Don't just pass on it. - RUN! - As fast as you can.

All the Smart Power I + II, True Power I + II + III, True Blue (and so on) models had the same problem. - Fuhjyyu caps
It was complicated by the fact that some caps were weird sizes that are impossible to find replacements for.
These photos are from some 550w I had.
I've rebuilt about 10+ of them and I've seen 100's of complaints about them online.
I've also seen two cases where Smart Powers caused the owner's house to catch fire.
.
Never trust PSU review sites. They only see brand new unused PSUs.
.

Any issues beyond the caps?

hmm, so if i open it up and replace the caps i should be fine then

gdjacobs wrote:

Any issues beyond the caps?

Some of them blow the PWM controller but I never had that happen to me.
The PSU 'click' over at Badcaps net worked out some kind of a mod for that.
For me, if a recap won't fix a PSU I would just get another one.
I've been lucky. It hasn't happened very often

One of the failure modes for the early Smart Powers was a bit unique.
They would develop a high pitch whine or similar noise.
If you heard it a cap was about to blow and there was a chance it would take the mobo with it.
It might whine for weeks before actually failing. Might not.
.

candle_86 wrote:

hmm, so if i open it up and replace the caps i should be fine then

I will say ~probably~ fine if you can find the caps.

(The following uF numbers vary some between models but the situation doesn't.)
(The length is by memory. I didn't write it down. They may have been 45mm long.)
There are usually six 10mm diameter caps that are packed in too tight to up-size the diameter.
= Four are 10x40mm 4700uF 10v. Their ESR is 0.038.
You will not find those. They do not exist short of ordering them custom made.
Best you can do is 3300uF 10mm with appropriate ESR. (Panasonic FK)
= There are two 3300uF 10x40mm 16v. Their ESR is 0.024.
The closest I know of that is still in production is 2200uF 16v 10mm. (Panasonic FK)

I used to get 3300uF 16v 10mm with good enough ESR from a fellow that custom ordered caps.
I would replace all six caps with those.
He no longer orders those and I don't know of an alternate source of new caps.
Sometimes I see NOS Teapo on Ebay ( 3300uF, 16v, 10x40mm, SC series) that will work but that depends on how much you trust Teapo.

I would look for a different PSU.
Last week when I looked there were several Enermax that will do the job.
Sometimes those blow caps too sometimes but you can find replacements. Usually 12.5mm fit fine.
Often I look for used Dell Precision PSUs from towers. The watts vary by system model.
.

PCBONEZ wrote:

Often I look for used Dell Precision PSUs from towers. The watts vary by system model.

Why do you look for Dell Precision PSU's? I have a Dell Precison Workstation 410 I bought in 1998. Now it runs dual PIII-850's. I've never opened up the PSU chassis or had any issues. Is this something I should be checking?

feipoa wrote:

PCBONEZ wrote:

Often I look for used Dell Precision PSUs from towers. The watts vary by system model.

Why do you look for Dell Precision PSU's? I have a Dell Precison Workstation 410 I bought in 1998. Now it runs dual PIII-850's. I've never opened up the PSU chassis or had any issues. Is this something I should be checking?

That last was recommendations for ~good PSUs~, not new projects.
Precisions are workstations so they tend to have healthy +5 and +3.3 volt rails. (There are exceptions.)
I like them because they are usually problem free.
.

PCBONEZ wrote:

All the Smart Power I + II, True Power I + II + III, True Blue (and so on) models had the same problem. - Fuhjyyu caps It was co […]
Show full quote

All the Smart Power I + II, True Power I + II + III, True Blue (and so on) models had the same problem. - Fuhjyyu caps
It was complicated by the fact that some caps were weird sizes that are impossible to find replacements for.
These photos are from some 550w I had.
I've rebuilt about 10+ of them and I've seen 100's of complaints about them online.
I've also seen two cases where Smart Powers caused the owner's house to catch fire.
.
Never trust PSU review sites. They only see brand new unused PSUs.
.

TP2-550EPS12V_.jpg

TP2-550_5vsb.jpg

At the time I bought it, the Antec TP3-650 was advertised as having Japanese capacitors (on the website, not the box) . I never actually looked inside to check (now I probably should).

darry wrote:

At the time I bought it, the Antec TP3-650 was advertised as having Japanese capacitors (on the website, not the box) . I never actually looked inside to check (now I probably should).

I suppose they could have told the truth for a change. I wouldn't put money on it though.
In another thread somewhere there are screen shots where Antec got caught doctoring the images on their own website.

Antec has a bad habit of changing the product without changing the model number.
If they were doing one of the changes right about when you bought then you could have ended up with either version.
.

well good info, i had a trueblue 480W back in 2005 that was rock solid, but good info didn't know about the future issues with them, I sold it in 2009 with an AM2 system running a 7950GT 🤣

Truepower "New" is a good PSU:

https://www.anandtech.com/show/3985/three-550 … erent-prices/11

yes but i need at least 32A on the 5V 🤣

candle_86 wrote:

well good info, i had a trueblue 480W back in 2005 that was rock solid, but good info didn't know about the future issues with them, I sold it in 2009 with an AM2 system running a 7950GT 🤣

If you got 4 years out of it you were lucky and did good.

I had two that I bought for emergency spares before I knew about their problems.
They were on the shelf still in the shrink-wrap for about a year before I needed one.
When I finally did need one and I opened it up I noticed the brown cap leakage through a fan grill.
So I opened the second one. It had the same problem.
I had read tons of online reviews before I chose those. I was really pissed.
.

candle_86 wrote:

yes but i need at least 32A on the 5V 🤣

Many of the Enermax have it.
EG365P-VE has 32a on +5v
EG465P-VE has 44a on +5v
They are the Emermax Whisper series. There are several others with similar model numbers that will also work.
Make sure you can read the label and verify the numbers.
At some point they upgraded their ATX spec without changing the model number so two with the same model number don't always have the same amps.

Stay away from Liberty and Noise Taker.
Those two have issues with the glue on the components turning conductive when it hardens.

candle_86 wrote:

yes but i need at least 32A on the 5V 🤣

There are models from Delta and Zippy that can deliver that. Also, there's a thread for this:
The PSU Matchup Thread

PCBONEZ wrote:

There are usually six 10mm diameter caps that are packed in too tight to up-size the diameter. = Four are 10x40mm 4700uF 10v. Th […]
Show full quote

There are usually six 10mm diameter caps that are packed in too tight to up-size the diameter.
= Four are 10x40mm 4700uF 10v. Their ESR is 0.038.
You will not find those. They do not exist short of ordering them custom made.
Best you can do is 3300uF 10mm with appropriate ESR. (Panasonic FK)

Some of the original caps for the 5v rail are 3300uF. Some of them have a mix of 3300uF and 4700uF. It probably just depends on whatever crap they had in their parts bin on the day the PSU was made. I recently recapped a TruePower 550 which had 1x 4700uF and 3x 3300uF. The 4700 was slightly taller than the 3300.

Since the original caps are such low quality (and probably with fake ratings), a brand name 3300uF should be even better than it was when new. I used 3300uF 6.3v Panasonic FR.

seanneko wrote:

PCBONEZ wrote:

There are usually six 10mm diameter caps that are packed in too tight to up-size the diameter. = Four are 10x40mm 4700uF 10v. Th […]
Show full quote

There are usually six 10mm diameter caps that are packed in too tight to up-size the diameter.
= Four are 10x40mm 4700uF 10v. Their ESR is 0.038.
You will not find those. They do not exist short of ordering them custom made.
Best you can do is 3300uF 10mm with appropriate ESR. (Panasonic FK)

Some of the original caps for the 5v rail are 3300uF. Some of them have a mix of 3300uF and 4700uF. It probably just depends on whatever crap they had in their parts bin on the day the PSU was made. I recently recapped a TruePower 550 which had 1x 4700uF and 3x 3300uF. The 4700 was slightly taller than the 3300.

Since the original caps are such low quality (and probably with fake ratings), a brand name 3300uF should be even better than it was when new. I used 3300uF 6.3v Panasonic FR.

Yes as I said it varies from model to model. The one that specific list is from was a TP-II 550w.
A TP-II 480w will be slightly different as would a TP-1 550w (if they made a 550w in TP-1)
What is consistent is was those annoying 10mm large uF Fuhjyyu that were packed in such that they didn't even get airflow.
(Caps last longer if they get some airflow/cooling.)

I think I was the one that originated the idea to use 3300uF to replace the 4700uF eons ago over at badcaps.net.
At the very least I was part of the conversation that worked it out and the one pushing for physically taking the ESR measurements of the Fuhjyyu as those can sizes are not even in Fuhjyyu's datasheets.
The idea spread from BCN and now just about everyone that rebuilds PSUs and runs into the jumbo uF 10mm problem does it.

I don't recall if they were part of that specific conversation but at that time jonnyGURU (creator) and OklahomaWolf (current owner) of jonnyguru.com were frequent visitors over at badcaps.net. I think they showed up because I was giving them shit on their site about how their TP-II review screwed me over. After that info about caps, ESR and ripple started showing up on their site and in their PSU reviews. Earlier they weren't doing all that. A lot of their information originated at BCN but it was a two-way street. Everybody was learning and experimenting with how far things can be pushed.

The problem is not exclusive to Antec.
IIRC those PSUs were built by Channel Well Technology (CWT) who was the builder for many other brands.
CWT builds to a cost point and/or the specifications the client wants.
The exact same design might show up badged as several brands but the individual components (and quality thereof) can vary.

It's preferred and good engineering practice to use the original uF if at all possible....
.... but (by looking at other PSU's) it's very clear 2x 4700uF were overkill insofar as the uF.
In other similar PSUs it's common to see a single 4700uF or a pair of 3300uF where that pair of 4700uF are.

ESR on the other hand is a critical spec for output filter caps.
- Never go higher. Within reason going lower is okay.

In OP filter caps...
The ESR goes to how much noise (ripple) is in the DC.
The uF primarily goes to how well the PSU handles DC transients.
- Too little Uf and the PSU will have a hard time with DC load changes. May not be able to keep DC voltage in spec during transients.
- Too much uF and the time and current to initially (on startup) charge the cap might cause problems.
----- If the inrush current (to the cap) is too high the over-current protection might think there is a short and turn the PSU back off.
----- If the DC voltage takes too long to get into spec (due to charging the huge cap) the time limit for PWR-Good to issue will pass and that will turn the system back off.
(This next is not as big a deal as the amount of words to explain it might suggest.)
The uF value in conjunction with the nearby inductor's value optimizes (tunes) filters to pass (to ground) the expected ripple frequency best. Usually the acceptable uF range is fairly wide. (+/-20% uF is a thumb-rule for replacements when you HAVE to change values..) Going up or down one standard uF value (approx 20%) doesn't throw the tune so far off as to cause a problem. (Should say I've never seen it cause a problem.) If you want to keep spot on the original tuned frequency then you need to change the inductor's value as well. Going form 4700uF to 3300uF exceeds that 20% but I've never seen it cause a problem.
.

Well I'm going to recap it