If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.
![]() |
|
|
#1 |
|
Regular
|
Place your bets
![]() ![]() I found this lying on the floor of the case, after the system re-booted and wafted a burning smell my way I was playing Operation Flashpoint at 2048x1536. Is that a memory module's heatsink? I haven't tried 3D (will wait until a replacement X1950Pro 512MB AGP arrives) but 2D is working... Jawed |
|
|
|
|
|
#2 |
|
Dangerously Mirthful
Join Date: Feb 2002
Location: Winfield, IN USA
Posts: 15,292
|
A memory chip fell off and it's still working? Damn, you one lucky dude for being unlucky enough to have it fall off!
__________________
Elite Bastards - Adminish “Be polite, be professional, but have a plan to kill everybody you meet.” - General James N. Mattis |
|
|
|
|
|
#3 |
|
Red-headed step child
Join Date: Jun 2004
Location: Guess ;)
Posts: 3,084
|
The x800 (and nearly every other video card made after about '00) uses BGA memory chips; that's not a memory chip that fell off. With so many chips these days on the PCB, there's no real way to tell... But with it being heatsinked, I'd wager it's something like a voltage regulator or the like.
Ouch!
__________________
"...twisting my words" |
|
|
|
|
|
#4 |
|
Senior Member
|
A clean shot of the board would be graceful.
__________________
Apple: China -- Brutal leadership done right.
Google: United States -- Somewhat democratic. Microsoft: Russia -- Big and bloated. Linux: EU -- Diverse and broke. |
|
|
|
|
|
#5 |
|
Regular
|
Looks like some kind of voltage regulators. Two of them per heatsink, with one still hanging on for dear life! :
![]() As you can see from this picture the one that fell off was at a bit of a disadvantage, facing the full heat of the hot air being blown off the core. The other heatsink over there on the left doesn't get the benefit of this "cooling".
The memory chips on this board are on both sides - the chips on the back of the board don't have any heatsinks on them, so I guess they're safe. This is the All in Wonder version of the board, in case anyone's wondering why it looks different. When I take the board out, I'll take a proper pic. I'm just keeping my fingers crossed this card will keep running until the replacement is delivered, sometime Thursday. I'm still in shock that it works at all! I was contemplating getting my trusty Radeon 32MB SDR out of storage... Jawed |
|
|
|
|
|
#6 |
|
Junior Member
Join Date: Apr 2004
Location: Calgary, Canada
Posts: 57
|
Hopefully it's something related to the AIW functions that allows it to run fine until the replacement's ready, don't think that original Radeon's gonna cut if for Flashpoint.
Never good when Sh... Stuff starts falling off, but at least you know that it's definitely time for an upgrade, no hehming and hawing and wondering if you could've gottan a better deal, etc. When I first saw it reminded me of a Radeon 9500P that had the more tragic fate of it being the BGA memory on the other side of that heatsink.
__________________
"I'm Sorry That would be playing God." "GOD Shmod, I want my Monkey Man! " |
|
|
|
|
|
#7 |
|
Locally Operating
Join Date: Dec 2004
Location: QLD, Australia
Posts: 1,773
|
That's just terrible... Look at that poor coil beside it. Is that just dust or did it cop some heat damage too? You could always peel off the heatsink and see what chip that was, then if you really want, find and solder a new one on
__________________
Valve Software - Giving me Episodic nightmares Bulletstorm - I Will Kill Your Dick! |
|
|
|
|
|
#8 | |||
|
Regular
|
Quote:
Oh, and the Radeon plays Flashpoint quite fine at 1024x768. 16xAF is stretching it, admittedly. I was amazed at how much more playable the frame rates got when I changed from an Athlon 1200 to an A64 3500 while leaving the Radeon unchanged... Quote:
Quote:
Jawed |
|||
|
|
|
|
|
#9 | |
|
Regular
|
Quote:
I wonder if the guy who loaned me this wants it back... Jawed |
|
|
|
|
|
|
#10 |
|
Tiled
Join Date: Oct 2003
Location: Kings Langley, UK
Posts: 2,675
|
Heh, the number of dead boards in my collection probably outweighs those that work, so when you're finally done you can send it back for that. Glad it did you reasonable service, sorry I don't have any more AGP to go your way. If you go PCIe, let me know and I'll throw you a board.
__________________
A major redesign of the core ALU pineapple boomerang fortress. |
|
|
|
|
|
#11 |
|
Regular
|
Cheers Rys! I do actually plan to shell out for the latest and greatest at some distant point in the PCI Express 2.0 future, ooh, prolly 18 months from now. Gotta get decent mileage out of the new toy first...
Jawed |
|
|
|
|
|
#12 |
|
Moderate Nuisance
Join Date: Feb 2002
Posts: 4,653
|
So did it fall off b/c the solder failed or its little metal fingers gave out? Seems like it's worth a try soldering it back on, if it's still working now. Of course, then you'd want to replace the glue with something fancier like Arctic Silver cement, so a slippery slope with no guarantee of success.
Funny, my 9800P gave up the ghost, too, though I can't tell if it's because the Arctic Silencer on it gave out (it was touchy, due to a dodgy jerry-rigged fan connector) or because something else failed. I can definitely understand the stock fan failing. I have a 9100 whose fan failed so many times that I gave up replacing it and let it run fanless. It's still working right now in this low airflow system. Of course, it's not even pulling 25W. |
|
|
|
|
|
#13 | |
|
Senior Member
Join Date: Dec 2006
Location: In front of the PC.
Posts: 1,063
|
Quote:
If I was to dare to speculate I'd say the voltage regulator failed and shorted out internally and got really really hot. Thus melting the solder to the board and falling off. It might not be a good idea to cement it in place and solder it back on because if it IS shorted it'd just keep drawing power if it can't fall off and possibly heat up to the point of causing a fire.. I'd consider the card a goner myself just to be safe. Who knows what's happened with it. The remaining voltage regulator of that pair isn't even getting any cooling anymore. If it hasn't died yet it probably will soon and then that'll likely be all she wrote for that card. Peace.
__________________
"Life is only but a long moment." -Moonwolf |
|
|
|
|
|
|
#14 |
|
Regular
|
Rainbow Man's theory sounds pretty good. It seems like a bit of bizarre good luck that it dropped off: the "solder melted", the heatsink came free from the other chip AND the heatsink didn't short anything to the point of destruction in its little suicide dive. AND the card still works. Blimey
I daren't go near 3D - I'm even leary of playing movies just in case they stress the GPU. Jawed |
|
|
|
|
|
#15 |
|
Naughty Boy!
Join Date: Mar 2004
Posts: 1,172
|
When are you going to get rid of that old piece of ....... ?
You do know video cards are used for more than browsing very long techinical threads on certain forums ?!! |
|
|
|
|
|
#16 |
|
Member
|
What your seeing is part of the power regulation. Its possible to repair it by replacing the missing parts. If Rys has bad video cards then you might be able to replacement parts on some of them to repair you X800XT. You will have to find out the part numbers on the IC used on your card. My guess is both of the ICs are the same part number so if you can read one, your in good luck there. I have about 15 dead cards because the GPU is fried and use them to scavenge parts off them to repair reparable cards with.
Last edited by {Sniping}Waste; 08-Aug-2007 at 19:10. |
|
|
|
|
|
#17 |
|
Dangerously Mirthful
Join Date: Feb 2002
Location: Winfield, IN USA
Posts: 15,292
|
Snipe is the GAWD at busted card repair!
*digitalwanderer bows humbly to the Snipe I still think grateful and amazed thoughts your direction Snipe every time I fire up a 3D game or play with my X1800GTOs, I still can't believe you pulled that off.
__________________
Elite Bastards - Adminish “Be polite, be professional, but have a plan to kill everybody you meet.” - General James N. Mattis |
|
|
|
|
|
#18 | |
|
Regular
|
Quote:
It does make me wonder if Rys is trying to build up a "musem" or if it could serve as a cache for someone who wants to make FrankenGPUs with the aid of a soldering iron... Jawed |
|
|
|
|
|
|
#19 |
|
Tiled
Join Date: Oct 2003
Location: Kings Langley, UK
Posts: 2,675
|
Hehe, my missus wonders too given the space they take up in my office. Nearly 100 boards of varying descriptions, ages, sizes and uses, which I'll catalogue and write about for the site one day. Some have interesting provenance, some are odd and some are downright interesting (multi chip board with four NVIDIA GPUs, built by 3dfx after they were bought, anyone?), so it should be a reasonable read!
__________________
A major redesign of the core ALU pineapple boomerang fortress. |
|
|
|
|
|
#20 |
|
Merrily dodgy
Join Date: Aug 2003
Location: The colonies
Posts: 1,398
|
After having seen my wife's Sapphire 1900XT go up in smoke a couple of weeks ago, and my own Asus refusing to do anything anymore since last night other than burn a little red led which indicates T_FAULT (which according to a large number of hits on google, means I'm buggered), I'm having some serious doubts about the build quality of some ATI OEMs...
|
|
|
|
|
|
#21 | |
|
Regular
|
Quote:
Generally with ATI there seems little variation in board design/cooler configuration - NVidia cards seem much more varied (though perhaps this varies by region - I'm used to UK/US products). Also, with so many of these companies producing both ATI and NVidia cards, is it actually down to the OEMs? They're often just branding stuff, aren't they? On the other hand the basic stuff from which a board is made, the GPU itself and core components like the power supply, should be common across a whole range of SKUs/brands. I look at the placement of that heatsink that fell off this X800XT AIW and think "bad design putting it in the (genuinely hot) exhaust of the cooler." That's solely down to ATI as far as I can tell, AIWs have always been solely reference design, haven't they?... Meanwhile Scan.co.uk decided to kick my order around the warehouse for an extra day, so gratification comes tomorrow... Jawed |
|
|
|
|
|
|
#22 |
|
Diamond Viper 550
Join Date: Feb 2003
Location: U.S.
Posts: 1,189
|
__________________
"I demand the constitutional right to eat my enemies" - RussSchultz |
|
|
|
|
|
#23 |
|
Junior Member
Join Date: Apr 2004
Location: Calgary, Canada
Posts: 57
|
LOL!
Yes, that's for sure. PS, Good to see you man, it's been too long.
__________________
"I'm Sorry That would be playing God." "GOD Shmod, I want my Monkey Man! " |
|
|
|
|
|
#24 |
|
Darlek ******
Join Date: Jun 2004
Posts: 9,497
|
It shouldnt die, just pine
|
|
|
|
|
|
#25 |
|
Regular
|
Looking more closely, while swapping in my new graphics card (which is working great) I discovered that the near death experience was more serious than I realised:
![]() That's a floppy disk power molex, not a standard molex. And the chip that failed was the chip that was left behind! not the chip that stayed attached to the heatsink: ![]() Note that looking at the pairing, the top chip has disappeared, along with the heatsink. And the bottom chip is a charred mess. I'm afraid to say the failure was prolly my fault: there was a "gauze" of dust across most of the heatsink fins, where the fins are exposed to the draught from the fan. It's interesting that the core itself didn't fail. I think the dust may have been biasing the airflow to the fins that exhausted air across this pair of chips and heatsink - so with the airflow across the rest of the heatsink reduced, the air was hotter than it would normally be. Stupidly, I didn't think to check this cooler when I cleaned out dust from the system a few months back. The CPU fan and heatsink each had an accumulation of dust, which I cleared. Jawed
|
|
|
|
![]() |
| Thread Tools | |
| Display Modes | |
|
|