Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

Geek Culture / Anything DX dies! (New GPU arrived!)

Author
Message
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 07:36 Edited at: 4th Nov 2007 07:38
Link
I think my card is dying, anything I run that needs dx is blowing up into sick and fantastical pixelations, then freezing, then eventually the pc reboots. Everything is the latest. Anyone had this happen before?

Card: NVidia 6800 Ultra PCI-E 256

Description: You start game, all ok. You run around when suddenly theres long objects coming from your chest thru the focal distance + white weird lines that look like 3d modelling lines indicating edges. then everything pixelates (blocky shite everywhere), if you dont try to exit the game at this point its "tick tock tick tock" BOOM! Frozen.

I tracked it back a bit to possibly a dx input problem on my sound card but to no avail. I dont want to have to go buy a new vid card atm, so any help or prior exp with this crapzor would be great.

tks
CR

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Phaelax
DBPro Master
21
Years of Service
User Offline
Joined: 16th Apr 2003
Location: Metropia
Posted: 4th Nov 2007 07:42
Link
Anyone you know have a card you could borrow just to test it out? I never had that problem, but it does sound like a faulty card. Maybe its overheating?


Back to top
Profile PM Email Website
Agent Dink
20
Years of Service
User Offline
Joined: 30th Mar 2004
Location:
Posted: 4th Nov 2007 07:47
Link
Definitely a faulty or overheating video card. I first suggest you find a nice big air compressor and blow all the dust out of your computer if you haven't done that in awhile. Dust holds the heat in big time. If it's still doing this afterward, monitor your temperatures a bit and see if they go really high. I had the GT version of your card and it was common for it to get around 70C while playing games without issue so see if it's running hotter than that.

Warning: Please be advised. Geek Culture is under lockdown. All mods are set to Indi mode. Any and all topics WILL BE LOCKED. Post at your own risk!
Back to top
Profile PM
Benjamin
21
Years of Service
User Offline
Joined: 24th Nov 2002
Location: France
Posted: 4th Nov 2007 07:53 Edited at: 4th Nov 2007 07:54
Link
I had a similar problem, initially where the graphics were messing up. After not so long my PC would freeze up seconds after running anything 3D. It turned out to be the card overheating, or a problem with it that was causing it to overheat. When it was serviced they replaced the graphics card.

Tempest (DBP/DBCe)
Multisync V1 (DBP/DBCe)
Back to top
Profile PM Email
Duplex
User Banned
Posted: 4th Nov 2007 08:59
Link
I have this same problem as we speak, I think it is due to overheating, at the moment, when I want to play a game, I take the side panel off the tower and blow a desktop fan into it. It works better than with the panel on, but I still have to put the game options on the lowest possible settings.

I was thinking on using my restore disk to put everything to the way it was when I first had this PC, cause everything worked fine back then, it's only resently this has been happening.

And it's not the games either, I can't run old games that used to work well on this just so you don't say that my PC most probably can't support the games I'm trying to run.


Professional, afforable, what more could you want?
Get your FREE Standard Edition Today!
Back to top
Profile
gamebird
17
Years of Service
User Offline
Joined: 13th Jun 2007
Location:
Posted: 4th Nov 2007 15:46
Link
Quote: "Definitely a faulty or overheating video card. I first suggest you find a nice big air compressor and blow all the dust out of your computer if you haven't done that in awhile. Dust holds the heat in big time. If it's still doing this afterward, monitor your temperatures a bit and see if they go really high. I had the GT version of your card and it was common for it to get around 70C while playing games without issue so see if it's running hotter than that."


Unfortunately this will not work because if he is expieriencing the effects of overheating then the components will already be damaged. Heat itself does not cause problems, but the damage it inflicts on the card does. All this will do is keep further damages from occuring.
Back to top
Profile PM
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 16:12 Edited at: 4th Nov 2007 16:22
Link
thanks. I dont know why it would be suddenly over-heating as nothing's changed hw/sw wise, unless the new games are slowly killing it (hl2:ep2, portal, tf2) but its been fine for the last two weeks, so it seems odd.

There is a ton of dust in this thing thats for sure, so I will need to get some compressed air and blow it out, to see if that helps.

Instead of playing video games I spent all of last night doing a million things to fix this, none of which worked, but I guess it ruled out a slew of possibilities:

(with play tests in between each item)
-disconnected everything unessential from PSU
-reinstalled latest vid drivers
-rolled back to last drivers
-rolled forward to newest
-reinstalled latest dx
-removed sb live sound board and ran on-board auidio
-reinstalled sb live
-removed gpu from windows and rebooted, reloaded all nvidia stuff and drivers

meh.

I could have sworn there was an nvidia utility that monitors the gpu temp, and can alarm before it crashes. Does anyone know where that went, I cant seem to find it?

also this is the first set of driver from nvidia that didnt want to default to the winXP2K folder, this set (163.71) wanted to make a Win2K folder, which originally I overrode and told it to install where all the previous sets were located. Last night during a reinstall I let it make its folder defaults etc but its still dead.

this really sucks. and whats worse is it happened on a saturday evening of all times lol. the one time a week I play video games Well that would make sense since its only happening in games, but you know what I mean - too bad I didnt run a quick tf2 or something during the week, then I could have rectified it before saturday, instead of wasting my "game night" - grrr

i need to find out if its overheating, if not maybe the mem went south?

thanks so far guys...

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Jess T
Retired Moderator
21
Years of Service
User Offline
Joined: 20th Sep 2003
Location: Over There... Kablam!
Posted: 4th Nov 2007 17:03
Link
Quote: "Unfortunately this will not work because if he is expieriencing the effects of overheating then the components will already be damaged. Heat itself does not cause problems, but the damage it inflicts on the card does. All this will do is keep further damages from occuring."


I beg to differ.

As the components heat (mainly talking silicon chips here), the likely-hood of an electron jumping tracks between silicon sections is greatly increased, which causes the corruption, and eventually completely locking up of the PC.
The hardware usually has built-in protections which will detect the heat up to a certain factory set threshold, then turn it off to stop it being damaged.


Quote: "find a nice big air compressor"


Don't be lazy - just get in there with your hands, pull out as much as you can, then blow the rest out with a few huffs and puffs

Nintendo DS & Dominos :: DS Dominos
http://jt0.org
Back to top
Profile PM Email Website
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 17:27
Link
I downloaded the nvidia diagnostics thing, and it crashed in the middle of the stress tests, or at least thats what it seemed to do.

the logging feature doesnt seem to work either, I select gpu, audible alarm, and try to turn on logging and it says something needs to be selected. wtf is up with that?

this is getting retarded already

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 18:24 Edited at: 4th Nov 2007 18:25
Link
ok I was the one that was retarded with regard to the logger.
I was able to log some statistics while running hl2dm as a test, and got it to crash the machine. Something very interesting is occurring. The GPU Temp is staying within normal ranges (i think, never goes beyond 90c, nvida alarm is set by default to 143c) but the memory usage slowly increases into the 98% range right before it dies. see the log below. I truncated it to show from the time I started the game until the crash.



I just have to figure out if thats system memory or gpu memory, I think its system. Since I never logged this data before this problem I have no idea if its normal or not, but at this point it doesnt seem like a heat issue - any ideas?

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Deathead
18
Years of Service
User Offline
Joined: 14th Oct 2006
Location:
Posted: 4th Nov 2007 18:44
Link
@Cattle Rustler: You can never keep up with technology. As quoted from Red vs Blue in the future there will be "DVDVDVDVDVDVDVDVDBVB"lol

[center]
Back to top
Profile PM Email
Agent Dink
20
Years of Service
User Offline
Joined: 30th Mar 2004
Location:
Posted: 4th Nov 2007 19:11
Link
90C seems pretty hot CR. I don't think I've ever seen my 6800GT above 85C. I still strongly suggest getting the dust out. It will make a world of difference. It's pretty much like putting a blanket around your video card. It just holds the heat in.

Quote: "Unfortunately this will not work because if he is expieriencing the effects of overheating then the components will already be damaged. Heat itself does not cause problems, but the damage it inflicts on the card does. All this will do is keep further damages from occuring."


My card's / processors have overheated plenty of times. They've worked just fine afterwards.

Of course it's not good for them and probably does damage them, but only over prolonged periods.

Warning: Please be advised. Geek Culture is under lockdown. All mods are set to Indi mode. Any and all topics WILL BE LOCKED. Post at your own risk!
Back to top
Profile PM
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 19:25
Link
I think youre right 90c is high based on reading some forums on the issue since I last posted. I will do a dust bust and see what happens. I dont have any compressed air here, should I risk the vaccuum? I know its possible to fry stuff with the static caused by a vaccuum but Im not sure what else to do. Its a disaster dust-wise in that case, lol

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Agent Dink
20
Years of Service
User Offline
Joined: 30th Mar 2004
Location:
Posted: 4th Nov 2007 19:36
Link
You can get compressed air at most office stores. Or you could use a vacuum. My cousin uses a small shop vac in his PC all the time to blow / suck dust out. I don't know if there are any risks involved though and don't blame me if you do it and fry something hehe.

Warning: Please be advised. Geek Culture is under lockdown. All mods are set to Indi mode. Any and all topics WILL BE LOCKED. Post at your own risk!
Back to top
Profile PM
bitJericho
22
Years of Service
User Offline
Joined: 9th Oct 2002
Location: United States
Posted: 4th Nov 2007 19:58
Link
I don't know if I'd want to use a vacuum on it, but a can of air would work perfectly.

On my last gfx card, I had to take a toothpick to the fan as the dust clumps wouldn't blow out. Once I did that, my card actually worked properly


The greatest multiplayer text adventure ever...
Back to top
Profile PM Email Website
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 21:25 Edited at: 4th Nov 2007 21:32
Link
I degunked this sucker. I took out the gpu and the cpu heatsink assembly (which I swore I would never do because its a b*tch to reinstall) and I cleaned out everthing from top to bottom. All of the fins inside the gpu and cpu heatsinks were clogged, thats all clean now . I have to go retest now and see if it makes a diff.

had a minor issue when booting the first time after the cleaning. apparently I didnt seat the cpu heatsink properly (I hate this thing) and by the time windows started up the temp monitor was beeping for the cpu I immediately shut it down and reseated that thing. I count 4 cuts on my right hand so far.

wish me luck in game test...

EDIT: just checked the log from earlier vs. right now for gpu Idle temperature:
before cleaning: 64c
after cleaning: 51c


My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
bitJericho
22
Years of Service
User Offline
Joined: 9th Oct 2002
Location: United States
Posted: 4th Nov 2007 21:36
Link
Quote: "I count 4 cuts on my right hand so far."


Mere battle scars!

Quote: "before cleaning: 64c
after cleaning: 51c"


That's not bad at all! Now you can OC that sucker!


The greatest multiplayer text adventure ever...
Back to top
Profile PM Email Website
Benjamin
21
Years of Service
User Offline
Joined: 24th Nov 2002
Location: France
Posted: 4th Nov 2007 21:39
Link
Quote: "Now you can OC that sucker!"

Yes and then have it overheat again.

Tempest (DBP/DBCe)
Multisync V1 (DBP/DBCe)
Back to top
Profile PM Email
bitJericho
22
Years of Service
User Offline
Joined: 9th Oct 2002
Location: United States
Posted: 4th Nov 2007 21:44
Link
Quote: "Yes and then have it overheat again."


You can always get a new heatsink


The greatest multiplayer text adventure ever...
Back to top
Profile PM Email Website
Benjamin
21
Years of Service
User Offline
Joined: 24th Nov 2002
Location: France
Posted: 4th Nov 2007 21:46
Link
Quote: "You can always get a new heatsink"

Or brain.

Tempest (DBP/DBCe)
Multisync V1 (DBP/DBCe)
Back to top
Profile PM Email
bitJericho
22
Years of Service
User Offline
Joined: 9th Oct 2002
Location: United States
Posted: 4th Nov 2007 21:51
Link
Quote: "Or brain."


You're the one here who needs a new brain!


The greatest multiplayer text adventure ever...
Back to top
Profile PM Email Website
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 21:54
Link
well, the temps now look good in the game, and the mem consumption was lower, but it still crashed. It was very different this time tho, there was no corrupt video at all, the game just froze and the audio went into a loop, then the machine shut down and wouldnt reboot by itself, I had to cold boot it

log


highest temp the gpu hit this time was 59c - a lot better than 90c like before. so now I am stumped again. Is it heat? Is it gpu memory? is it something else? It would be lame to buy a new card to find out its something else. meh.

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Benjamin
21
Years of Service
User Offline
Joined: 24th Nov 2002
Location: France
Posted: 4th Nov 2007 21:57 Edited at: 4th Nov 2007 22:06
Link
Sounds like it's an actual fault with the card now, perhaps caused by overheating before.

Quote: "It was very different this time tho, there was no corrupt video at all, the game just froze and the audio went into a loop, then the machine shut down and wouldnt reboot by itself, I had to cold boot it"

The same thing happened to me. For me, the graphics only messed up the first time, and all other times from that it just froze like you say. Of course eventually the PC crashed while I was playing solitaire (yeah, that's really GPU intensive) and after that it wouldn't boot up with a normal screen, so that was when I had to have it serviced. Be warned.

One thing that I did notice is that I could run a spinning cube demo fine if the framerate was capped to 60, but when running uncapped it froze my PC within a couple of seconds. Maybe you should try this, it might help determine if it is actually caused by high memory usage (dodgy memory perhaps) or just intensive calculations.

Tempest (DBP/DBCe)
Multisync V1 (DBP/DBCe)
Back to top
Profile PM Email
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 22:35 Edited at: 4th Nov 2007 22:36
Link
thanks for that info Ben, unfortunately now the plot thickens. My cpu fan wont ramp up to the right rpm and the cpu is getting hot (75c). now I dont know if this was occurring before all this started or as a result of me taking it apart and cleaning it. I have no idea now. I normally never run the asus monitor in windows so I dont know if this cpu fan thing is new or was present earlier, but I do recall the fans ramping up like normal yesterday and earlier today (during boot up, when a game is run, or some other cpu intensive stuff raises the cpu temp - otherwise the stay at a lower, quieter rpm). I went into the bios and disabled q fan so it would stay in high gear always and it wont even ramp up then. Im letting it cool off as I write this on my old pc Maybe my bios is taking a crap, or the battery is dying. who knows anymore. This officially sucks.

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
bitJericho
22
Years of Service
User Offline
Joined: 9th Oct 2002
Location: United States
Posted: 4th Nov 2007 22:54 Edited at: 4th Nov 2007 22:58
Link
I thought that second issue was the cpu! I didn't say anything cuz I thought that would be *too* coincidental to be the case

It was probably always there, but only was getting hot now because your gfx card isn't killing the computer any more.

Make sure the cpu fan is clean nothing is obstructing it (I assume you did).

If you're sure it's not dust or anything, you may need a new fan. You can get some nice ones on newegg. Your local comp shop (not best buy, but a small comp shop) will have spare stock fans that I'm sure wouldn't cost much.

(I doubt the bios, and the battery wouldnt figure into it)


The greatest multiplayer text adventure ever...
Back to top
Profile PM Email Website
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 4th Nov 2007 23:04 Edited at: 4th Nov 2007 23:05
Link
yeah, but I think I would notice the fan right away if it was malfunctioning prior to today, since its quiet until I run a game then it ramps up and nothing like this has ever happened before. I am hoping its not the mobo itself, and the cpu fan and the graphics crashes are just symptoms. I built this pc in 2005 with good quality parts, like I always have in the past, and I have been lucky to not have any major problems, until now that is

All I can do at this point is see if I can get the cpu fan up to speed while in the bios, if not then I will have no choice to replace that fan, and see. Luckily its dirt cheap (like 10 bucks). Its for a Intel LGA Socket 775 p4 3.2 ghz cpu.

know any really good brands for third party quiet fans ?

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
bitJericho
22
Years of Service
User Offline
Joined: 9th Oct 2002
Location: United States
Posted: 4th Nov 2007 23:26 Edited at: 4th Nov 2007 23:26
Link
At least it's a modern socket, that way if it is a faulty mobo, you can use your fan on the next mobo you get.

I'm really happy with my silenx 80mm fans (non-cpu fans), they're completely quiet and work really well.

I wouldn't hesitate to buy one of their cpu fans:

http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&N=2010110574+50002081&name=SILENX

If you want something cheaper, I'd just check out the reviews for the various fans sold on newegg and go that route.


The greatest multiplayer text adventure ever...
Back to top
Profile PM Email Website
gamebird
17
Years of Service
User Offline
Joined: 13th Jun 2007
Location:
Posted: 5th Nov 2007 00:35
Link
Quote: "Quote: "Unfortunately this will not work because if he is expieriencing the effects of overheating then the components will already be damaged. Heat itself does not cause problems, but the damage it inflicts on the card does. All this will do is keep further damages from occuring."

I beg to differ.

As the components heat (mainly talking silicon chips here), the likely-hood of an electron jumping tracks between silicon sections is greatly increased, which causes the corruption, and eventually completely locking up of the PC.
The hardware usually has built-in protections which will detect the heat up to a certain factory set threshold, then turn it off to stop it being damaged.
"


I stand partly corrected. I forgot about the electrons jumping. However what you said about the hardware is mostly only true with fairly new stuff. The geforce 7 series does have a heat sensor but it does not auto-shutdown. (well most of the geforce 7 series) And usually components do not shut themselves down. It is the motherboard that takes care of this.
Back to top
Profile PM
Jess T
Retired Moderator
21
Years of Service
User Offline
Joined: 20th Sep 2003
Location: Over There... Kablam!
Posted: 5th Nov 2007 01:20 Edited at: 5th Nov 2007 01:25
Link
A new fan might not be necessary.
Since you re-seated the heat-sink, did you notice that the heat-transference gel was a little patchy (or completely rock-hard) on the die?

If it's either, then you need to replace that. I did mine the other day, and instantly the temp dropped 10c.

As Jerico said, local computer shop will have some Arctic Ice, or
the like, for about $5-$10.

Clean off the old stuff with a firm (but bendable) bit of rubber (like an eraser, etc), then rub clean with a tissue.

Don't scratch the heatsink or die, as it reduces the effectiveness of the gel, but then again, scratching it doesn't actually do anything (bad) to it.

Only put on just enough (on the die) to cover it, squish it down, and clamp it without pulling it off again.

Other than that, if it's still failing whilst being cool, I can imagine that the heat has eventually damaged a component (vid-card, more than likely).

If you can, swap it out with an old one / a friends / use the on-board motherboard one (if you have it).

A good test would be to nab yourself an old game that has the option to use a software rendered (HardWar comes to mind), and put it into that mode.
If it plays fine, then the CPU is ok (it's doing the rendering), and the GPU must be at fault (it's got next to no work to do, so it's not failing).

Nintendo DS & Dominos :: DS Dominos
http://jt0.org
Back to top
Profile PM Email Website
bitJericho
22
Years of Service
User Offline
Joined: 9th Oct 2002
Location: United States
Posted: 5th Nov 2007 01:22
Link
Jess, I think he reseated his gfx card heat sink, not the cpu. If the fans not running at full RPM, and you're sure of it, then you'll definitely need a new fan, (or mobo if it's a bad fan controller)


The greatest multiplayer text adventure ever...
Back to top
Profile PM Email Website
Jess T
Retired Moderator
21
Years of Service
User Offline
Joined: 20th Sep 2003
Location: Over There... Kablam!
Posted: 5th Nov 2007 01:27
Link
Either way, a heat-sink is a heat-sink, and needs transference gel to work effectively.

Also, it sounds like you've got a program in Windows that monitors the fan RPM / heat, etc, yeah CR?
Maybe that is the one that is stopping it from spinning up to full RPM as it realises it doesn't need it.

However, that doesn't explain why it's not spinning up on boot when you have it set to go full-pace.

Are you sure that it's simply not quieter now that you removed dust from the fan? It could be spinning even faster, but not making as much noise (ie, running smoother).

Nintendo DS & Dominos :: DS Dominos
http://jt0.org
Back to top
Profile PM Email Website
bitJericho
22
Years of Service
User Offline
Joined: 9th Oct 2002
Location: United States
Posted: 5th Nov 2007 01:30
Link
Quote: "Either way, a heat-sink is a heat-sink, and needs transference gel to work effectively."


I agree with you here, but it wont affect the RPM speed, which is all I'm saying, if it's visually slower, than it's safe to say it's a hardware issue.


The greatest multiplayer text adventure ever...
Back to top
Profile PM Email Website
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 01:55 Edited at: 5th Nov 2007 01:59
Link
thanks for all the info guys, its all sound advice as you guys are saying what I have been reading on a bunch of sites.

the cpu fan and heat sink was remounted (yes I noticed the thermal grease and it probably needs replacing, thanks Jess). When I boot the fan is spinning, but the bios HW monitor is reporting a cpu temp of 60c + and it increases slowly as it runs. If I let it go to windows, by the time I start the asus monitor its already in the 70's c. I am starting to think maybe the mobo is misreporting the data because if the pc has been off for over an hour would it really hit 60c + in the time it takes to enter the bios?. not sure. I think I will just get a better fan to replace this one. and see if it makes a diff before I fry the cpu.

j2d, yeah I saw that on newegg, but I prefer to take advantage of the PWM capability of my mobo (4 pin cpu_fan connector) That fan you mentioned is a 3 pin jammer. I have beeneyeing this one:
http://www.newegg.com/Product/Product.aspx?Item=N82E16835887002
quieter, better cooling, need to regrease anyway, hate this stupid stock fan/hs (loud, hs insane to correctly seat) its an Intel also btw. Only thing about the one I mentioned above is that I will need to remove the mobo from the case to put the bracket on the back, so that kinda sucks. If I can find a duplicate of the intel setup I have, somewhere local, I might try that (since its cheap) just to see if I am barking up the wrong tree. Or a regrease of the contact to see what happens, since that stuff is cheap.

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Jess T
Retired Moderator
21
Years of Service
User Offline
Joined: 20th Sep 2003
Location: Over There... Kablam!
Posted: 5th Nov 2007 06:42
Link
Quote: "but it wont affect the RPM speed"


True

Quote: "would it really hit 60c + in the time it takes to enter the bios?"


Possibly.
I really recommend replacing the heat gel first. Especially, if like I said, it was hard, then when re-seated, the contact is broken and the old heat gel actually acts negatively in transferring to the heatsink.

Quote: "its an Intel also btw"


Urg.
I've only ever had bad experiences with them (replacing them always gives better results). I don't see why they don't just look at what's good about the others around and duplicate even a cheap design. Oh well.

Quote: "If I can find a duplicate of the intel setup I have, somewhere local, I might try that"


That's a good idea, so you can tell if it is exactly that fan, or the style


Good luck with it!

Nintendo DS & Dominos :: DS Dominos
http://jt0.org
Back to top
Profile PM Email Website
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 14:13
Link
thanks.
Ill try and head over to the shop where I buy my parts and see if they have some gel and a duplicate of this fan/hs assembly. I Think that stupid gel is all messed up and the contact is partially broken.

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 15:59
Link
I'm starting to think my cpu/hs was never properly disipating the heat correctly, based on what I am reading about other people with the same processor, and even overclocked ones, they seem to all list cooler idle/load temperatures than mine ever was. They seem to list 30c+ idle and 40c+ loaded, while mine always idled in the 50's c, and went into the 60's loaded. Maybe my system was on edge to begin with (not that I ever had a crash under load) but now things are starting to make sense. Maybe the newer games were enough to put it over the edge temp-wise. It probably wasnt the gpu temp all along, although the cleaning seemed to reduce the gpu temp considerably, which is a good thing either way.

Anyway I found a few resellers in my area that stock Arctic Silver Thermal Paste (ranked #1 in thermal compounds, super thermal transfer, not electrically conductive) so I am definitely going to get a hold of a tube of this stuff and see what happens with this current fan/hs setup. If this stuff brings down the idle temp considerably, and the fan isnt blorkt, then there is no need to replace this fan/hs because it will run very quiet when cpu temps are low. Altho the ZeroTherm BTF90 is cheap and from the tons of reviews I read, it seems to kick serious arse, some even report idle temps below ambient, and lower than board/bus temps lol.
Cant go to shop until later however. I'll keep you all updated with news in this ongoing saga

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 17:21 Edited at: 5th Nov 2007 17:22
Link
Arctic Silver 5, in stock at radio shack a 1/2 mile from my house, woohoo. Cant go until later tho . In the meantime I have re-removed my cpu fan/hs unit and am getting busy with the isopropyl alcohol and coffee filters, removing the old gunk from it, and from the cpu top itself. Its dry and unevenly distributed, which tells me that it was barely ok from the original install, and completely shot after I removed it yesterday and reinstalled it. I had no idea how important this process was. I have been building my own computers since 1990 or so, and prior to this processor it was not such an exacting science. I guess thats due to the amount of heat these newer processors generate (Intel P4 LGA775 3.2ghz). Previously it was slap in the cpu, slap on the fan/hs - done, or the fan/hs was already mounted to the cpu. Now its like micro-science

live and learn

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Arkheii
21
Years of Service
User Offline
Joined: 15th Jun 2003
Location: QC, Philippines
Posted: 5th Nov 2007 17:30
Link
I can't imagine the CPU causing problems only because it's running a more intensive game. If that's the case then anything that can take it to 100% load should kill the system, like file compression. RAMs shouldn't be a problem unless they're unbranded or overclocked beyond its limits.

Back to top
Profile PM
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 17:58
Link
evrything in this system is branded, and I never overclock anything in it. I hear what youre saying, but until I can get the cpu back to where it was at least I wont know. so thats the step I am at. If I apply this silver stuff and notice a diff in temp in bios, then I can at least run windows, and run a monitor there to watch it. Then if all ok I can run a logger on cpu temp as well as the gpu to see where the temps go when it dies (if it continues to die in games).

so basically I am no closer to knowing exactly what the problem is but I will/have been ruling a ton of stuff. I really dread one scenario: I apply the arctic silver and reseat the cpu fan/hs, and in the bios the cpu temp is still immediately in the 60c + range. If that happens then I will have to go to drastic measures, and will be pc-less for a few days in the least.

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Preston C
21
Years of Service
User Offline
Joined: 16th May 2003
Location: Penn State University Park
Posted: 5th Nov 2007 18:23 Edited at: 5th Nov 2007 18:26
Link
CR: If you do find your 6800 Ultra is damaged in some way, if you want, I could probably mail you my old 6800 128MB, that I bought when you helped me build my computer two years ago, until you can find a suitable replacement for your card.

Might have a bit of dust to get out, but the temps on it have never reached the levels you've been describing.



AMD Opteron 185 2.6 Ghz | 2 GB RAM | 8800 GTS 640MB | Vista Home Premium
Back to top
Profile PM Email
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 18:58
Link
Wow, thanks for that offer Preston, I will definitely keep that in mind - lets hope it doesnt come to that

do you happen to know what is a normal full load temp for that gpu? like I said mine was hitting 90c but now its around 60c after a good cleaning. I think the gpu temps can go into the 130c + range before it is approaching "overheat" levels. Anyone have a link to a thermal spec sheet for an nvidia 6800 256 pci-e card, before I go looking?

I just found the spec sheets for my cpu and it seems that its max safe temp is 67.7c @ 84 watts consumption, according to intel. Yesterday it went over 70c in win before I quickly shut it down, but this seems due to the unseat/reseat debacle because that never happened before. I'll be able to go get the gel in a couple of hours and re-test.

thanks again Preston (and everyone) I'll keep this updated.

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 22:11
Link
got the arctic silver thermal grease
gonna go apply and see what happens...

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
NeX the Fairly Fast Ferret
19
Years of Service
User Offline
Joined: 10th Apr 2005
Location: The Fifth Plane of Oblivion
Posted: 5th Nov 2007 22:27
Link
My hardware crashes if it goes over 55 degrees C, something I discovered before installing two extra fans. Thankfully, it's now rarely above 35.


Since the other one was scaring you guys so much...
Back to top
Profile PM Website
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 23:00
Link
Normally my cpu would idle at 56c...

I applied the Arctic Silver 5 Thermal Grease, reseated the HS/FAN assembly and the temp in bios was low (45c, was off all night and day today), I let it boot to windows, started the Asus monitor and the temp by then was 55c, I watched and expected it to either rise or stay the same. I am happy to report it has DROPPED and is now idling at 41c!!!

The manufacturer also says to expect a 2 to 5 c drop over the break-in period as the material sets (3 phase). That should put me at idle temps somewhere in the mid to upper 30's c. Thats phenominal.

Well, it seems that the problem I introduced is gone (crosses fingers), now I can get back to the original issue. Hopefully when I post again I will be posting from my "real" computer.

So far I highly recommend this stuff for anyone with a cpu that requires a thermal grease bond between the HS and the CPU

http://arcticsilver.com

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Preston C
21
Years of Service
User Offline
Joined: 16th May 2003
Location: Penn State University Park
Posted: 5th Nov 2007 23:02
Link
I don't remember the usual full load temp for the 6800, but my 8800 GTS has yet to even go higher than the temp your 6800 Ultra was pulling



AMD Opteron 185 2.6 Ghz | 2 GB RAM | 8800 GTS 640MB | Vista Home Premium
Back to top
Profile PM Email
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 5th Nov 2007 23:25
Link
yeah but the power consumption is lower on the newer cards, so logic of "faster card = more temp" actually doesnt hold true in this case. But get this....

As I posted before I went back into my office to find my pc on but crahed, monitor in standby! I have it set to an hour in windows so it wasnt that it went into standby....

Anyway I quickly rebooted to the bios (pc was on this whole time), in the bios the cpu temp was still a cool 43c, so luckily its not a cpu heat issue. But now it brings back that its most likely a crapping out video card. At this point I dont even care. Its more about knowing what exactly is the issue, instead of all this trial and error, ruling out crap. It was fun for about 5 minutes on saturday lol. Its now end of monday. I guess Ill start researching new video cards which I cant afford atm.

meh again

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
Agent Dink
20
Years of Service
User Offline
Joined: 30th Mar 2004
Location:
Posted: 6th Nov 2007 00:53
Link
My 8800GTS idles in the mid 50s all stock cooled no overclocking. It only gets into the mid to upper 60s while I was playing the Crysis demo with all graphics set to high in XP.

My 6800GT would usually play games in the mid 70s to about 80c.

Warning: Please be advised. Geek Culture is under lockdown. All mods are set to Indi mode. Any and all topics WILL BE LOCKED. Post at your own risk!
Back to top
Profile PM
Jess T
Retired Moderator
21
Years of Service
User Offline
Joined: 20th Sep 2003
Location: Over There... Kablam!
Posted: 6th Nov 2007 01:45
Link
Well, I'm glad the thermal gel helped, but it's a shame that you're still getting issues!

Looks like you might have to swap out that card (test another one in there before making the new purchase! You'll regret it if it's not actually the card!).

Nintendo DS & Dominos :: DS Dominos
http://jt0.org
Back to top
Profile PM Email Website
CattleRustler
Retired Moderator
21
Years of Service
User Offline
Joined: 8th Aug 2003
Location: case modding at overclock.net
Posted: 6th Nov 2007 02:57 Edited at: 6th Nov 2007 03:17
Link
yeah, that Arctic Silver is the sh*t!
too bad that was a separate problem I introduced on top of the video issue, but I am glad I did. I learned alot and my cpu runs 15c cooler now and should drop even more (2-5c) over the 200 hour "burn in" period for the gel.

I am on my machine as we speak, I dug out an old ATI Rage II + DVD pci video card, and slapped it in my machine. I am going to make sure I dont see any odd problems in normal windows usage, then I can say for sure the vid is what went pear shaped, and buy a new card.

Since I dont use Vista is it safe to assume I will need a pre 8800 card. I want to stick with Nvidia. Whats a good dx9/xp card above the one I had (6800)? Hows the 7900 series? and whats the deal with the lettering scheme at the end (gt, gts, xt or w/e) I know it usually means "very shite, somewhat shite, top notch, but I havent researched video cards since I was buying this 6800, any advice?

On a side note, this old card is running 1280 x 1024 32bit, not bad for a temporary old pos, haha


EDIT: I just realized I can still get a hi end 8800 card and use it with XP/dx9, but it will support dx10/vista if and when I choose to move up, is that right??

My DBP plugins page is now hosted [href]here[/href]
Back to top
Profile PM
gamebird
17
Years of Service
User Offline
Joined: 13th Jun 2007
Location:
Posted: 6th Nov 2007 03:28
Link
Something to remember- the heat sensors on processors (and graphics cards) aren't always correct.

Yes, if a card is in the 8xxx series it will support directx 10. It doesn't get locked out or anything. However *looks at vista*, the software is a different story...
Back to top
Profile PM
Back to top

Login to post a reply

Server time is: 2024-11-19 13:21:37
Your offset time is: 2024-11-19 13:21:37