Nicon caps

[_JP_]

My PSU (no 80-plus cert.) has been whining for a couple of months now and 3 days ago scared the crap out of me when I walked in my room and there was a disturbing stench in the air.
Amongst other things, I eventually thought the caps had started leaking...
Since the computer was on stand by, I woke it up and everything was normal.
Still, I opened the computer up and took the thing out. Normal. No bulging caps, no leaking. Some dust, but nothing serious.
So I plugged it up again, voltages were normal, everything was fine. The whining was still there. Great.
According to what I've read, LC-Power PSUs usually have Huntkey internals (I know, bad rep. They're actually fine if you keep the load ~50W below the advertised). Not sure if mine is the case.
When I bought this PSU, I already knew that I would have to replace the caps in it eventually, but 4 years have passed and the thing never gave me a reason to rush to the electronics store....except the recent whining.
Do you guys know anything about Nicon caps and their reliability?
I've never seen or heard about Nicon caps and can't find info on them with the exception that they're made in China...

The whining is low-pitched, although more noticeable while the computer is running a game (or anything that taxes my 5770). It is also there when the computer is off, but barely audible. Yanking the cord (so to speak) for a while and then plugging it on again results in no whining until I power the pc on. This leads me to believe the whining comes from the secondary side. I can't tell if it comes from caps or coils (although some coils on the secondary side lack glue on them).

Should I waste any time on it or just go for a new PSU?
I'm looking at a XFX Core 550W. Best I can afford, atm.

 

[Frick]

Coil whine or capacitor whine is often high pitched I think. At least coil whine (which is the only kind I've actually heard for a longer time). But it might as well be the coils. Or something else. Glue wouldn't affect it.

If you end up getting something new the XFX Core PSU's are rather good actually.

 

[_JP_]

I'm a little rusty when it comes to bad capacitor diagnostics, but I seem to recall that when caps whine, the noise is random and it only tends to get worse with time, while coil whine varies with load/freq. going through the coil...which I think is my case. Still, because I'm not sure, I want to leave all options on the table.

 

[McSteel]

Actually, that specific PSU is made by China Great Wall Computer Shenzhen Co., Ltd. A solid piece of work, if you count it as a ~350W model. The caps are, as you may have suspected, abominable.

I've seen quite a few Nicons (too many, in fact) fail. After a bit of digging, it turned out they're a Fuhjyyu knock-off (who'd wanna copy those?!), and they're about a step below Chang and BH. They may not be bulging, but there's a good chance they're at least partially dried-up. Also, these caps sometimes leak at the bottom, through the shoddy rubber seal, as opposed to bulging up.

The PSU may be worth recapping if you have some spare caps lying around, but buying a cap set is only justified if you can get them dirt cheap. I'd suggest going for Panasonic FR, as they're quite good and pretty inexpensive, and they'll be miles above what's in there now. Also, you don't actually wanna go too low in ESR as you may introduce some unwanted hi-freq oscillations between the caps and the regulation coils.

Regarding coil whine, I wrote a wall of text on it recently, here. The XFX you were eyeballing is based on a modified Seasonic S12-II platform. It's pretty good, but there actually may be some better choices in it's price range, so if you're intent on buying a new PSU, why not link to the shop(s) you plan on visiting? Perhaps we could suggest some alternatives worth considering...

 

[de.das.dude]

i think u saw unicon

https://www.google.co.in/search?cli...rceid=opera&ie=utf-8&oe=utf-8&channel=suggest.


anyways my dad used to use philips capacitors in the audio amps he made and one such amp is around 25years old and still going strong. so they make (or atleast used to) good caps.

 

[Frick]

That wall of text was excellent, i learned a fair bit.

So could a splash of glue fix the whine, if it is the coils? And if youre willing to wait a bit ebay is a good start to buy cheap caps from. Or is it?

 

[de.das.dude]

btw, the wall of text got me thinking and explain a lot of things i see. i was aware of physical vibrations, but not the electrical part

i have seen a lot of the coils glued in or having heatshrink on them (big heatshrink over the whole coil). also is that why the VRMs on our motherboards have coils in those plastic cubes now?

 

[McSteel]

i think u saw unicon

https://www.google.co.in/search?cli...rceid=opera&ie=utf-8&oe=utf-8&channel=suggest.


anyways my dad used to use philips capacitors in the audio amps he made and one such amp is around 25years old and still going strong. so they make (or atleast used to) good caps.

Nah, it really is Nicon. They're increasingly popular in cheap PSUs (including single-voltage AC/DC adapters), slowly replacing the ~$0.05 more expensive BH, Chang, 12KJ, HEC, etc.

That wall of text was excellent, i learned a fair bit.

So could a splash of glue fix the whine, if it is the coils? And if youre willing to wait a bit ebay is a good start to buy cheap caps from. Or is it?

Absolutely. Non-conductive glue, hot plastic, non-corrosive caulk, whatever is able to dampen vibrations and isn't conductive or thermally reactive. A dab (or two) of glue inside the toroidal gap of a regulation coil works really well, but the upright filter inductors in pi-filters are best treated with heatshrink rubber tubing and gratuitous amounts of glue applied to their base.

## EDIT ##

btw, the wall of text got me thinking and explain a lot of things i see. i was aware of physical vibrations, but not the electrical part

i have seen a lot of the coils glued in or having heatshrink on them (big heatshrink over the whole coil). also is that why the VRMs on our motherboards have coils in those plastic cubes now?

The plastic cubes are there to hold the dielectric, which also does reduce vibrations. Outer shell of the "cube" actually contains metal or is made of metal, so that it provides EMI shielding to the coil. See here. It is also a convenient way to package coils for soldering and protecting them from deformations. The number written is inductance in micro-Henry ("2R0" = 2 uH, "1R5" = 1.5 etc).

 

[_JP_]

Actually, that specific PSU is made by China Great Wall Computer Shenzhen Co., Ltd. A solid piece of work, if you count it as a ~350W model.

The PSU I linked shares the manufacturer and model name only.
The rest doesn't match. Mine doesn't have an EPS connector, isn't 80-plus certified, has the 120v/230v switch....and some other stuff. It is 4 years old, lol.
Makes me think the manufacturer is different. I'll try to upload a photo of the internals. For now the computer is in use and I don't really have a lot of time.

The caps are, as you may have suspected, abominable.

I've seen quite a few Nicons (too many, in fact) fail. After a bit of digging, it turned out they're a Fuhjyyu knock-off (who'd wanna copy those?!), and they're about a step below Chang and BH. They may not be bulging, but there's a good chance they're at least partially dried-up. Also, these caps sometimes leak at the bottom, through the shoddy rubber seal, as opposed to bulging up.

Fuhjyyu knock-offs? How hasn't this thing failed to work yet??!? (Yes, I suspected they were utter crap)
I agree, I mean, I thought HEC was as low as you could go........to sell something presumably working out of the box. (and I've had my share of HEC removals)

The PSU may be worth recapping if you have some spare caps lying around, but buying a cap set is only justified if you can get them dirt cheap. I'd suggest going for Panasonic FR, as they're quite good and pretty inexpensive, and they'll be miles above what's in there now. Also, you don't actually wanna go too low in ESR as you may introduce some unwanted hi-freq oscillations between the caps and the regulation coils.

I was thinking between 0.1 and 0.15 ESR. I have no idea on the prices. I'd have to take the PSU apart to list the caps and then hunt for replacements. The I would check with a store I have close by, or farnell. I usually resort to Rubycons for my needs, but I'll consider those Panasonic as well...if I can find them here.

Regarding coil whine, I wrote a wall of text on it recently, here. The XFX you were eyeballing is based on a modified Seasonic S12-II platform. It's pretty good, but there actually may be some better choices in it's price range, so if you're intent on buying a new PSU, why not link to the shop(s) you plan on visiting? Perhaps we could suggest some alternatives worth considering...

That was delightful to read. Very informative. Thank you for the time you took to write it.

anyways my dad used to use philips capacitors in the audio amps he made and one such amp is around 25years old and still going strong. so they make (or atleast used to) good caps.

They do, but they're not useful for all situations.

Nah, it really is Nicon. They're increasingly popular in cheap PSUs (including single-voltage AC/DC adapters), slowly replacing the ~$0.05 more expensive BH, Chang, 12KJ, HEC, etc.

That is disturbing to hear. I've seen my share of dead units, including one I purposely blew-up *literally*, knowing it was the problem (the computer wouldn't start) and still pressed the power button until something happened. It was so bad, 2 caps (don't remember the brand, sorry) exploded and the PSU took nothing with it to the grave. Remarkable. It was that bad.

Let it be noted that I bought this PSU because it was the best I could afford at that time. It was carefully chosen to last me enough time to make the switch to a better unit (it ended up having to last twice as longer) and allow for some headroom for more hardware (since then I've gone from 1 to 3 HDDs and from a x1650PRO to a HD5770).
If I had had more greens in my pocket, I WOULD have bought something way better.
I always intended to replace it with something better and it shouldn't take long now...

Absolutely. Non-conductive glue, hot plastic, non-corrosive caulk, whatever is able to dampen vibrations and isn't conductive or thermally reactive. A dab (or two) of glue inside the toroidal gap of a regulation coil works really well, but the upright filter inductors in pi-filters are best treated with heatshrink rubber tubing and gratuitous amounts of glue applied to their base.

Will do, thanks!

 

[McSteel]

The PSU I linked shares the manufacturer and model name only.
The rest doesn't match. Mine doesn't have an EPS connector, isn't 80-plus certified, has the 120v/230v switch....and some other stuff. It is 4 years old, lol.
Makes me think the manufacturer is different. I'll try to upload a photo of the internals. For now the computer is in use and I don't really have a lot of time.

Internals pics would be great. No rush, of course... I'm interested because LC switched to GreatWall a while ago, I'd say even more than 4 years ago. You must have picked up some old stock... We'll see.

Fuhjyyu knock-offs? How hasn't this thing failed to work yet??!? (Yes, I suspected they were utter crap)
I agree, I mean, I thought HEC was as low as you could go........to sell something presumably working out of the box. (and I've had my share of HEC removals)

Well, as long as the caps are cooled properly and effectively, they will live much longer than usual. All caps heat up while working, for two possible reasons. One is inherent to the way they're built, and is dictated by a value called "tangent-delta", which is the power dissipation factor. This would be self-heating, and you can try to minimize it by varying the internal layout of the aluminum foils and other techniques, but you can't avoid it. Secondary heating comes from nearby heat sources, and is pretty simple and self-explanatory.

While crap caps do suffer from high tan-delta, and their electrolyte can't really survive the declared 105°C for much longer than a few hundred hours, they do live long if cooled properly. Actually, the lifespan of a capacitor is prevalently influenced by temperature. The formula that determines the relation between temperature and lifespan is very nearly logarithmic: for every 10°C of temperature difference, you apply a factor of 2 to the lifespan. So, if you have a capacitor with an MBTF (mean time between failures) of 2000 hours @ 105°C, that same capacitor will be expected to last 4k hours @ 95°C. Now try going down to, say, 65°C, which is more than realistic, and you end up with an estimated 32k hours of expected service. Even with the crap caps being capable of just about 1000 hours at 85°C, having a fan set to full blast (and in most crap PSUs they are simply directly connected to the 12V output of the PSU) will mean they get to see around 6k hours of lifetime before starting to degrade. That's about 3 years of 8 hours of work per day, every single day. Not too bad, actually...

Now, just for fun, take into consideration that most high-quality Jap-caps are rated at 5k+ hours at 105°C and do the math

 

[_JP_]

Another great read. Thanks!

Internals pics would be great. No rush, of course... I'm interested because LC switched to GreatWall a while ago, I'd say even more than 4 years ago. You must have picked up some old stock... We'll see.

More than 4 years ago? Er...then maybe it actually is a Great Wall unit. :\
I've also found that some LC-power branded PSUs were made by CWT. That would be an improvement. But then again, I don't recall seeing green wrapped coils.

Well, as long as the caps are cooled properly and effectively, they will live much longer than usual.

Well, the PSU does have a 140mm fan. That hasn't failed so far. *knocks on wood*

EDIT: It looks an awful lot like this one (going by the heatsinks and component layout)...which is made by Great Wall.

 

[_JP_]

Nicon caps. Nicon caps everywhere...

Time to revive this.
I got my new XFX PSU. It's awesome and it is powering the computer that I am using to type this.
About the LC-Power unit. Yes, it IS like the one I showed just above this post.
I took the cover off...and reassembled it again, because I don't have the caps. I'm also not sure if I want to go ahead with this, because I'm not having a lot of time right now to tinker with stuff just for heck of it. Took me forever to do this as you can notice...for example.

Spoiler: Anyway, pics:

So, nicon caps everywhere and a pair of HECs.
After reassembly, I tested it. Still makes noise. Both on and off. Also, PSU takes its time until the noise goes away after I turn it off from the mains.
Measurements with multimeter:
+12V: 11.71V
+5V: 5.12V
+3.3V: 3.34V
-12V: 11.43V
-5V: 5.24V

 

[itsakjt]

Just to share, my Zebronics 600W PSU also has Nicon caps. It also makes a slight whine noise when it is cold after I power on the computer. After some time, it goes away. But it was like that from the very first day.
The thing is that it is serving me quite good. I am running a highly overclocked Phenom II X4 955 and overclocked Radeon HD6770 as well with 2 HDDs, one ODD. Almost 2 years has passed and the performance of the PSU is the same as it was on the day I bought it.

 

[McSteel]

Well, it has been a while
Thanks for not forgetting about it. I'm glad to hear you are satisfied with your new PSU... I'm sure your PC is even more so

About the LC-Power you have there... The Nicons in it are, for the most part, unloaded. Meaning, they're used primarily as token (function-specific, for IC capacitive loading etc.) capacitors, rather than filter caps. This greatly reduces their stress level, so it's not too bad. The bulk of the filter duties is delegated to Teapo caps (as far as I can see - those green sleeve/gold markings caps with three-pointed star/dashes between rays scoring on top). I suppose they're LowESR "SC" or "SG" series. They can last 3-4 years in high-stress situations, and easily double that in a well-designed PSU. GreatWall generally makes PSUs that don't ask too much of their filter caps.

The noise that you mention, does it change pitch (rising with time) when the PSU is unplugged (from the wall)? If so, you're hearing resonant vibrations from the flyback transformer (the one closest to the wonderful turquoise common-mode filter choke on the right). Not much can be done about this, I'm afraid. You could try swapping those big primary HEC caps out and putting something with lower ESR (higher quality), then swapping out the bleeder resistor (the one soldered in between them) for one with higher resistance. But I suppose it's probably not worth it.

As for the voltages, that was measured with the PSU unloaded? I can't quite make out the minimum-load resistors on the secondary, but they might be poorly chosen, with the 5V high and 12V low indicating that 5V isn't quite loaded enough. Shouldn't be a problem once you connect the PSU to an actual system.

 

[_JP_]

Well, it has been a while
Thanks for not forgetting about it. I'm glad to hear you are satisfied with your new PSU... I'm sure your PC is even more so

Well, it sure hasn't complained.

About the LC-Power you have there... The Nicons in it are, for the most part, unloaded. Meaning, they're used primarily as token (function-specific, for IC capacitive loading etc.) capacitors, rather than filter caps. This greatly reduces their stress level, so it's not too bad. The bulk of the filter duties is delegated to Teapo caps (as far as I can see - those green sleeve/gold markings caps with three-pointed star/dashes between rays scoring on top). I suppose they're LowESR "SC" or "SG" series. They can last 3-4 years in high-stress situations, and easily double that in a well-designed PSU. GreatWall generally makes PSUs that don't ask too much of their filter caps.

The Teapo caps are "SC" series, 105ºC rated. And that's good to hear.

The noise that you mention, does it change pitch (rising with time) when the PSU is unplugged (from the wall)? If so, you're hearing resonant vibrations from the flyback transformer (the one closest to the wonderful turquoise common-mode filter choke on the right). Not much can be done about this, I'm afraid. You could try swapping those big primary HEC caps out and putting something with lower ESR (higher quality), then swapping out the bleeder resistor (the one soldered in between them) for one with higher resistance. But I suppose it's probably not worth it.

It does change pitch, but also "slows down" (lack for a better term) as it goes away.
It whines while on, plain noise while plugged in to the wall on standby, and when I switch it off, the noise begins to lower its frequency until it is gone (there's a moment while it goes "wow-wow-wow-wow-wow" ). Better get it on camera, if the microphone can pick it up.
The two big HECs seem easy to replace...I doubt I'll be able to get premium japanese caps here in a store...but I'll try.
The bleeder resistor seems to be a 820kΩ resistor. 1/1.5MΩ should do the trick...no?

As for the voltages, that was measured with the PSU unloaded? I can't quite make out the minimum-load resistors on the secondary, but they might be poorly chosen, with the 5V high and 12V low indicating that 5V isn't quite loaded enough. Shouldn't be a problem once you connect the PSU to an actual system.

Yes, the PSU was unloaded.
I do remember measuring the voltages when the PSU was installed and powering the computer, my motherboard was reading 11.7v and I was becoming concerned.
I like to keep a molex handy for these kind of situations. So I popped out the side panel and connected the multimeter. I remember the readings were around 12.2v and 4.9v, with the PC idle.

 

[de.das.dude]

sounds like the caps just take sometime to charge up.

i'd say never turn off the PSU. leave it connected to the mains. hardly consumes any power.

Nicon caps are definitely cheap since the local Indian manufacturer seems to use them as well. and there is no trace on the internets.

since you are a bit into electronics, you might want to re cap those with better caps later on.

Corsiar uses Panasonic/Matshusita caps which are in the top 5 reliable brands.
so you may want to look into those.

other good ones include, nippon, philips. etc.

 

[itsakjt]

sounds like the caps just take sometime to charge up.

i'd say never turn off the PSU. leave it connected to the mains. hardly consumes any power.

Nicon caps are definitely cheap since the local Indian manufacturer seems to use them as well. and there is no trace on the internets.

since you are a bit into electronics, you might want to re cap those with better caps later on.

Corsiar uses Panasonic/Matshusita caps which are in the top 5 reliable brands.
so you may want to look into those.

other good ones include, nippon, philips. etc.

Good info. You know where to find the Philips or Panasonic capacitors here in Kolkata? I may recap my PSU. Also will be any problem if I recap with a slightly higher capacitance capacitor like 2200 micro Farad instead of 2000 and with a higher temperature tolerance?

 

[McSteel]

It does change pitch, but also "slows down" (lack for a better term) as it goes away.
It whines while on, plain noise while plugged in to the wall on standby, and when I switch it off, the noise begins to lower its frequency until it is gone (there's a moment while it goes "wow-wow-wow-wow-wow" ). Better get it on camera, if the microphone can pick it up.
The two big HECs seem easy to replace...I doubt I'll be able to get premium japanese caps here in a store...but I'll try.
The bleeder resistor seems to be a 820kΩ resistor. 1/1.5MΩ should do the trick...no

You can rest easy using Teapo on primary as well. Samxon will also work nicely... Japanese are preferred, but not mandatory in these circumstances. 1-1.5MΩ will be just fine. Just watch the power rating - it would be nice to use at least a 0.5W resistor there, as you don't want it cooking your caps. As for the sound, probably part of it is coming from the PPFC coil, but what you described is the opposite of what I was expecting... Well, anyway, new caps and a new bleeder resistor should help. Don't be afraid of using a cap rated higher than 200V, if funds permit.

Yes, the PSU was unloaded.
I do remember measuring the voltages when the PSU was installed and powering the computer, my motherboard was reading 11.7v and I was becoming concerned.
I like to keep a molex handy for these kind of situations. So I popped out the side panel and connected the multimeter. I remember the readings were around 12.2v and 4.9v, with the PC idle.

Those readings are perfectly fine.

## EDIT ##

Good info. You know where to find the Philips or Panasonic capacitors here in Kolkata? I may recap my PSU. Also will be any problem if I recap with a slightly higher capacitance capacitor like 2200 micro Farad instead of 2000 and with a higher temperature tolerance?

Concerning recapping, more bulk capacitance for purely filter caps is always welcome. One should be careful not to overdo it (no more than 50% increase). When it comes to functional caps, you should respect the original spec as much as possible, to avoid throwing off resonant frequencies, or introducing unwanted oscillations, especially in the feedback circuits. In the filtering part, going too low in ESR could be detrimental to ripple filtering, as you can make the pi-filters miss their mark of ~100 Hz. If you upsize the caps, the inductor needs to follow.

 

[itsakjt]

You can rest easy using Teapo on primary as well. Samxon will also work nicely... Japanese are preferred, but not mandatory in these circumstances. 1-1.5MΩ will be just fine. Just watch the power rating - it would be nice to use at least a 0.5W resistor there, as you don't want it cooking your caps. As for the sound, probably part of it is coming from the PPFC coil, but what you described is the opposite of what I was expecting... Well, anyway, new caps and a new bleeder resistor should help. Don't be afraid of using a cap rated higher than 200V, if funds permit.



Those readings are perfectly fine.

## EDIT ##



Concerning recapping, more bulk capacitance for purely filter caps is always welcome. One should be careful not to overdo it (no more than 50% increase). When it comes to functional caps, you should respect the original spec as much as possible, to avoid throwing off resonant frequencies, or introducing unwanted oscillations, especially in the feedback circuits. In the filtering part, going too low in ESR could be detrimental to ripple filtering, as you can make the pi-filters miss their mark of ~100 Hz. If you upsize the caps, the inductor needs to follow.

Ok. I will keep that in mind.

 

[de.das.dude]

Good info. You know where to find the Philips or Panasonic capacitors here in Kolkata? I may recap my PSU. Also will be any problem if I recap with a slightly higher capacitance capacitor like 2200 micro Farad instead of 2000 and with a higher temperature tolerance?

all shops keep philips caps. a little bit here and there doesnt make a diff. just make sure the voltage is not less.
the higher the rated voltage is the better your life.


i wouldnt advice recapping a PSU that cant even supply its rated wattage.

 

[brandonwh64]

Does it look like the bottom of the first cap has popped open leaking out?

 

[GC_PaNzerFIN]

Does it look like the bottom of the first cap has popped open leaking out?

http://imageshack.us/a/img844/7513/sany0736.jpg

No that is the glue-like substance they use to secure the components on PCB in addition to soldering.

 

[_JP_]

sounds like the caps just take sometime to charge up.

i'd say never turn off the PSU. leave it connected to the mains. hardly consumes any power.

Nicon caps are definitely cheap since the local Indian manufacturer seems to use them as well. and there is no trace on the internets.

since you are a bit into electronics, you might want to re cap those with better caps later on.

Corsiar uses Panasonic/Matshusita caps which are in the top 5 reliable brands.
so you may want to look into those.

other good ones include, nippon, philips. etc.

Caps do take some time to charge up. But it's so fast, under normal conditions, you can't notice.
I turned it off, because I was testing it. As in, not installed in a system.
Also, if I'm not going to use a certain type of electronic device for a period of time longer than 3 days, I do usually turn it off from the wall. No power consumption at all is better that minimal.
I am planning in recapping the PSU, since I have no use for it, atm, and will probably do another low-power build with it or sell it, once the noise is gone...or at least lower in volume.
It's really a shame that it is making this noise, because otherwise this PSU would be dead-silent. It was before. Awesome for HTPCs.
My favorites are Rubycons. I know about the others, but Matshusita caps are rare around these parts, available Philips aren't usually for these kind of electronics and never saw a Nippon Chemi-con around here.

Also will be any problem if I recap with a slightly higher capacitance capacitor like 2200 micro Farad instead of 2000 and with a higher temperature tolerance?

Higher temperature, definitely, if you can. And I'd favor higher voltage to higher capacitance. There's a reason the capacitance value was defined as such and usually it is wise not to alter that.

You can rest easy using Teapo on primary as well. Samxon will also work nicely... Japanese are preferred, but not mandatory in these circumstances. 1-1.5MΩ will be just fine. Just watch the power rating - it would be nice to use at least a 0.5W resistor there, as you don't want it cooking your caps. As for the sound, probably part of it is coming from the PPFC coil, but what you described is the opposite of what I was expecting... Well, anyway, new caps and a new bleeder resistor should help. Don't be afraid of using a cap rated higher than 200V, if funds permit.



Those readings are perfectly fine.

## EDIT ##



Concerning recapping, more bulk capacitance for purely filter caps is always welcome. One should be careful not to overdo it (no more than 50% increase). When it comes to functional caps, you should respect the original spec as much as possible, to avoid throwing off resonant frequencies, or introducing unwanted oscillations, especially in the feedback circuits. In the filtering part, going too low in ESR could be detrimental to ripple filtering, as you can make the pi-filters miss their mark of ~100 Hz. If you upsize the caps, the inductor needs to follow.

Yes, I was planning on using either a 1W resistor (which is like the one already there) or 2W.
IIRC, not all Samxon caps are suitable for primary filters, only certain series...I do not remember which ones. I'll see what I can find.
Those readings were still within ATX spec, so yeah.

i wouldnt advice recapping a PSU that cant even supply its rated wattage.

Why not? As long as you know its boundries it is still safe to use.
I already knew this PSU would not handle 550W. I thought I would be lucky if I could use 400W tops, without blowing something up. So I didn't push it.

Does it look like the bottom of the first cap has popped open leaking out?

http://imageshack.us/a/img844/7513/sany0736.jpg

No. That's just snot-colored glue.

 

[de.das.dude]

recapping will be too tedious for something you dont plan on using for long time. caps arent that cheap. and you will end up spending a couple of hours unsoldering and soldering as you need to do it one type at a time or you risk screwing up.

even cheap ones come for about rs 15 a piece... considering 5mm med power leds are 5 a piece... its quiet a bit.

then the price increases with rarity as well. the rarer the rating the more the price they will demand for it.

 

[itsakjt]

Caps do take some time to charge up. But it's so fast, under normal conditions, you can't notice.
I turned it off, because I was testing it. As in, not installed in a system.
Also, if I'm not going to use a certain type of electronic device for a period of time longer than 3 days, I do usually turn it off from the wall. No power consumption at all is better that minimal.
I am planning in recapping the PSU, since I have no use for it, atm, and will probably do another low-power build with it or sell it, once the noise is gone...or at least lower in volume.
It's really a shame that it is making this noise, because otherwise this PSU would be dead-silent. It was before. Awesome for HTPCs.
My favorites are Rubycons. I know about the others, but Matshusita caps are rare around these parts, available Philips aren't usually for these kind of electronics and never saw a Nippon Chemi-con around here.

Higher temperature, definitely, if you can. And I'd favor higher voltage to higher capacitance. There's a reason the capacitance value was defined as such and usually it is wise not to alter that.

Yes, I was planning on using either a 1W resistor (which is like the one already there) or 2W.
IIRC, not all Samxon caps are suitable for primary filters, only certain series...I do not remember which ones. I'll see what I can find.
Those readings were still within ATX spec, so yeah.

Why not? As long as you know its boundries it is still safe to use.
I already knew this PSU would not handle 550W. I thought I would be lucky if I could use 400W tops, without blowing something up. So I didn't push it.

No. That's just snot-colored glue.

Yeah I will try then. Also the only drawback of my PSU is that the 3.3V rail is kinda weak. The voltage is never 3.3 but 3.26 idle and at load drops to 3.20 V. Any solutions for it?

recapping will be too tedious for something you dont plan on using for long time. caps arent that cheap. and you will end up spending a couple of hours unsoldering and soldering as you need to do it one type at a time or you risk screwing up.

even cheap ones come for about rs 15 a piece... considering 5mm med power leds are 5 a piece... its quiet a bit.

then the price increases with rarity as well. the rarer the rating the more the price they will demand for it.

No that will be not a problem for me. I am only concerned about the filter caps on the secondary side. So there will be hardly 10 caps I need to replace. Consider one being 20 Rs average, then also I will be able to do it in just around 200 Rs. And as for the soldering part, I am good enough.