safety. All of these are good parts that can be used to rebuild another cheap PSU that is perhaps built better on the output but lacks the input filtering. The PSU above also has an SB3040PT schottky rectifier, and that can be used to upgrade another PSU as well.

[Wester547]

Re: UsiCase 330U-SCE

^ Interesting. I was of the impression that it meant +3.3V was filtered using the coupled inductor that's part of the magamp circuit on the DC side (super group regulation...). Hence the high +3.3V reading. Of course it could be taken directly from the +5V transformer tap after the output tap is rectified and then filtered by the output capacitors or lack thereof. Unless I misunderstood your post altogether?

[momaka]

Re: UsiCase 330U-SCE

^ Interesting. I was of the impression that it meant +3.3V was filtered using the coupled inductor that's part of the magamp circuit on the DC side (super group regulation...)

If the 3.3V rail on this PSU used a mag-amp design, then the 3.3V rail would have had its own output toroid *and* also the mag-amp AC-side saturation toroid. Neither of these are optional for a mag-amp design.

The only time you see a single output toroid (what people often call here "supergroup regulation", which is misleading IMO) is if the PSU derives the 3.3V rail linearly (through a MOSFET, typically from the 5V rail) *OR* like in this case, through its own winding from the main transformer *without* any regulation whatsoever (now that is truly supergroup regulation... and its supper crappy ).

[Wester547]

Re: UsiCase 330U-SCE

If the 3.3V rail on this PSU used a mag-amp design, then the 3.3V rail would have had its own output toroid *and* also the mag-amp AC-side saturation toroid. Neither of these are optional for a mag-amp design.

The only time you see a single output toroid (what people often call here "supergroup regulation", which is misleading IMO) is if the PSU derives the 3.3V rail linearly (through a MOSFET, typically from the 5V rail) *OR* like in this case, through its own winding from the main transformer *without* any regulation whatsoever (now that is truly supergroup regulation... and its supper crappy ).

I don't think I'd call it group regulation, as there is no output inductor to regulate the +3.3V rail in that case. I'm a bit perplexed though, as to what this means for other PSUs (I know, I'm getting off-topic again...).

Like the linear regulator design in the older Hipro PSUs. As I understand it they have their own transformer tap, their own rectifier, their own winding in the output toroid, and their own output filter before the linear regulated +3.3V rail. So I'm not sure where the +5V and +3.3V combined rating on those PSU's label comes from unless it's just an "ATX mandated" standard (or unless that refers to how much power you can "waste" in the linear regulator before it burns out under a relatively high PSU load). I don't think they're referencing the limit of the main toroid or transformer.

Going back to the +3.3V magamp design though, I understand that the choke on the AC side of the heatsink is used to attenuate the AC emanating from the +5V transformer pins so it can bring the voltage down to +3.3V and thereafter is rectified to DC by the rectifier, then regulated by its own toroid and filtered by the output capacitors. Is this part of a separate topology from a forward converter or half bridge design (would a forward converter still need a freewheeling diode for the +3.3V magamp circuit to discharge the output inductor and preclude reverse voltage spikes)? In some half bridge PSUs with a +3.3V magamp design, there appears to be *a second* choke that looks identical to the AC side choke, but it's on the DC side, wedged between the two toroids. No idea what it's for.