• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Seriously P O'd about EVGA GTX 460s...

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

ssjwizard

Has slightly less legible writing than Thideras
Joined
Mar 12, 2002
Ok so yesterday I was playing a game as well as a few other things and suddenly my system shut down...

After much fiddling I finally restored power only to have 1 of my 2 460s catch fire.... I spend most of the evening/night dismantling my rig to remove the burnt card.

So today im playing dirt 2 just about 15 mins ago after 3-4 matches suddenly my computer shuts down again.... I pull the side panel off and start feeling around(I had HW Monitor open on my 2nd monitor the whole time watching temps) to see if anything felt abnormally hot. Low and behold I touch the VRM heatsink and it burned me instantly...

So I let the rig sit for several mins and hit the power button, guess what happens next?

My remaining GTX 460 VRM section decides to catch fire.......

WTF EVGA Im running 2 of your cards AT STOCK DAMN SPEED with water cooling in a case with 4x 140mm yate highs, and 2x coolermaster 120mm medium speed fans. My CPU temp never broke 42c and the GPU temp never topped 50c WTFITS! I have had this particular pair of cards for just over a month and outside of a single benching session they have been running stock on water cooling(XSPC RazR FC blocks).

This is total BS there is absolutely no reason that TWO cards should fall victim to this same failure so fast. I have one of the stock HS but it has a broken fan, IF I order a fan I can RMA 1 card at a time and MAYBE in 2 months Ill have my rig up and running again WTF!
 
Ok so yesterday I was playing a game as well as a few other things and suddenly my system shut down...

After much fiddling I finally restored power only to have 1 of my 2 460s catch fire.... I spend most of the evening/night dismantling my rig to remove the burnt card.

So today im playing dirt 2 just about 15 mins ago after 3-4 matches suddenly my computer shuts down again.... I pull the side panel off and start feeling around(I had HW Monitor open on my 2nd monitor the whole time watching temps) to see if anything felt abnormally hot. Low and behold I touch the VRM heatsink and it burned me instantly...

So I let the rig sit for several mins and hit the power button, guess what happens next?

My remaining GTX 460 VRM section decides to catch fire.......

WTF EVGA Im running 2 of your cards AT STOCK DAMN SPEED with water cooling in a case with 4x 140mm yate highs, and 2x coolermaster 120mm medium speed fans. My CPU temp never broke 42c and the GPU temp never topped 50c WTFITS! I have had this particular pair of cards for just over a month and outside of a single benching session they have been running stock on water cooling(XSPC RazR FC blocks).

This is total BS there is absolutely no reason that TWO cards should fall victim to this same failure so fast. I have one of the stock HS but it has a broken fan, IF I order a fan I can RMA 1 card at a time and MAYBE in 2 months Ill have my rig up and running again WTF!

Did you have some sort of airflow over the VRM Heatsinks?

They are not designed to be passively cooled... They are designed to have airflow over them just like the factory GPU heatsink would....
 
I had an 80mm blowing air down across the two cards from the HDD cage to the back of the case. It wasnt directly blowing on the VRM but I would have thought it to be sufficient... What makes this worse is the pair of 460-SE I had before this didnt even have a HS on the VRM and I ran those at 1.2v OC to 975mhz and had no problems.
 
Wow, have you called EVGA yet? How old are they? I read somewhere recently that 460 VRM sections run pretty darn hot, but it sounds like your airflow should have been more than enough, especially on stock voltage. I can't believe that both died, seems like there has to be some other factor at work to kill both cards..
 
I cant imagine it to be the PSU although I suppose it COULD be the culprit, but I find that unlikely. Its a TX 750 that I got just over a month ago and my +12V reading in HW Monitor reads 12.1v pretty much constantly except for extreme GPU and CPU loads together when it dips down a tiny bit.
 
Thats terrible news... that could have happened to Gigabyte, Zotac, Galaxy, Evga, Asus.........etc.
 
While I do have to send a disproportionate number of EVGA graphics cards away for warranty I am very surprised that you managed to get two that combusted because I'm talking about roughly a 3% failure rate in the first 3 years and even then it's often due to abuse. (I'm not suggesting you were abusing yours)

Anyway, if you don't have a fan blowing a fair amount of air on that area of a 460 I would expect problems to arise over time. I wasn't dumb enough to touch mine before adding heatsinks but after adding heatsinks I was able to burn 4 thin lines into the tip of my finger(heh, not as smart as I thought I was!). That's with a 230mm intake fan blowing directly onto the card. If you've got an 80mm 5 inches away blowing air at the cards then I would really say that's not enough unless that 80mm fan is spinning mighty fast. Adding a VRM heatsink also lowered the GPU temp on my 460 by 10 degrees so you can just guess how hot that area of your card would be getting. Of course, you touched it so you know. It should serve as a lame lesson for the future though. There is more to cooling than just covering processors and memory.

By the way I know you have heat sinks on them so don't misinterpret my dumb comment.
 
I am considering a few options as to what I'm going to do when I get these cards back. Obviously I need to do something to keep these VRMs cooler. The simplest option is going to be attaching a fan to the HS on the VRM but how much airflow do I really need. Taking a few rough measurements it looks like the mounting holes around the VRM sink(attached with thermal tape) is 52mm across with 10mm horizontal spacing.

So I have a few potential ideas,
1 bend up a simple and small bracket using the two holes around the VRM sink to hang a 40mm fan(PWM?)

2 pull out some of my leftover acrylic and make a half shroud that screws down into the attachment points for the VRM part of my FC blocks(didnt fit these cards) and hooks over the edge of the card with something like say a 60mm fan on it.

3 I do still have machining tools(though I would need a bit) and could probably make a NEW VRM part that attaches to the waterblock for these VRMs. Hand machining something as detailed as this is kind of a PITA though.. I just dont have the cash to hire CNC work right now.
 
Yes they will RMA the cards as long as I return it with a complete factory HS/F. I am waiting for the replacement fan to arrive so that I can get the RMA process rolling.

Looking at the block and card earlier It looks like I can actually whip up a VRM attachment relatively easily. I wont be able to tie it directly to the copper base but whatever. My plan is to replace the three top screws closest to the VRM with ones that are 1/4" longer. Then I can take a 4" length of 1/8x2" aluminum flat and bolt it into the steel top. I need to then come down about 9mm to hit the VRMs so I plan to take 3 1" lengths cut 3/4" off of two of them to allow bolting down to the PCB. Ill take the two cut parts and join them in the center to the 3rd uncut 1" strip and bond them with thermal epoxy. Then I can attach it to the cross bar at the top the same way. So ill use up about 9" of flat bar per card so Ill be ordering a 2' length.
 
Back