Does Antec replace PSU's that burned & melted?

[c627627]

Found that the cause of my trouble here
http://www.ocforums.com/vb/vb/showthread.php?s=&postid=2400014
was burned PSU connector that melted into the mobo socket.

Is this kind of thing covered under warranty?

[c627627]

http://c627627.com/IMG/Overclockers/PSUburn1.jpg

On the flip side:
http://c627627.com/IMG/Overclockers/PSUburn2.jpg

[Ritalin]

wow that sux :(

[_slh_]

Probably depends on what caused it to burn out in the first place

[L337 M33P]

All the wires that are crispy are 5 volt wires. I say phone them up and request an RMA, assuming you haven't overloaded the PSU and it's still under warranty.

Crummy connectors or a poor contact between motherboard and the ATX connector can lead to this, and as you have a motherboard that dervies Vcore from the 5v line there would have been a hefty current flowing. The current would have heated the connectors as they had just a slight bit more resistors, and the heat would have increased the electrical resistance of the contact a little more and b00m, it runs away and you smell burning :(

[Oklahoma Wolf]

Another possible cause of this type of damage is a failure in the motherboard's voltage regulators causing it to pull too much current, though if the 5v rail didn't read abnormally low in BIOS/MBM (usually I advise a voltmeter on the PSU, but in this case we'd want to know what the board itself is up to as well) it's possible the connector just wasn't making good contact on the one ultra fried pin.

I would RMA the board though too... once Molex connectors get this way, they should be replaced on both sides.

[c627627]

Yeah but how's the mobo manufacturer going to justify the RMA for damaged caused by Antec? Certain battery companies compensate you for damaged equipment if the batteries leak, my hope was Antec would do the same.

Sparkle is the same as Fortron, can I use this for a week:

http://c627627.com/IMG/Overclockers/235sparkle.jpg

for
2100+ JIUHB 0308 UPMW
3 x 256MB K-Byte PC2700 (SpecTek chip)
Epox 8RDA+ nForce2
Thermalright SK-7 with variable speed Thermaltake Smart Fan 2
BFG GeForce4 Ti4200 8X 128MB
Two 120GB WD 7200 RPM
tv card, modem, ethernet card, sound card,

I'll disconnect, the CDRW, DVD drive and floppy.

[Oklahoma Wolf]

Any other 235w unit I'd say forget it, but the Fortron may do the job for a little while non overclocked. I'll leave that one for those with more experience on lower watt Fortrons ;)

The Antec may not be at fault - if there was a problem with the board's regulators that caused excessive current draw, it would then be the board's problem. Though the Antec has overcurrent protection, it's possible in my mind the board drew enough to damage the ATX connector yet not enough to trigger the protection.

[L337 M33P]

I'd say turn it all down too. Run the chip at 1.33 GHz or something with a 133 FSB and put the voltages down low. Your power consumption should be less than 200W if you underclock and undervolt. Also to reduce power, especially on the 3.3v line which isn't exactly beefy try removing 2 of the 3 RAM sticks.

[c627627]

Is there a concensus now that if you had to choose between the PSU and the Epox 8RDA+ nForc2 mobo cuasing this problem, you would all go with the "board's regulators that caused excessive current draw..."

Originally posted by Oklahoma Wolf
The Antec may not be at fault - if there was a problem with the board's regulators that caused excessive current draw, it would then be the board's problem. Though the Antec has overcurrent protection, it's possible in my mind the board drew enough to damage the ATX connector yet not enough to trigger the protection.

[c627627]

Hooked up that 235 watt Sparkle.

Left only one RAM stick in and lowered voltage to 1.6 volts form 1.65 volts and set my 2100+ B to [133] x 12.5 = 2000+.

CPU temp 30 C

+5 at 4.95
+12 at 11.86
-12 at -12.27
-5 at -5.09

Wow those Sparkle's (a.k.a. Fortrons) are damn good.

[The Spyder]

Yep, I only use sparkel because of how well built they are.

[doc6886]

Originally posted by c627627
Hooked up that 235 watt Sparkle.

Left only one RAM stick in and lowered voltage to 1.6 volts form 1.65 volts and set my 2100+ B to [133] x 12.5 = 2000+.

CPU temp 30 C

+5 at 4.95
+12 at 11.86
-12 at -12.27
-5 at -5.09

Wow those Sparkle's (a.k.a. Fortrons) are damn good. Sparkles are the same things as Fortrons?????

[larrymoencurly]

Unless the PSU burned and melted because of a fire outside of it, I think that Antec should cover it because they advertise that it's protected against overloads. It's possible that they recommend a limit of 6-8A per wire, but since I couldn't find that information, it must mean that even if a wire melts from having > 30A flow through it, it should still be covered.

[c627627]

We'll find out tomorrow.
Regardless, the question still stands if there's anybody that disagrees that : "board's regulators caused excessive current draw..."

In other words, should I stop using this mobo pronto?


Originally posted by doc6886
Sparkles are the same things as Fortrons?
Yes other names made by the same companies (I believe) Antecs (Channel Well), Fortron (Sparkle), Sirtec (Enlight, Vantec, Thermaltake), Herolchi (Heroichi), Zalman (Aopen).

[Oklahoma Wolf]

I have to admit, your 5v line looked in good shape in the other thread - didn't mean to cause panic, just wanted to present another possible cause of the damage. Judging from your 5v reading, though not as accurate as a DMM measurement, your board's regulators are probably ok. It may be worth inspecting the mosfets while the board's running to see if any run significantly hotter than the others.

It occurs to me I don't recall if you measured the PSU with a meter or not... if you were able to do so, how was the 5v? There are three causes I can think of for something like this, excessive current draw, bad connection, or faulty PSU - I'd ruled out the PSU after seeing that it had been running for quite some time with normal readings, but if not metered it should be checked anyway.

The ATX connector on the board itself though is likely damaged to the point of not having much contact area left on some of those pins, and should be replaced IMO to prevent future happenings.

[c627627]

Thank you.

mosfets appear to be normal and not hot. I have no way to measure the PSU with a meter.

The only thing that scares me about RMA'ing the mobo is the time it takes fro mobo RMA turnarounds.

After reading your post, it is possible that Gautams explanation is the closest to what's behind this:

Originally posted by Gautam

I'm thinking that your case got jerked too hard one day and the connector got a little loose. That's more than enough to cause melting like you encountered and the shutdowns. Whatever the cause may be, you should RMA both. If the ATX connector on the motherboard is warped alone, its a recipe for future disaster.

The PSU connector has that plastic latch so I don't see how it could get disconnected + the case wasn't moved or hit though...


Question: Can overclocking a loaded system cause this by itself?

I do have two 7200 RPM 120 GB drives on + all the cards + scanner and two printers connected along with a TV card & camera.

Was doing 195 x 11.5 = 2242 MHz.

[CrashOveride]

I would RMA the board, when it happend to me the ATX connector was FUBAR, the board might still work if I hadn't tried to fix it by soldering all the wires onto the back :rollyes:

[Oklahoma Wolf]

I think you may be right about the connector getting jostled a bit, though if the wires are stiff and are sort of pulled out of the way for asthetic reasons, that may also cause poor connection. I do believe your board is fine electronically now though.

One thing I noticed was that the pin most damaged was the one with the 5v sense wire - two wires running to one connector sleeve could very well deform the sleeve enough to not make good contact if these wires were strained. The second most damaged connector seems to be the 5v line adjacent to it - an interesting coincidence, perhaps the sleeve on that one got deformed as well.

Originally posted by c627627
Question: Can overclocking a loaded system cause this by itself?

Possible, but unlikely in my mind... at least I hope not, anyway. My own system runs from the 5v, and with BIOS readings as low as mine get it's a mild worry for me. Fortunately I can replace the connectors if need be, but I still don't want it happening in the first place ;)

[CJ145]

Here is what happened to mine... just a few days ago. Antec TP 350 running a 8kha+ with a tbird 1.95vcore (only hdd, cpu, gforce2)

http://ufo.zerobrains.com/images/burnt.jpg

[CrashOveride]

Holy freaken Antec PSUs and Epox motherboards!!!

maybe Antecs just don't have the 5v juice for it... I havn't seen this happen to anyone with a non-antec PSU and non Epox board... Has anyone else?

[Oklahoma Wolf]

Had to go to another forum to find this, but figured it might throw a bit more light on the issue:

http://forums.amdmb.com/showthread.php?s=&threadid=86737

[c627627]

The basic question is: Does this mean that you simply need a brand name PSU rated a lot higher than 350W to avoid this problem or what?

Antec RMA dept is answering machine only today, left a message, if I replace this PSU, should I simply sell it and get a higher rated one?


Thanks for the link Oklahoma Wolf

"Reason for the prob : The resistance of contacts was too high
because of corrosion, dust, not properly seated, loosened over
time, too low power of female spring contact (on cable), or other
reasons."
E-Blitz


Once again thank you all who posted, you've helped a lot.

[c627627]

I just realized I bumped that thread at amdmb forums....three years after it was started... oh well :santa2:.

[c627627]

Talked to Antec, finally, they need to determine if PSU gave out to allow a claim for mobo, but get this, I go, I don't want this happening again, can I get a better PSU from you and just pay the difference and they're like, yeah all right. (!)

So:

My burnt SL350 +

$10 for TruePower 330.
$30 for TruePower 380.
$40 for TruePower 430.

$ was a prob so I chose the TruePower 380, should I have saved $20 and gone with the 330 or paid $40 for the 430 watter?

[CrashOveride]

I would get the 40 dollar more one if you have the money.

[CJ145]

I did notice a bit of dust inbetween the wires but I don't see how it could have worked it's way down into the connectors since it haden't been removed in a long time. I can't rma the psu cause I had a solder point on the AUX power connector for my 5v mod (5v to MOSFET's).

[Dubbin1]

This makes me think twice about getting a Antec PSU :-/

[MameXP]

here is another antec PSU fail situation. Its just blown up yesterday. Its *high rated* Antrec True 480. Way worse than yours :http://www.ocforums.com/showthread.php?s=&threadid=275811

I called Antec but couldnt talk to anybody so i will call again tmr. :( Read the thread and you'll see, i hope my mobo and CPUs are still working

[puji]

I would like to point out again, what could be deducted reading through the other link: burnt out wire/connector is caused buy too much heat. This is only caused by either too much amps or a resitance ( through oxydation poor contact etc, ) So do not blame the power supply for those burns( except it has a defect power connector). Overclocking can be definatly a cause (too much amps above specifications ) or as mentioned loose connectors, bad seated connectors oxydised connectors a defect motherboarde etc. I have encounterd quite oftenburned contacts in cars and never blamed the alternator or battery for it.