Tetrium wrote:

I have a question about an old PSU I opened up a couple days ago (It's still in my living room actually). […]
Show full quote

I have a question about an old PSU I opened up a couple days ago (It's still in my living room actually).

Most of the time I kinda know what I have in hands (I'll add right away that I'm not at all as knowledgeable as most of the others here in this thread, but I do have some basic knowledge about what's good and what's not so good) but with this PSU I kinda have some doubts which I will try to explain.

This PSU came out of a P4 I got at a thrift store a month or 2 ago (Northwood RDRAM, just so you can place it chonologically a bit) and wasn't anything special (just some lower-end AGP card, single harddrive, optical drive, etc).

At first I kinda had high hopes it would be a better quality unit as the PSU fan exhaust grill looked very much like the ones FSP models used a lot in those days.
However...
The very first thing I did when I got the PSU out (the label wasn't readable due to part of the case obscuring it, that's why I couldn't ID the PSU before I removed it from the case) was to read the brand and to check the label.
Brand: Sam Cheer. Could've been better and as I knew very little about this brand (found maybe 1 or 2 other Sam Cheers in like 10 years so I didn't know a lot about them from top of my head).
Label: The print on the label looked very similar to the ones used by FSP models of the same era but the model number was this one: CWT-300ATX12 (PFC) and CWT was Channel Well Technologies (iirc), which was somewhere below FSP I think, but not on the bottom.
Stats on the label: 3.3V 14.0A, 5V 30.0A, 12V 15.0A, +5Vsb 2.0A, -5V 0.5A, -12V 0.5A

And for some reason a PSU with a colory sticker on it which includes the word "Pentium 4" always raises my suspicion 🤣

Anyway, I opened it up and 1 of the caps had leaked pretty bad, the rest of the PSU seemed to look alright what damage is concerned (so no burn marks on the PCB or anything, not a lot of dust) but I have not checked it thoroughly as I typically never bother when I see theres leaking caps inside.

One more thing I did find obvious, and that's how tiny the 2 largest caps looked. The typical ones are a good portion of the height of the PSU but these looked like they were dwarfs...too bad.

But anyway, would it be worth it to repair this unit? Main reason is of course it's stated 30A and that the rest of the unit didn't seem to be used a lot.

I can post pics if this could help, but these would need to be taken with a mobile phone 😵

Interested in hearing opinions about this PSU 😀

edit:
Almost forgot, this particular PSU also has one of these extra motherboard power connectors that looks a lot like 1 of the 2 plugs used on typical AT-class PSUs, so that alone might be a reason to repair it?

Depending on where you take it this also could be it's own topic.
A better camera would help... 🤣

Channel Well Technologies is highly variable in so far as the design.
They build a lot of PSUs to order for other companies so the designs range from so-so to high quality.
They do (or did) typically use crap caps even in their better designs. (Good PSUs after a recap.)
When looking at PSU caps became 'a thing' among enthusiasts some of CWT's clients started specifying what caps to use so some CWT models have all Jap caps.
.
[When looking at PSU caps became 'a thing']
I think I played a part in making that happen. I got tired of PSU review sites not mentioning caps at all so I went on a mission and started writing the editors and authors (a lot of them on a lot of sites) telling them their reviews sucked because they were incomplete for not mentioning a critical part with a high failure rate. I don't remember all of them but i had direct communications regarding PSU caps in reviews with JonnyGURU of JonnyGURU.com (now retired as site webmaster), Oaklahoma Wolf (current master of the JonnyGURU site) and Gabriel Torres of Hardware Secrets. This was all a very long time ago when Hardware Secrets was a fledgling site, Gabriel was the the only staff member IIRC and Hardware Secrets didn't even have a PSU tester yet. (IIRC later they acquired JonnyGURU's old tester when JonnyGURU upgraded theirs.) JonnyGURU and Oaklahoma Wolf are (or were) both members over at Badcaps.net so sometimes I saw them over there as well.
A few of the people I bitched at started including capacitor information in their reviews and many other sites followed suit soon after because having the best reviews is a competitive thing for many people that do it professionally.
Even before they had a tester Gabriel Torres of Hardware Secrets was doing some of the best reviews ever. I think he was the first to disassemble PSUs and look at them component by component. Other reviewers followed suit with that too but Hardware Secrets is still doing a great job so I always remember to check them (in addition to JonnyGURU) to see if they have reviewed some PSU I'm interested in.
.

PCBONEZ wrote:

Depending on where you take it this also could be it's own topic. A better camera would help... :lol: […]
Show full quote

Tetrium wrote:

I have a question about an old PSU I opened up a couple days ago (It's still in my living room actually). […]
Show full quote

I have a question about an old PSU I opened up a couple days ago (It's still in my living room actually).

Most of the time I kinda know what I have in hands (I'll add right away that I'm not at all as knowledgeable as most of the others here in this thread, but I do have some basic knowledge about what's good and what's not so good) but with this PSU I kinda have some doubts which I will try to explain.

This PSU came out of a P4 I got at a thrift store a month or 2 ago (Northwood RDRAM, just so you can place it chonologically a bit) and wasn't anything special (just some lower-end AGP card, single harddrive, optical drive, etc).

At first I kinda had high hopes it would be a better quality unit as the PSU fan exhaust grill looked very much like the ones FSP models used a lot in those days.
However...
The very first thing I did when I got the PSU out (the label wasn't readable due to part of the case obscuring it, that's why I couldn't ID the PSU before I removed it from the case) was to read the brand and to check the label.
Brand: Sam Cheer. Could've been better and as I knew very little about this brand (found maybe 1 or 2 other Sam Cheers in like 10 years so I didn't know a lot about them from top of my head).
Label: The print on the label looked very similar to the ones used by FSP models of the same era but the model number was this one: CWT-300ATX12 (PFC) and CWT was Channel Well Technologies (iirc), which was somewhere below FSP I think, but not on the bottom.
Stats on the label: 3.3V 14.0A, 5V 30.0A, 12V 15.0A, +5Vsb 2.0A, -5V 0.5A, -12V 0.5A

And for some reason a PSU with a colory sticker on it which includes the word "Pentium 4" always raises my suspicion 🤣

Anyway, I opened it up and 1 of the caps had leaked pretty bad, the rest of the PSU seemed to look alright what damage is concerned (so no burn marks on the PCB or anything, not a lot of dust) but I have not checked it thoroughly as I typically never bother when I see theres leaking caps inside.

One more thing I did find obvious, and that's how tiny the 2 largest caps looked. The typical ones are a good portion of the height of the PSU but these looked like they were dwarfs...too bad.

But anyway, would it be worth it to repair this unit? Main reason is of course it's stated 30A and that the rest of the unit didn't seem to be used a lot.

I can post pics if this could help, but these would need to be taken with a mobile phone 😵

Interested in hearing opinions about this PSU 😀

edit:
Almost forgot, this particular PSU also has one of these extra motherboard power connectors that looks a lot like 1 of the 2 plugs used on typical AT-class PSUs, so that alone might be a reason to repair it?

Depending on where you take it this also could be it's own topic.
A better camera would help... 🤣

Channel Well Technologies is highly variable in so far as the design.
They build a lot of PSUs to order for other companies so the designs range from so-so to high quality.
They do (or did) typically use crap caps even in their better designs. (Good PSUs after a recap.)
When looking at PSU caps became 'a thing' among enthusiasts some of CWT's clients started specifying what caps to use so some CWT models have all Jap caps.
.
[When looking at PSU caps became 'a thing']
I think I played a part in making that happen. I got tired of PSU review sites not mentioning caps at all so I went on a mission and started writing the editors and authors (a lot of them on a lot of sites) telling them their reviews sucked because they were incomplete for not mentioning a critical part with a high failure rate. I don't remember all of them but i had direct communications regarding PSU caps in reviews with JonnyGURU of JonnyGURU.com (now retired as site webmaster), Oaklahoma Wolf (current master of the JonnyGURU site) and Gabriel Torres of Hardware Secrets. This was all a very long time ago when Hardware Secrets was a fledgling site, Gabriel was the the only staff member IIRC and Hardware Secrets didn't even have a PSU tester yet. (IIRC later they acquired JonnyGURU's old tester when JonnyGURU upgraded theirs.) JonnyGURU and Oaklahoma Wolf are (or were) both members over at Badcaps.net so sometimes I saw them over there as well.
A few of the people I bitched at started including capacitor information in their reviews and many other sites followed suit soon after because having the best reviews is a competitive thing for many people that do it professionally.
Even before they had a tester Gabriel Torres of Hardware Secrets was doing some of the best reviews ever. I think he was the first to disassemble PSUs and look at them component by component. Other reviewers followed suit with that too but Hardware Secrets is still doing a great job so I always remember to check them (in addition to JonnyGURU) to see if they have reviewed some PSU I'm interested in.
.

Thanks for this info, maybe the fluctuating quality of CWT PSUs is what is making me have doubts about this particular PSU 🤣
And thanks for the backstory, I always find these fun to read 😀

wave wrote:

Please post some high resolution well focused pictures of tne psu

And yes, I know how important/handy good and clear pics are 😁! Perhaps I can get a family member to part with an older camera as I don't have the funds to get a new camera atm, but for now my mobile will have to do (the pics may still be usable if I don't move the "camera" too much while taking pics).

I will take some pics of it (especially it's internals and anything which may include details that may be interesting) but have some RL stuff to do first 😀

Pics (part 1 of 2)

It was really hard to get clear shots, even had to use a flashlight because otherwise the pics would've been even more useless.

Here's the first batch of pics

The VRM thingy (I know there are 3) has this reading:
ST
3116 or 3118
MOROCCO
N40Nf0

The other 2 also seem to have MOROCCO on it and the 3rd one seems to be fairly larger with the 2nd one being kinda identical in shape, but it did have a different reading. I could'nt make out the reading on the other 2 VRMs though as the font was just too small for me to read

^ Can't see much but doesn't look too bad (save for crappy Crapxons).

The oozing cap is on +5VSB, don't use this PSU without recapping, it may nuke the mobo.

This is an old design. The add on board is probably active pfc.

The psu seems to have a full AC filter.

The blurriness of the photos is too damn high 😜

Does it work if you short green to black with a paperclip?

TELVM wrote:

^ Can't see much but doesn't look too bad (save for crappy Crapxons).

The oozing cap is on +5VSB, don't use this PSU without recapping, it may nuke the mobo.

That's not good, it may indicate the board that came out of the same system this PSU was part of, may already be dead. But oh well.

Any idea whether it's possible to visually spot motherboard 5Vsb damage?

wave wrote:

This is an old design. The add on board is probably active pfc. […]
Show full quote

This is an old design. The add on board is probably active pfc.

The psu seems to have a full AC filter.

The blurriness of the photos is too damn high 😜

Does it work if you short green to black with a paperclip?

I haven't tried to power it up, I typically don't bother if I see the PSU is already damaged

And about the pics, this is literally the best I could make of it, I even used a LED flashlight to kinda increase the visibility in the pics but with this old mobile it's not of much use.

And thanks for the info, I guess I will be storing this one for now, with proper labels to not use it till it's repaired of course 😀

Tetrium wrote:

Any idea whether it's possible to visually spot motherboard 5Vsb damage?

It is sometimes visually evident by shinny spots or pits on the chipset.
Anything that looks like melting or popping happened inside the package.
It's nowhere near 100% though.

PCBONEZ wrote:

It is sometimes visually evident by shinny spots or pits on the chipset. Anything that looks like melting or popping happened in […]
Show full quote

Tetrium wrote:

Any idea whether it's possible to visually spot motherboard 5Vsb damage?

It is sometimes visually evident by shinny spots or pits on the chipset.
Anything that looks like melting or popping happened inside the package.
It's nowhere near 100% though.

The southbridge seemed to look normal. The northbridge is heatsinked, so I will remove it and have a look at it at a later time.
Too bad for me the mainboard was basically the only other interesting part in this untested system, good thing it was only €10 to begin with 🤣

Tetrium wrote:

Too bad for me the mainboard was basically the only other interesting part in this untested system, good thing it was only €10 to begin with 🤣

Try out the motherboard with a known working powersupply. Chances are if it boots to the bios the motherboard is fine, then just to make sure I like to Prime95 questionable boards. If they dont fail after 24hours then I call it good 😀

ODwilly wrote:

Tetrium wrote:

Too bad for me the mainboard was basically the only other interesting part in this untested system, good thing it was only €10 to begin with 🤣

Try out the motherboard with a known working powersupply. Chances are if it boots to the bios the motherboard is fine, then just to make sure I like to Prime95 questionable boards. If they dont fail after 24hours then I call it good 😀

The board is still in the case it came in when I bought it, that system is already partially disassembled and I don't actually have a testbench right now (working on it, along with some other projects of mine) so the testing will have to be postponed, but when I do I will do as you suggested, except that I don't use prime95, I prefer to actually use it (gaming) and see how things go 😁

Northwood P4 with RDRAM sounds like a lovely combo

ODwilly wrote:

Northwood P4 with RDRAM sounds like a lovely combo

Very much true! 😁

I kinda did make a slight slip-up yesterday as the chip that came with this board was actually a s478 1.7GHz Willamette and not a Northwood, but the board should definitely support Northwood, though only the 400MHz FSB ones.
If this board actually still works, I can pick either a 2.2GHz P4 or one of the P4 s478 Celeron ES chips I never got around to find a use for 😀

Enermax NAXN ENP500AGT recapping

This is a budget PSU just above the 'passable' mark that uses the ubiquitous CWT GPA platform (APFC, two-transistor forward and group regulation).

Most unfortunately it comes OEM stuffed full of Su'scon and Jun Fu crapacitors, unreliable bottom of the barrel junk 😵 . To add insult to the injury they are way undersized for the application (i.e. just 1000uF caps on +12V for a 500W PSU is anemic).

___

So well-sized quality jap caps to the rescue. 😎

+3.3V output filtering:
Jun Fu WG 6.3V 2200uF 10mm Ripple=1530 ESR=41 ===========> Panasonic FC 10V 2200uF 10x30mm R=1815 ESR=35
Su'scon MF 16V 1000uF 8mm Ripple=1210 ESR=60 ============> Panasonic FC 10V 2200uF 10x30mm R=1815 ESR=35

+5V output filtering:
Su'scon MF 16V 1000uF 8mm Ripple=1210 ESR=60 ============> Panasonic FC 10V 2200uF 10x30mm R=1815 ESR=35
Jun Fu WG 16V 1000uF 10mm Ripple=1350 ESR=46 ============> Panasonic FC 10V 2200uF 10x30mm R=1815 ESR=35

+12V output filtering:
Jun Fu WG 16V 1000uF 10mm Ripple=1350 ESR=46 ============> Rubycon ZL 16V 2200uF 12.5x25mm R=2770 ESR=18
Jun Fu WG 16V 1000uF 10mm Ripple=1350 ESR=46 ============> Rubycon ZL 16V 2200uF 12.5x25mm R=2770 ESR=18

-12V output filtering:
Jun Fu WG 25V 470uF 8mm Ripple=980 ESR=64 ===============> Rubycon ZL 16V 470uF 8x16mm R=995 ESR=56

+5VSB output filtering:
Chemicon KY 10V 1000uF 8x20mm Ripple=1400 ESR=46 ========> The one and only OEM decent cap onboard, stays and soldiers on
Su'scon MF 16V 1000uF 8mm Ripple=1210 ESR=60 ============> Sanyo WG 16V 1500uF 10x20mm R=2550 ESR=13

Small caps:
Jun Fu WG 50V 1uF 5x11mm ============> Panasonic FC 50V 1uF 5x11mm
Jun Fu WG 50V 4.7uF 5x11mm ==========> Panasonic FC 50V 4.7uF 5x11mm
Jun Fu WG 50V 10uF 5x11mm ===========> Panasonic FC 50V 10uF 5x11mm
Su'scon SD 25V 220uF 8x12mm =========> Panasonic FC 50V 220uF 10x20mm

As a curiosity the PCB of this PSU comes with spark-gaps for extra protection against line AC voltage surges (like not-too-close lightning strikes):

This cool EEVblog video shows how spark-gaps work: https://www.youtube.com/watch?v=vfP_65gSSBU&t=12m24s

C_Hegge reviewed the internally identical (also CWT GPA inside) Antec VPA-500P in this HardwareInsights article.

Spark gaps and a zener! Someone wears belt and suspenders at Channel Well.

Well done. Excellent work

I need a PSU for some testing and thought I might aswell take a look at some dumpster finds I have not yet tested.

The Seventeam ST-250PIIIP "PS3" small form factor 250W PSU from year 2000 or thereabout.

Its very cramped inside so its hard to get a good look.

I have not tested it yet but the build quality seems excellent, lets hope the rest is as good.

EMI suppression can plus a MOV and some spare caps. Excellent. I don't see a coil for PFC. Is it hidden or not featured?

gdjacobs wrote:

... Someone wears belt and suspenders at Channel Well.

^ 🤣 🤣 🤣

Skyscraper wrote:

... Seventeam ST-250PIIIP "PS3" small form factor 250W PSU from year 2000 or thereabout ...

Looks good ... but I spot Crapxons and PCE-Turs on the secondary, I wouldn't bet my money on those craps after 16 years.

You know the drill, after a recap with decent fresh caps it would be a nice retrocomp PSU.

gdjacobs wrote:

EMI suppression can plus a MOV and some spare caps. Excellent. I don't see a coil for PFC. Is it hidden or not featured?

It was too cramped for me to see much... lets pretend the coil is hidden as Im to lazy to reopen the unit.

TELVM wrote:

Looks good ... but I spot Crapxons and PCE-Turs on the secondary, I wouldn't bet my money on those craps after 16 years.

You know the drill, after a recap with decent fresh caps it would be a nice retrocomp PSU.

At some point this PSU will get recapped as it's compact but seems easy to fully disassemble, until then lets hope the shady caps still are within their rated specifcation or at least reasonable close to beeing so. 😉