Got my GPU back from warranty (560ti)....NOT HAPPY

[mkristian]

So i made a thread a while back talking about how my GPU all of a sudden booted into 100% fan mode whilst playing skyrim despite temperatures not exceeding 60 degrees clecious at load or 40 degrees celcius at idle. My computer shut off a minute after that happening and whenever i tried to boot my PC with the GPU from then on it would automatically go to 100% fans straight away and nothing would show on the screen. I did all the usual stuff including an entire re stall of windows and I as well as members here determined it was indeed a GPU problem.

So i take my less than 4 month old card into the shop. In email discussions with them I was informed they would test the product themselves, and if they found a problem with it they would send it off to Asus. So a month later it has been to Asus and back and i was told the GPU was in normal working order which is one hell of a contradiction considering, according to the shops policy, they test the products themselves so one can only assume they found a problem with the card . Sceptical i took the card home and tried it out. Now the weird thing is, i had spent almost a week trying to get this card to work, and all of a sudden i put it into my PC and it works. Strange right, I know something must of been done to it yet they passed it off as a fully functioning card which IMO is a straight up lie because the card simply would not boot before i had sent it off. As well at this the fan speed was in check. Great, whatever, it works what do i care right?

Well within an hour, whilst watching some NBA highlights the card boots into 100% fan mode again despite a temperature of 41 degrees. Weird, whatever i kept on browsing until i started smelling the same smell that occurred when it broke in the first place. Something was burning. I immediately turn off my PC, and dismantle my PC to check for burning. Nope, it was definitely a card problem. I put it all back together making sure my wiring is super stringent, this time the PC turned on for 2 seconds then swithed off. From there it would not start at all. So at this point i just started using the onboard video on my mobo which works fine. Before i go to bed i decided to try it one more time.

I put my GPU back in my PC, and as soon as turn it on (it actually booted this time) a big spark flys right out from the side of the card on the MOBO side. Before i had a chance to turn my PC off, something i can only describe as a jet flame was emerging from the place it had just sparked and i could see the bright yellow flame lighting up a large area of the case. Needles too say this scared the absolute bonazas out of me and it scares me even more to think this could of happened whilst the computer was running unmonitored.

I have emailed the shop and told them of the inherent contradiction concerning their policy and Asus's statement that the card is fine. If the card was damaged which I can only assume because of the shops own testing policy, why and how could Asus state the opposite without any concern? I told them there is noway i am putting this card back into my PC repaired or not and that i demand a replacement. Inspecting the card i can see where the damage is, on the side with the two black insulated cables, one of them is very frayed and i assume that was the source of the spark/power surge.

Sorry about the wall of text, anyway the point is do you think my request for a replacement card is reasonable? I have waited over a month for this problem to be rectified only to be told that there is no problem despite an inherent assumption that there is a problem.

 

[amuffin]

You must have a crap power supply.

 

[Rockdpm]

I agree, sounds like either a Power supply issue. or faulty PCI express ports on the card could have blowed. However your theory could be right. I still would make sure the PSU isnt at fault

 

[mkristian]

antec neo eco 520c, never overcloked once or changed voltages, been told many times this would be a suitable PSU for a sinlge 560ti. I run a modest system with a 95w phenom, 1 HD and 1 DVD drive and a couple of case fans. If it is a PSU problem ive dun goof'd. I guess ill go buy another power supply and see if it works.

edit: how does an inadequate PSU cause a GPU to start flaming? I are confused

 

[biscuitasylum]

If you knew how to use a multimeter, you could check voltage from the PSU

 

[mkristian]

just put it back in, seems to working ok so far. Sure as hell won't be leaving this thing on over night or when im out of the house tho.

 

[biscuitasylum]

I would still check that power supply with a multimeter.

 

[mkristian]

wish i had one or could afford one

 

[biscuitasylum]

wish i had one or could afford one

Every man should have a multimeter and learn to use it.... its one of the most basic tools to keep around the house.

You can get them really cheap too. you dont need an expensive one... you can get them under 15 dollars easily. I've had one since I was 10 years old.

http://www.sunstons.com/index.php?main_page=product_info&products_id=5222&i=B000EVYGZA

of course now Ive upgraded to a Fluke which cost me several hundred dollars... but I do a lot of extensive testing.

this is mine...

http://www.amazon.com/Fluke-FLUKE-1587-Insulation-Multimeter/dp/B000VRFJU8


Save a few pennies and pick one up for ten bucks and learn to use it. Seriously... its the best under 20 dollar tool in a tool box next to the crescent wrench and hammer.

 

[mightymaxio]

I agree test it with a multi-meter if you don't know or care to learn how to use one you shouldn't be working with custom computers period. Heck i know elementary school kids that use multi-meters.

 

[biscuitasylum]

I agree test it with a multi-meter if you don't know or care to learn how to use one you shouldn't be working with custom computers period. Heck i know elementary school kids that use multi-meters.

lol, Yeah My father made me learn how to use one. The old needle styles.

I do agree... its very wise to have and know how to use a multimeter while building computers... in the long run you'll save a lot of headaches and money.

 

[mightymaxio]

My parents are both engineers and taught me at a very young age to use one. I started with testing batteries and then self taught myself on other things such as current with surge strips.

They are not that expensive either, cheaper than a power supply tester for sure by about 5-10$

 

[mkristian]

didnt think they were that cheap, used them plenty of times at school, ill go pick one up

 

[biscuitasylum]

Thats awesome Mightymaxio... Good for you.


Good deal mkristian... Yeah, you can get them cheap. The Needle meters as I call them which are non digital can be as low a 5 or 8 bucks.

 

[mkristian]

just got one, 20 dollars which i imagine is a good price within Australia considering everything costs a bit more over here. Thanks for the help everyone i really appreciate it. Just going to learn how to do this now and will report back with results.

 

[mightymaxio]

Sure no prob post back with your results and we will be glad to assist you.

 

[biscuitasylum]

Congrats! Yeah, let us know.

 

[mkristian]

Hey so I have been up all night just learning about voltages and playing the odd game to try out my GPU. So far its working without a hitch but i went ahead and tested some voltages anyway.

I have tested the 5v and 12v rails from a spare molex adapter whilst my PC is running and got 5.12 and 12.28 volts respectively. Is that it? If so does that officially mean my PSU is ok? Or do i have to test out the the 6 pin adapters that plug into my GPU as well?

Still scares me though that something was burning in that GPU, maybe it was some residue or something I have no idea all i know is it works fine atm at very nice temps too.

 

[biscuitasylum]

I would test the 6 pin GPU as well. Just to be safe. Its good to know anyway.

You know, this may sound odd... But I've had issues where Ive installed a GPU card, had issues... then took it out and reseated or reinserted the card and bam... ran like a champ. lol

Computers sometimes tend to have a mind of their own.

Yeah Check the voltage for the GPU and write your voltages down. Keep track, that way next time something comes up, Immediately check those voltages with what you recorded from your check. That way you can see if you have a fluctuation in your readings.

Good to hear you are up and running so far. Cross your fingers... lol

 

[xtcx]

Hey so I have been up all night just learning about voltages and playing the odd game to try out my GPU. So far its working without a hitch but i went ahead and tested some voltages anyway.

I have tested the 5v and 12v rails from a spare molex adapter whilst my PC is running and got 5.12 and 12.28 volts respectively. Is that it? If so does that officially mean my PSU is ok? Or do i have to test out the the 6 pin adapters that plug into my GPU as well?

Still scares me though that something was burning in that GPU, maybe it was some residue or something I have no idea all i know is it works fine atm at very nice temps too.

Did you try cleaning your GPU card since that incident?. If you suspect some residue left, just remove the fan and heatsink which doesn't void your warranty. (Make sure you dont damage the warranty sticker covering the screws or other parts).

One more point is, since you said your are able to boot, goto Bios and there you can check the all voltages that come from PSU. You may not need multimeter in this case. However if your PC does not boot\other such cases, you can use a multimeter then as "biscuitasylum" suggested. But a low quality multimeter will show more variations and results will not be more accurate. Also never attempt to measure the current .

One more, make sure you are always inserting probe into Voltage socket (V Ohm) of a multimeter, not into current (20A\mA) measurement socket. Because in current measurement mode, it will short circuit your PSU lines , so be careful
....Just an advice

And please make yourself seated in a proper ground\conducting material that will ground you, your ESD can still make things worse when you deal with motherboard\other accessories. Also you can discharge yourself first by standing on ground for a moment and step-in.

 

[xtcx]

antec neo eco 520c, never overcloked once or changed voltages, been told many times this would be a suitable PSU for a sinlge 560ti. I run a modest system with a 95w phenom, 1 HD and 1 DVD drive and a couple of case fans. If it is a PSU problem ive dun goof'd. I guess ill go buy another power supply and see if it works.

edit: how does an inadequate PSU cause a GPU to start flaming? I are confused

In this case, it doesn't make me think much about PSU until the PSU is too unrated brand. When there is more load requirements at the output of PSU above the PSU rating, it can only damage the PSU not the load. However the wrong voltage\higher voltage can definitely ignite the lighting shows, but how much in terms of voltage?, say 12.2V@45A as normal, is the PSU such a crap\poorly designed model to deliver some 20V@same 45A?. Definitely when the voltage increases, the current rating of 45A will drop a bit. So this may not be the reason for the flames, etc.

I strongly believe this can't be much of PSU problem. Anyway, the 12V rail is common for all such as 26-pin\6-pin. Just open your bios and keep monitoring your 12V rail for some time and take a log. Also try checking this voltage with GPU and without GPU as well as with\without HDD\DVD drive, etc. This is to test the PSU whether the loading is the actual problem or not. The more load you add, you can see more fluctuations in voltage.

GPU card, when inserted wrongly can also be the reason.

Is your card working right now?.

 

[biscuitasylum]

Did you try cleaning your GPU card since that incident?. If you suspect some residue left, just remove the fan and heatsink which doesn't void your warranty. (Make sure you dont damage the warranty sticker covering the screws or other parts).

One more point is, since you said your are able to boot, goto Bios and there you can check the all voltages that come from PSU. You may not need multimeter in this case. However if your PC does not boot\other such cases, you can use a multimeter then as "biscuitasylum" suggested. But a low quality multimeter will show more variations and results will not be more accurate. Also never attempt to measure the current .

One more, make sure you are always inserting probe into Voltage socket (V Ohm) of a multimeter, not into current (20A\mA) measurement socket. Because in current measurement mode, it will short circuit your PSU lines , so be careful
....Just an advice

And please make yourself seated in a proper ground\conducting material that will ground you, your ESD can still make things worse when you deal with motherboard\other accessories. Also you can discharge yourself first by standing on ground for a moment and step-in.

A low quality multimeter wont fluctuate to a point you wont get a good reading. Ill argue this till the death bed. This isn't the 70's anymore.

 

[xtcx]

A low quality multimeter wont fluctuate to a point you wont get a good reading. Ill argue this till the death bed. This isn't the 70's anymore.

:lol: Agreed
But when the precision comes to 3 digits are more such as 12.535 etc, how better can a basic multimeter be precise?.

 

[biscuitasylum]

:lol: Agreed
But when the precision comes to 3 digits are more such as 12.535 etc, how better can a basic multimeter be precise?.

With a keen eye, you can average out readings per say a number of times and see what ya get. the average of 5 separate readings should give you a decent average on what its doing.

Quite honestly I would be more concerned with a whole volt over lets say a half a volt.

 

[xtcx]

With a keen eye, you can average out readings per say a number of times and see what ya get. the average of 5 separate readings should give you a decent average on what its doing.

Quite honestly I would be more concerned with a whole volt over lets say a half a volt.