Antec True Power Exploded

Chris

Weaksauce
Joined
Dec 16, 2002
Messages
82
I have a strange incidence. My Antec 480w TP exploded when not in use.
Last night I still used my computer without a problem. And today at 6 PM when I tried to use my computer, it refused to power on. Then, I smell something burning. Apparently last night when I slept or when I left home for work, my Antec exploded by itself and killed my motherboard and 1 harddisk. The burning mark is extreme. It even left black gunk around the area of the case near the PSU. The burnt PSU leaked some brown oil, which I assume the liquid inside the transistors. Some internal component of the PSU fell off, and it seems the burning was so severe, because some transistors seems to burn out in extreme heat and became black powder. I have no idea what was happening. Frankly I become very worried leaving my computer at home now without plugging off the power cable from the back of PSU. It looks like there was big fire when the PSU exploded.
Thank goodness, it was only my PSU....what if...............{speechless}
The PSU was connected to a UPS, which was OFF because of battery exhaustion when I got home. So, I suspect the short circuit should happen inside the PSU itself. The question.....WHY and WHEN exactly this happened ??? It was only 1.5 years old.

I will try to borrow some digital camera and post the pics for you to see. It looks extremely bad.
 
I forgot to mention that in the morning there was a power outtage right before I went to work. I only heard my UPS was giving alarm sound, but I didn't care and left home.
I wonder what exactly killed my PSU. Spike during power outtage ?
But my UPS has line conditioner, and it should protect the PSU, right ?
The fact is that my UPS is still running fine, so it should not be a spike that killed my PSU. If the spike was so bad, the UPS should be killed first before reaching my PSU.

When I got home, the power was back. Some lamps in my front door was on. Nothing seems to trigger my fuse box. Also, my UPS was turned off because of battery exhaustion, which means no power could possibly reach my PSU, even when there was a big spike. So, how could it be possibly burnt out ?
 
Antec's CWT made units are notorious for substandard capacitors. While they don't often fail this catastrophically, 1.5 years at Antec's slow fan speeds is just about right for these caps to start bulging and leaking. I just got done recapping three CWT units not long ago - they all had more than one cap needing replacement. The more loaded down they were, the more help they needed. The 420W model that overheated I had been using in my main rig needed 4 of them for sure, so I did the whole supply.

There may have been some external component to the failure, but my money's on the caps dying young.
 
I'm trying to wrap my head around an OFF PSU killing a board.

I mean, if it was receiving power, even an "OFF" system is getting power from the +5VSB. But if the system is off, the only juice in the PSU should be in the caps that would almost immediately discharge if still hooked up to a motherboard.

Very strange.
 
True... the standby caps have also been known to go, but that shouldn't have given much of a lightshow. All I can figure is it turned itself on somehow, maybe via the state after power failure bios setting. Can't think of anything that could have produced such a dramatic failure from power off. Unless it got blasted by lightning. Need pics.
 
*OFFTOPIC* Well i have an Antec SmartPower 500W should i change it?

*ONTOPIC* If you can give us some pics in the case and out and PSU open.
 
LOCO LAPTOP said:
*OFFTOPIC* Well i have an Antec SmartPower 500W should i change it?

Why automatically do people do this? They see someone with a problem and automatically want to sell their kit cos it might have the remotest part of failing. Theres a guy in the video card section doing the same thing when he has no problems with his cards!

How many hundreds of thousands of units are there out there that work perfectly. Of course there are going to failures but the percentages are low in compared to the number of happily working units.

To the OP, I think that Antecs have a 3yr warranty so I would contact Antec with the pics and see what they say :) think they should replace it under warranty :)
 
Oklahoma Wolf said:
Antec's CWT made units are notorious for substandard capacitors. While they don't often fail this catastrophically, 1.5 years at Antec's slow fan speeds is just about right for these caps to start bulging and leaking. I just got done recapping three CWT units not long ago - they all had more than one cap needing replacement. The more loaded down they were, the more help they needed. The 420W model that overheated I had been using in my main rig needed 4 of them for sure, so I did the whole supply.

There may have been some external component to the failure, but my money's on the caps dying young.

Glad to know it's not just me seeing this.
Though if you think the cheap caps are a demonstration of how cheap they are.. the frigging bastards at Antec still owe me a goddamned True380 RMA from a Sonata - which fell apart itself - after I went through the absurd song and dance of faxing all that stupid paperwork. Sure as shit, bad caps on top of the dead fan. They still haven't replied, or said a thing, even after I asked for a replacement filter assembly for the Sonata since the one in it fell apart as well.

Hooray Antec Kwality. I honestly am tempted to just haul their asses to my lawyer's office and let them tell him why they can't be assed to honor warranty.
 
There is an Antecrep here on the forums under that handle. If you guys have a problem with an RMA....contact him politely and he may be able to help you.
 
I will go to my friend's house tomorrow to get his digital camera. I really have no idea why this happened. This is an extremely embarashing incidence to me, because just 3-4 weeks ago, I have a friend who had exactly the same experience of self-blowing PSU (a very cheap one). He killed practically EVERYTHING , his mobo, his memory, AGP card, 3 PCI cards, 2 harddisk. He consulted me, and I was the one who explained the difference between a good PSU and a generic one. I highly recommended Antec. In fact, I took many comments from this forum about the danger of a generic PSU, how a bad PSU could SELF-BLOW and kill everything, how good Antec is, and bla ...bla...bla...
I sent him many related threads from this forum thru email. He bought Antec 480w True Power like mine, despite his tight budget for buying the whole new system.

And now......the "guru" who told him to buy Antec apparently "proved" to be wrong.....
It's like slapping my own face in front of him.
Geez, frankly I think I will hide this incidence from him. I;m just too embarasshed......just 3-4 weeks after all my "preaching" to him. I think I'd rather shoot my head with a gun, than telling him the truth.
 
I wouldn't say they're that bad... even with bad caps, these often last years. Failures like yours aren't all that common with them thankfully. I've also been hearing about CWT going to Teapo for some of their newer units, which is much better than their usual Fuhjyyu selection.
 
Chris said:
And now......the "guru" who told him to buy Antec apparently "proved" to be wrong.....
It's like slapping my own face in front of him.
Geez, frankly I think I will hide this incidence from him. I;m just too embarasshed......just 3-4 weeks after all my "preaching" to him. I think I'd rather shoot my head with a gun, than telling him the truth.
If you can't admit you're wrong, you have no business recommending anything.

And how old are you?
 
Oklahoma Wolf said:
I've also been hearing about CWT going to Teapo for some of their newer units, which is much better than their usual Fuhjyyu selection.
Didn't XClio stop using Fuhjyyus?
 
Firelord said:
Why automatically do people do this? They see someone with a problem and automatically want to sell their kit cos it might have the remotest part of failing. Theres a guy in the video card section doing the same thing when he has no problems with his cards!

And there's a guy at Anandtech that swapped out his PSU and the side panel on his case that was having no problems either. What's funny is after he did so, his GPU temps went up 4C! :rolleyes:

Yep. I see it all of the time. People don't realize that even as crappy as Antec may or may not be, a failure within a year of use of any kind only accounts for less than 5% of total units sold, and failures that cause damage to other components only comprise of another 5% of those failures... AT THE MOST. There's no reason to swap the unit out because of what happened to another guy. Swap out parts because you're having a problem or you're doing an actual upgrade (CPU, GPU, memory, etc.)

There might be as many as five posts a day on this website from people that had their PSU blow up and take other parts with it. But that doesn't account for the other THOUSANDS of people that are having no issues at all with the same brand PSU.

If a part in my computer works the way it should, should I post how much I love how it does what it's supposed to do? Maybe that will help equalize things. ;)
 
Holy crap!!!

That caught FIRE!!

I'm fixin' (Florida Cracker here, people) to put a TP 2.0 550W on the load tester, but now you've got me scared!!!! :eek:
 
Run the PSU away from the new SunMoon and have some marshmallows ready just in case ;)

That's a bad failure - by far the worst I've seen for an Antec/CWT. Hard to say what caused it.
 
Oklahoma Wolf said:
Run the PSU away from the new SunMoon and have some marshmallows ready just in case ;)

That's a bad failure - by far the worst I've seen for an Antec/CWT. Hard to say what caused it.

Not really hard to say what caused it, actually. The second picture tells the whole story; point of origin? Those cheapass capacitors. Exact cause is of course, much harder to pinpoint. But if I had to wager a guess I'd say that the capacitor towards the front was leaking at the base, and created a short. Short heated up, and voila, fire.
How can you tell point of origin was the capacitor towards the front? Easy - it's melted and collapsed. But the second capacitor right next to it, isn't. The PCB is also burnt away significantly around the first capacitor - enough to probably expose the mount - but the rear corner is clean, and the input toroid is intact.
 
I don't know if the cap was leaking, but I would agree it may not have been too healthy possibly leading to the regulator/switching transistor behind it going up in flames.
 
Oklahoma Wolf said:
I don't know if the cap was leaking, but I would agree it may not have been too healthy possibly leading to the regulator/switching transistor behind it going up in flames.

Yeah; can't say for sure if it was leaking. Can say for sure that it was the point of origin approximately, and likely the underside. May have been a short, may have been any number of things; defective switch stuck open, etcetera. Either way, the fire definitely originated either on the underside of that capacitor, or just forwards of it. And just based on the extent of burning, there had to have been additional accelerant on the underside of just that capacitor. If it was the base flammability of the capacitor itself, then the second capacitor would show more damage than it does. As it is, it only appears to have lost it's PVC covering with almost no deformation.
Of course, I'm no fire marshal. But I do deliberately push things to the point of catching fire as part of stress testing, so I've seen similar before. (What? If instead of popping off, it catches fire, I wanna know!) :D

Unfortunately, I misinterpreted the photo - forgot that I'm looking at a plastic sheet on the top, which helps identify where the fire exactly was. Those sheets have a very low melting point - an ordinary pocket lighter'll burn one. If you take a look, you see that the surface of the PCB - not the PCB itself - is severely scorched from the front capacitor to about the input toroid - and then abruptly stops. There's much more significant burning just forwards of the front capacitor, with a very clean area on the edge. That would pretty clearly indicate to me the presence of a flammable substance on the surface of the PCB in that specific area. However, the plastic itself shows only minor burning forwards of the capacitors. The fire was pretty localized. You can see how it's actually melted to the first capacitor, and has a gouge melted out of it between the capacitors, with the PCB scarred between them - lack of accelerant between capacitors. Now given the tendency of PSUs to have a 1-2 degree angle leaning downwards towards the front, I'd have to guess the capacitor was leaking. The fire spread backwards, and probably the epoxy or sealant or cover on the other components there was flammable, causing the rearmost gouge on the plastic. I'd wager there's more damage to the PCB towards the heatsinks forwards of the first capacitor, but only singing going towards the rear.

Doesn't mean a leaking capacitor caused it though, the more I look. It may have been a heatsink getting too hot, and touching the capacitor's PVC, causing the fire. However, going back to that front capacitor, it's deformed significantly while the second shows almost no deformation. So I'd wager there was definitely something wrong with it. It'd also take a lot more heat than normal, and the heatsink likely would have deformed - which it didn't.

Can't wait to hear what they say about RMAing this one.
 
A brownout is often more destructive than a spike.

I have seen many UPSes cause some serious damages after the battery is used up.
If the UPS does not have enough power, the power supply will suffer a brownout.

You are far better off with a simple power surge than with a UPS if you cannot automate the PC shutdown. A good UPS should have a monitoring software that will allow you to shut down connected machines in timely fashion.
I have the belkin 1500va and it allows you to configure actions based on events or states.

I find it totally senseless to have a UPS and not being able to communicate with it.
 
For the record: those are DEFINITELY Panasonic caps. Not the Fujjhyu's that we're used to seeing. My TP 2.0 has Panasonics too. I guess Antec is learning from their mistakes. ;)
 
jonnyGURU said:
For the record: those are DEFINITELY Panasonic caps. Not the Fujjhyu's that we're used to seeing. My TP 2.0 has Panasonics too. I guess Antec is learning from their mistakes. ;)

You mean mine are Panasonic caps and they are bad ?
Then how come TP 2.0 still uses Panasonic caps afterall ?
 
jonnyGURU said:
I guess Antec is learning from their mistakes. ;)

I'm not so sure about that after seeing the Fuhjyyus on the secondary side of your review sample ;)
 
Oklahoma Wolf said:
I'm not so sure about that after seeing the Fuhjyyus on the secondary side of your review sample ;)

Yeah. I guess they thought we wouldn't notice if they threw Panasonics on the primary. I had to pull the cables all out in order to even SEE the secondary caps. So not only are they Fuhjyyu's and the PSU runs really hot, but the caps are insulated by a bunch of wires draped over them so they're going to run super super hot. :rolleyes:

EDIT: Said "PSU" instead of "caps" in last sentence. Fixed.
 
well, at least it wasn't a cold cathode inverter going bad.. :D

http://www.casemodgod.com/ccfl_inverter_rant.htm
invertersuck04.jpg
 
jonnyGURU said:
Yeah. I guess they thought we wouldn't notice if they threw Panasonics on the primary. I had to pull the cables all out in order to even SEE the secondary caps. So not only are they Fuhjyyu's and the PSU runs really hot, but the PSU are insulated by a bunch of wires draped over them so they're going to run super super hot. :rolleyes:

Jonny, remind me to get you pictures of my NOT disassembled True380 out of the Sonata. It's disgusting. Literally, there's glue everywhere inside this thing. (Yeah, it's bad enough to be that visible.) It looks like a couple glue bombs went off inside of it.
Capacitors are unfamiliar to me, which is not a common thing, too. M with a squared-off circle around it, UQ, best I can do on partial part number is 32190. There's an indigo/white over near output stage, but the bulk of them look like they're the black/whites with the M logo. But, can't open it up - FINALLY going back for RMA.

(EDIT) Just found that thread over at BadCaps you posted, Jonny.
Wow; the TP550 is basically identical to the True380 single fan version. That's.. I'm just really not sure what to say. TP550's got bigger caps, but yep, that's them. No doubt. And the Fuckyou's all over the rest of it. Can't tell if the Fuckyous are leaking, but it sure wouldn't surprise me.
 
AreEss said:
Jonny, remind me to get you pictures of my NOT disassembled True380 out of the Sonata. It's disgusting. Literally, there's glue everywhere inside this thing. (Yeah, it's bad enough to be that visible.) It looks like a couple glue bombs went off inside of it.

I believe that... and the glue they use turns such an interesting shade of brown after it's been heated a year or two...

Only caps I can think of matching that description would be Matsushita/Panasonic. Not sure about the indigo one, but I've never seen Antec/CWT use anything but mostly Fuhjyyu and the odd Teapo or Capxon in the output side, so my money would be on one of them. Probably Fuhjyyu.
 
Oklahoma Wolf said:
I believe that... and the glue they use turns such an interesting shade of brown after it's been heated a year or two...

Only the stuff that's up against the 5W resistor in there. I thought the cap had blown it's entire side out when I first saw it, but nope. That stuff has to be hitting at least 90C since the PSU was peaking around 70C.

Only caps I can think of matching that description would be Matsushita/Panasonic. Not sure about the indigo one, but I've never seen Antec/CWT use anything but mostly Fuhjyyu and the odd Teapo or Capxon in the output side, so my money would be on one of them. Probably Fuhjyyu.

Yeah, see my edit. The True380 is disturbingly similar to the TP550. Right down to the garbage Fuhjyyu's. Antec's finally taking care of all the problems with this Sonata, but talk about purchases one regrets. Wish I could've gotten it without the PSU; no way the True380's going back in when I have a Silencer360 and 10C lower temps.
 
The M with the square around it IS Matsushita. That's what the primary caps in my Antec are as well.

But it seems like it's the secondary caps that almost always fail, and those are all Fuhjjyu.

Despite being told primarys do not fail, I have had two Antecs where the primarys blew up, and they were fairly new units (less than a couple months) Went off with a deafening boom. Those primary caps were Fuhjyyu, but the caps probably didn't fail because they suck. I think there was a manufacturing defect with the PSU that caused a short that IGNITED the caps. :D
 
The primaries can be awful too, though they rarely fail as obviously as the secondary caps... I once pulled two 470uf at 200v Viva caps out of a Deer 250W and tried to use them to repair an old 1975 vintage Hitachi TV. The replacements were noticably worse than the 30 year old caps they replaced... it just blew my mind how bad the Vivas were - there was very obvious line interference present when I ran the Vivas in there.

Anyway, I love the Nichicon primaries in this here ST56ZF :D
 
Oklahoma Wolf said:
The primaries can be awful too, though they rarely fail as obviously as the secondary caps... I once pulled two 470uf at 200v Viva caps out of a Deer 250W and tried to use them to repair an old 1975 vintage Hitachi TV. The replacements were noticably worse than the 30 year old caps they replaced... it just blew my mind how bad the Vivas were - there was very obvious line interference present when I ran the Vivas in there.

Said it in another thread, I'll say it again; anyone who says something doesn't and can't fail ever, needs to be shot in the face. Parts fail. They may fail RARELY, they may not NORMALLY fail, but they DO fail period. Primaries definitely can and will fail, as we can all prove repeatedly, I'm sure.

Anyway, I love the Nichicon primaries in this here ST56ZF :D

I have too much trouble getting Nichicons, so I've been favoring CDE and Panasonic T-tops. I need to see if I can get my one source to use CDE primaries if I drop-ship them.
 
I just use whatever Digikey has that will work... used Panasonic and UCC to recap the three CWT based units I had sitting around here. Mostly Panasonic though - UCC only where I couldn't get the Pannies to fit.

The Silverstone came with the Nichicons from the Etasis factory... the other caps in there I could identify for sure are Teapo and Rubycon. Not worried as much about Teapo as most folks over at Badcaps are - my two year old FSP 530W uses nothing but Teapo and still works like new.
 
Oklahoma Wolf said:
I just use whatever Digikey has that will work... used Panasonic and UCC to recap the three CWT based units I had sitting around here. Mostly Panasonic though - UCC only where I couldn't get the Pannies to fit.

Yeah; I source from a few places, but I get CDE's easier. Especially the 10,000uF beasties - mmm, only $20 each! Panasonic occupies the bulk of my BP6's, but I'm STILL trying to find someone decent who makes 1500uF 6.3V 8mm. (There is NO way to get 10mms in there, I don't care what anyone says. NOT in these systems. And I'm not doing the 1200uF mod either.)

The Silverstone came with the Nichicons from the Etasis factory... the other caps in there I could identify for sure are Teapo and Rubycon. Not worried as much about Teapo as most folks over at Badcaps are - my two year old FSP 530W uses nothing but Teapo and still works like new.

Yah; Etasis uses Nichicons, Teapo, and Rubycon typically. For the hotswap stuff, it used to usually be high temp (but high tolerance) Nichicon exclusively. Might have changed; the tolerances on the hotswaps are within limits, but certainly not great.
Agreed on Teapo partly; they're a mixed bag. Older Teapos which are physically identical to known bads tend to be bad. However, if it isn't identical to a known bad, they seem to be just fine. Anything newer than I'd say 2002 or 2003 seems to be fine regardless of size, too. Rubycon on the other hand.. no. Never again. Not a chance in hell.
 
AreEss said:
Older Teapos which are physically identical to known bads tend to be bad. However, if it isn't identical to a known bad, they seem to be just fine. Anything newer than I'd say 2002 or 2003 seems to be fine regardless of size, too.

Cool - the FSP has some huge Teapos on the secondary... five 6800uf. I hope I never have to replace them... not sure I could find replacements that easily. They're the usual green and gold stuff. The Silverstone's Teapos are black.

Edit - UCC does have some KY's that will replace some of the FSP's Teapos (12.5mm diameter), but I can't find any thin replacements for the 16v caps.
 
Back
Top