possible wrecking of my graphics card???

  • Thread starter Thread starter dave_b
  • Start date Start date
D

dave_b

dear al

i have just installed a wooter cooling unit into my pc and replace
the cpu and gpu fan with a waterblock that provides sufficien
cooling- i've checked the CPU, and graphics temps. My problem is tha
since i've done it i'm getting gaps in the rendered video..

on boot i get several strings of exclamation marks across parts of th
screen as well as multiple holes in the asus bios startup picture

i've taken a couple of screenshots from the ati catayst that show
this problem and put them on uploadr.com... Any thoughts as to what'
gone wrong??? I was fairly careful when changing the fan to th
waterblock and its got heatsinks on the memory..

Pretty gutted, but at least i've got a quiet pc

http://file.uploadr.com/741
http://file.uploadr.com/741
 
....
on boot i get several strings of exclamation marks across parts of the
screen as well as multiple holes in the asus bios startup picture.

i've taken a couple of screenshots from the ati catayst that shows
this problem ....
Click to expand...

The video card seems to have a problem. Maybe re-seating the card will help.
So the new water block evenly touches the GPU? Look at the thermal paste
pattern to check. Perhaps when you removed old GPU cooler the flexing of
board caused a marginal solder connection to break. I suggest you put old
heat sink back on to see if fixes proiblem. If not, RMA card.
 
Kent_Diego said:
The video card seems to have a problem. Maybe re-seating the card will help.
So the new water block evenly touches the GPU? Look at the thermal paste
pattern to check. Perhaps when you removed old GPU cooler the flexing of
board caused a marginal solder connection to break. I suggest you put old
heat sink back on to see if fixes proiblem. If not, RMA card.
Click to expand...
RMA card ? so the card gets damaged by buyer, and they send it
back. Not. This is what drives the cost of hardware up. Unethical.

To the OP, it looks like missing textures, and sounds to me like
defective ram on the card. I doubt it is heat, because it shows up on a
cold boot, and the gpu wont get blazing hot that fast. You sure you took
good precautions when handling the card ie preventing ESD or static ?

Short of removing the card and reinserting, and checking power
cable maybe trying another one, you could have a permanently damaged
video card. let us know what you find.
 
I would argue that send out the cards with difficult to remove and
inadequate cooling systems (not efficient enough, too loud, etc), so you
need to replace them. And since they're difficult to remove, you risk
breaking the card when doing so.

The CPU manufacturer's learned how to make it relatively easy to upgrade the
cooling systems; it would be nice if the graphic card manufacturer's
followed suit. Really, why do they have to glue the frickin' things on?
Bolts, clamps, something other than a "permanent" glue or tape would be
really nice. Especially when the thing sounds like a jet engine when it
kicks in under full load. I wouldn't be considering taking mine apart if it
wasn't noisy, and I'm not even overclocking it. And I've got the side of my
case off, so it's not like it's got limited circulation of air.

None of the above justifies lieing to the manufacturer to get an RMA,
however. But I didn't see anybody post a message suggesting lieing to them.

Clint
 
thanks for the replies.. i tried reseating the card many times stil
no joy, but i remembered i hadn't put some clear plastic washer
inbeween the standoff or "nipple" in their words! - and th
PCB of the video card. The metal peg was touching the surface of th
board so its possible that it was causing a mild short or formed a
inductance loop cos it seems to be working now...

Pretty odd, i'll have to keep an eye on it

Unfortunately the card is manufactured by Grandmars if you're eve
heard of such a company - i think they're korean or chinese - i go
it off ebay 2nd hand so no RMA i'm afraid :

Oh well, fingers crossed.
Thanks
Dav
 
Clint said:
I would argue that send out the cards with difficult to remove and
inadequate cooling systems (not efficient enough, too loud, etc), so you
need to replace them. And since they're difficult to remove, you risk
breaking the card when doing so.
Click to expand...

In all the cases I've seen, the stock cooling is adequate for the intended
speed
of the card. If you overclock, you do so at your own risk.

Tom Lake
 
Well, in my case, I'm running a Sapphire X1800GTO. The cooling system will
keep it under 70C without the cover on my case (Antec Sonata II). Haven't
tried it yet with the cover on, but I don't imagine it will get much better.

The bad part is that it does it with a lot of noise. The factory fan/cooler
unit is quite loud at full speed, which it hits trying to keep it cool under
full load.

Clint
 
Clint said:
Well, in my case, I'm running a Sapphire X1800GTO. The cooling system will
keep it under 70C without the cover on my case (Antec Sonata II). Haven't
tried it yet with the cover on, but I don't imagine it will get much better.
Click to expand...

That sounds way too hot - is that idle in the OS or after running some
graphics intensive processes? OK, so my 1800GTO has a non-stock cooler
on it as standard from HIS, but I doubt that's reducing my temps by
20-30'C over your Sapphire!
The bad part is that it does it with a lot of noise. The factory fan/cooler
unit is quite loud at full speed, which it hits trying to keep it cool under
full load.
Click to expand...

Again, my HIS never goes above 57'C under a hefty load, and that's with
my fan at 36% (and in an SFF case - Sugo SG01-E)
 
That's 70C under load (using the ATI Tray Tool to test). The temp is kept
under control at about 60% fan speed, I believe. I'm at work now, and the
card is in my home machine.

And for Tom, I was never desiring that Sapphire and others would warranty my
screw-ups or mistakes. Like if I'm dancing around a carpeted room,
listening to Cyndi Lauper, and using my card as a lip-sync mic. Or if I
install the new HS improperly, and figure the extra heat will do the card
good, so I'll use it to keep my coffee warm. Or if my screwdriver slips as
I'm prying off the old HS and runs a nice gouge accross the traces on the
PCB.

But what I do wish is that they wouldn't use the fact that someone puts a
different HS on as an excuse to void their warranty. The story I've heard
is that they can determine if damage is caused by overclocking a card, or if
it's simply defective. If they can get down to that level of detail, they
should be able to determine if the after-market HS caused any issues.

I find it amusing that they offer the ability to overclock the snot out of
the cards so easily through software, but putting on a new HS involves
breaking out the hammer and chisel.

Clint
 
Back
Top