Author Topic: Tagan TG480-U01 PC ATX power supply  (Read 5088 times)

0 Members and 1 Guest are viewing this topic.

Online HwAoRrDkTopic starter

  • Super Contributor
  • ***
  • Posts: 1456
  • Country: gb
Tagan TG480-U01 PC ATX power supply
« on: May 23, 2016, 01:38:35 pm »
This isn't so much a how-to or a request for help in repair, but instead I thought it might provide some entertainment to share the horror that is the internals of this PC power supply. I do have a couple of questions, but they're more related to trying to understand "WTF is this?". :D

I originally bought this PSU around 2004 and used it in my own PC for a few years, but it then got relegated to a spare part when I upgraded, and has sat around ever since. Recently I dug it up again for use as a general-purpose +12V and +5V supply for a project. But, one day last week I turned it on to use, and after a few minutes it emitted a hissing noise and a sweet-smelling odour wafted into my nostrils. It appeared to carry on working, but I thought better of it and determined to open it up and see what happened. I suspected a leaky cap - and I was proved correct.

First, here's the culprits:




I am going to attempt to replace all these blue caps. As you can see, if they're not already leaking, they're bulging at the top.

Question: does anyone know the best way to remove that white gunk? It's not the usual kind of rubbery stuff that peels off fairly easily. It's quite hard - I can barely make a dent in it with a screwdriver tip. Maybe it was flexible when new, but has been baked hard over the years. Either way, I need to remove some of it, but not sure how.

This will be the third time now it's been in need of repair. I think it's going to become the proverbial ship of Theseus, with so many parts being replaced. :)

First repair was to replace one of the fans. The original failed after only one year. The plastic of the fan blades developed a crack, which meant it was no longer a tight fit over the rotating central part (I'm not sure what it's called - the bit with the windings, etc.). The centre would rotate, but the fan blades would not. :D I replaced the fan myself, as the postage cost to send it back under warranty would have been half the cost of the PSU itself!

Second repair was after about 2 years. It would turn on, run for about 10 seconds, then turn off again. I had no idea what was wrong, so I gave it to my dad (a retired PSU engineer) to have a look. He said he traced it to, and replaced, a faulty IC on a daughterboard - for which he fortunately happened to have an exact replacement to hand from his old stash of parts.

I hadn't seen what he'd done until I cracked the case open today. I think it's fairly obvious which part here was replaced:



;D Some kind of quad voltage comparator. That part is almost as old as I am. They obviously don't make 'em like they used to!

The more I look at this PSU, the more I think I was ripped-off when I bought it. It was originally marketed as a high quality power supply, and received quite good reviews, where it apparently performed quite well. But it appears even to my inexperienced eye that this is a bit of a dog's breakfast of a power supply. It seems to me as though someone has taken an existing design and modified, added and bodged in such a way as to turn it into a 'good' PSU. But then the silkscreen on the main board says it was designed in 2003! ???





The build quality is horrible. There are bodge wires and components all over the place. The soldering on the bottom of the main board is fairly horrific. Everything is extremely densely packed on the main board.

Some parts of it I'm not sure of the function of. Take, for instance, this additional board:



It has a four-pin connector coming off the main board, labelled 'AC', 'AC', '+V', 'GND'. I'm not sure what the its function is, apart from it taking in AC and feeding a DC voltage back out. Something to do with power factor correction judging by the presence of an 8-pin DIP marked L6561. My noob knowledge only gets me as far as tracing the AC input through a large bridge rectifier (with large cap across +/- out), then through an enormous inductor, and then branches off to three large transistors on the heatsink. Only one of them (the smallest of the three) has a pin going to the '+V' output. Apart from pin 1 of the L6561 which joins it (via a 1M resistor), nothing else on this board goes to that '+V' output. :-//

Then there is this board directly on the mains IEC connector:



Some kind of filtering, perhaps? It features a couple of inductors, a large cap and two smaller disc-shaped ones. The large cap bridges L and N. Both L and N pass through an inductor each before going out to the switch on the back. The small caps bridge between L/N and earth.

I do kind of like the way they implemented the variable-speed fan control, though. They've done it as a plug-in optional thing. Instead of directly connecting the fans to the two +12V connectors provided on the main board, they've instead ran one to a little PCB attached to one of the heatsinks. This board has (what I presume is) a thermistor attached to a tiny bit of aluminium, which nuzzles against the main heatsink to sense its temperature, and does something with a small TO-92 transistor to regulate fan speed accordingly. There are then three (one spare!) output connectors for the fans.
« Last Edit: May 23, 2016, 01:41:44 pm by HwAoRrDk »
 

Online HwAoRrDkTopic starter

  • Super Contributor
  • ***
  • Posts: 1456
  • Country: gb
Re: Tagan TG480-U01 PC ATX power supply
« Reply #1 on: May 23, 2016, 04:14:45 pm »
Hehe, discovered another amusingly bad thing as I was inspecting the leaky caps to read their values.

There's an inductor that's missing it's core! :-DD



Also, I happened to hit one of the two huge caps on the main board with my screwdriver. Surprised to hear it ring like a bell. They sound rather... hollow. Should a legitimate 680uF 200V large capacitor sound like that? Shouldn't they be a bit more 'solid'? Are these dodgy components too? :D
 

Offline ovnr

  • Frequent Contributor
  • **
  • Posts: 658
  • Country: no
  • Lurker
Re: Tagan TG480-U01 PC ATX power supply
« Reply #2 on: May 23, 2016, 04:25:09 pm »
Haha. That's terrible.

Mystery board is indeed an active PFC boost converter. It was p. much bodged on, as you say.


As for repairing it... well, it's terrible, so I wouldn't have bothered. But hey, why not.
 

Online HwAoRrDkTopic starter

  • Super Contributor
  • ***
  • Posts: 1456
  • Country: gb
Re: Tagan TG480-U01 PC ATX power supply
« Reply #3 on: May 23, 2016, 05:51:57 pm »
I think I may have hit an obstacle with my plan to replace the bad caps. :(

All of them are 10mm in diameter. But, looking at the usual vendors (Farnell, Mouser, Digi-Key, TME), they don't have any 4700uF caps that are 10mm in diameter. Smallest available is 12mm, which will never fit due to the tightly-packed nature of the board.

Does anyone know where I can find - in, or supplied to, the UK - any decent quality 4700uF, 10v, 105C alu. electrolytics that are 10mm diameter, 5mm lead spacing, 30mm max. height?
 

Online HwAoRrDkTopic starter

  • Super Contributor
  • ***
  • Posts: 1456
  • Country: gb
Re: Tagan TG480-U01 PC ATX power supply
« Reply #4 on: May 23, 2016, 08:49:26 pm »
After much searching, I have found that all of the quality manufacturers don't even make a 10mm diameter 4700uF alu. electrolytic cap. Panasonic, Rubycon, Nichicon, Wurth, Vishay - all nothing. I even looked at Samxon...

Seems like the only damn people who make a 10mm dia. 4700uF is the original manufacturer, Fuhjyyu. (Which I have since found out have a terrible reputation for quality... with the nickname ****yu. ;D)

Does anyone know of any other brands of capacitor that may do the right size?

Gah, if only the board weren't so tightly packed that a 12mm cap won't fit... >:(
 

Offline janoc

  • Super Contributor
  • ***
  • Posts: 3781
  • Country: de
Re: Tagan TG480-U01 PC ATX power supply
« Reply #5 on: May 23, 2016, 09:14:21 pm »
I think you could get away with smaller caps - e.g. 2200uF or so. The exact capacity is not all that important, it is just a filter. Moreover, those large caps have huge tolerances anyway. Just take the largest capacity one you can find for the given voltage and diameter.

The PFC being bodged on is a common thing on cheap power supplies that are/were imported into Europe. EU requires PFC since some years ago, so they bodge this on using an add-on board and sell the supply without it on markets where PFC isn't required. Saves quite a bit of money - copper for those coils is expensive.

That core-less inductor looks pretty much like an air-core inductor to me and not missing its core. There is no glue nor anything else around that would hold the missing core in place there.

However, as others posted before - I wouldn't bother repairing this. If the PSU has failed 3 times already, it is very likely you are just pouring money down the drain. A new ATX PSU will not cost you more than about 80 EUR, for decent quality one. Cheap junkers can be had for half of that. Saving money on these is a bad idea - if it blows, it can take out whatever is connected to it too.

« Last Edit: May 23, 2016, 09:19:38 pm by janoc »
 

Offline TheMG

  • Frequent Contributor
  • **
  • Posts: 866
  • Country: ca
Re: Tagan TG480-U01 PC ATX power supply
« Reply #6 on: May 24, 2016, 02:21:59 am »
However, as others posted before - I wouldn't bother repairing this. If the PSU has failed 3 times already, it is very likely you are just pouring money down the drain. A new ATX PSU will not cost you more than about 80 EUR, for decent quality one. Cheap junkers can be had for half of that. Saving money on these is a bad idea - if it blows, it can take out whatever is connected to it too.

I agree with this. I've replaced capacitors on a few PC power supplies over the years, and in most cases they only lasted a few more months before BANG! something else goes (usually the switching transistors).

Usually the capacitors going all bulgy and leaking means one of two things:

1- defective capacitors

2- capacitors that were pushed beyond their limits

In the case of #2, if the manufacturer pushed the capacitors so hard in the design that they failed prematurely, chances are high that other components in the power supply are also being pushed very hard and it's probably not long before something else fails.

Also, I happened to hit one of the two huge caps on the main board with my screwdriver. Surprised to hear it ring like a bell. They sound rather... hollow. Should a legitimate 680uF 200V large capacitor sound like that? Shouldn't they be a bit more 'solid'? Are these dodgy components too? :D

That's fairly normal. All that means is there's a bit of a gap between the rolled-up capacitor and the can/housing. Those large high voltage capacitors rarely fail as they are not subjected to the high frequency high current ripple like the output filter capacitors.
 

Offline imidis

  • Frequent Contributor
  • **
  • Posts: 426
  • Country: ca
Re: Tagan TG480-U01 PC ATX power supply
« Reply #7 on: May 24, 2016, 08:23:07 am »
My vote would be to toss it, that thing looks aweful in more ways than one.  :-//
Gone for good
 

Online HwAoRrDkTopic starter

  • Super Contributor
  • ***
  • Posts: 1456
  • Country: gb
Re: Tagan TG480-U01 PC ATX power supply
« Reply #8 on: May 24, 2016, 12:06:33 pm »
I think you could get away with smaller caps - e.g. 2200uF or so. The exact capacity is not all that important, it is just a filter. Moreover, those large caps have huge tolerances anyway. Just take the largest capacity one you can find for the given voltage and diameter.

The PFC being bodged on is a common thing on cheap power supplies that are/were imported into Europe. EU requires PFC since some years ago, so they bodge this on using an add-on board and sell the supply without it on markets where PFC isn't required. Saves quite a bit of money - copper for those coils is expensive.

Ah, okay. :-+ So a 3300uF would probably do instead? That's the greatest capacitance value I can find in 10mm diameter.

You might be right about the PFC. I was looking at some old reviews on a few websites, and saw that a photo of the internals of the version reviewed by a particular US website didn't look the same. What was present instead of the PFC board was a much, much smaller thing - presumably doing DC rectification and not much else.



I doubt I'd be wasting money repairing it for a third time. After all, the total cost of all previous repairs was zero. ;D But, on the other hand, I think you guys may be right - that who knows what other components may now be marginal as a result of these failed caps, so probably not a good long-term prospect if it was to be used again in another computer. Plus, looking at it again today, it's going to be a big pain in the arse to get to one of the caps that is hiding under the heatsink, probably necessitating removal of the heatsink - which would mean de-soldering 6 components.

That's fairly normal. All that means is there's a bit of a gap between the rolled-up capacitor and the can/housing.

Good, they're not fakes then. Given some of the crap inside this PSU, it wouldn't surprise me if the manufacturer tried to pull a fast one and put in some fakes - especially as, unlike the rest, they're actually a good brand, Matsushita/Panasonic. :D
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 7695
  • Country: de
  • A qualified hobbyist ;)
Re: Tagan TG480-U01 PC ATX power supply
« Reply #9 on: May 24, 2016, 01:02:29 pm »
I had two failed Tagan ATX PSUs. I bought them because they seemed to be decent based on several reviews. IIRC the price was about +50% of common ones. After the first one failed I had a look inside and was very disappointed. Built to survive the 3 year warranty by one day. >:( The third one I had went directly to e-junk recycling also.
 

Offline janoc

  • Super Contributor
  • ***
  • Posts: 3781
  • Country: de
Re: Tagan TG480-U01 PC ATX power supply
« Reply #10 on: May 24, 2016, 09:18:31 pm »
Ah, okay. :-+ So a 3300uF would probably do instead? That's the greatest capacitance value I can find in 10mm diameter.

Yeah, that will likely be fine. Elkos have commonly -20 to +80% tolerances anyway. Just get the special low ESR type or a model specifically designed for switch mode power supplies. Otherwise you will be changing the failed caps soon again.

I doubt I'd be wasting money repairing it for a third time. After all, the total cost of all previous repairs was zero. ;D But, on the other hand, I think you guys may be right - that who knows what other components may now be marginal as a result of these failed caps, so probably not a good long-term prospect if it was to be used again in another computer. Plus, looking at it again today, it's going to be a big pain in the arse to get to one of the caps that is hiding under the heatsink, probably necessitating removal of the heatsink - which would mean de-soldering 6 components.

Well, good quality low ESR caps are not cheap. I am not sure where are you sourcing them, but recapping a PSU like this can easily be 20-30 EUR in caps alone here if you are changing all of them.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf