• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

BFG what are you thinking!!!

v_lestat

Gawd
Joined
Sep 13, 2004
Messages
718
ya know ive seen some pretty scary thermal paste jobs, but i got sick of my NEW 2 day old 6800 GT OC run 60+ idle and 70+ load. and i cant overclock it much at all.

so i was gunna install a probe and some AS 5.
well take a look here. this is what i found when i took off the heatsink.
holy crap. everyone with a bfg product take the damn hsf off and reapply thermal paste properly!!

041011120050.JPG


041011120118.JPG


041011120130.JPG
 
you guys recomend any liquid i can use to GENTLY whipe this crap off. ? i got most of it off. but ther eis still some on the cpu area i would like to get off.
i cant believe that the manufacturing plant does this. its rediculous.
 
I believe high purity isopropyl alcohol would do the trick. 90%+, which you can find in a pharmacy.
 
I don't get it. What is wrong with the way they applied the paste? Looks like full coverage of the GPU die to me. What am I missing?
 
wow metal i hope that was a joke.

i forgot to take pics after i put on as 5 but believe me i used about as but as you see in the picture thats NOT on the gpu die. scoop all that out, and then spread it around the ram AND gpu thats abut all i ever use on my CPU and video card stuff.

maybe im taking "a thin layer" a little too far, but they took "covering the gpu die are" WAY too far.
 
haha, holy sh*t that is some overdone thermal compound right there...lol

that is one honkin heatsink too (don't mind the alliteration)

well, post with the new temps, wanna see how a "conservative applier" does with the temps...:p
 
wow. i don't know what happened to yours; my bfg 6800gt's thermal paste was applied fine/with the correct amount lol
 
Met-AL said:
I don't get it. What is wrong with the way they applied the paste? Looks like full coverage of the GPU die to me. What am I missing?

with the pics, it's hard to see but I recognize it from my BFG 6800 GT OC...I was shocked when I removed the stock heatsink to install my NV Silencer Rev. 5. I think you may have a bit more TIM than I did. The TIM is hardly there, and the coverage is far from adequate on the ram. The GPU has visible (although not in his pic) sections where there were varying thicknesses of TIM implying air pockets (TIM is somewhat hard, not an issue of pulling off when removing the heatsink).
 
These things are mass produced, its not like they apply thermal paste one at a time and very carefully.
 
Damn ! eheh
Please let us know your new temp with AS5 applied correctly!!
 
No... On fridays you dotn get thermal paste!!! :p

Seriously though... The card worked right? Thats all BFG cared about...
 
temps are a constant 59c idle and seem to peak around 75 during 3dmark 05.
overclocking went up quite a bit on the memory but nothing on gpu.
427/1100
i been searchin for the "right" bios to use.
im not happy with the bios thats on it.
i took it and upped the stock clocks to 425/1025 and voltages to 1.4v

im wondering though,, dont the Ultra cards use 1.5v ? the editing program allows you to change it to 1.5v if its safe to go to 1.5 screw it ill do it, but unfortunately with that comes higher temps.
the NV 5 is crap frankly. your lucky if you get 5 deg cooler. which dont help at all.
this is the BFG new Silver dual fan and heatsink. and i have it plugged into my fan controller so the fans are on all the time. 100%.

anyone know about the voltages of the ultras and if i should be good to go to flash it with a 1.5v bios ?
im currently looking thru the fs/ft threads looking for a good gpu block for this but no luck.
 
deathBOB said:
No... On fridays you dotn get thermal paste!!! :p

Seriously though... The card worked right? Thats all BFG cared about...

For the average user, the amount of TIM was probably dead on. They won't be OC'ing, and they certainly won't be taking the heatsink off. Thermal throttling isn't set to kick in until 130C, so BFG still has plenty of room, even with inferior paste application. To top it all off, they could slap peanut butter in there as TIM and not hurt the end user too much: they offer a lifetime warranty for replacement if it fails.
 
v_lestat said:
temps are a constant 59c idle and seem to peak around 75 during 3dmark 05.
overclocking went up quite a bit on the memory but nothing on gpu.
427/1100
i been searchin for the "right" bios to use.
im not happy with the bios thats on it.
i took it and upped the stock clocks to 425/1025 and voltages to 1.4v

im wondering though,, dont the Ultra cards use 1.5v ? the editing program allows you to change it to 1.5v if its safe to go to 1.5 screw it ill do it, but unfortunately with that comes higher temps.
the NV 5 is crap frankly. your lucky if you get 5 deg cooler. which dont help at all.
this is the BFG new Silver dual fan and heatsink. and i have it plugged into my fan controller so the fans are on all the time. 100%.

anyone know about the voltages of the ultras and if i should be good to go to flash it with a 1.5v bios ?
im currently looking thru the fs/ft threads looking for a good gpu block for this but no luck.

Dude, the GT runs at a stock voltage of 1.3v.....the Ultra runs at 1.4v.....and the Ultra Extreme is sposed to be nothing more then an Ultra running at 1.5v. You're already at Ultra voltage....and if you're seeing those kind of temps already i would NOT recommend flashing your card to 1.5v. Load temps take quite a dramatic jump by about 10C when gonig from 1.4v or 1.5v.....the "move" should be saved for something like an NVSilencer, preferably with mods.....or some decent water cooling.
 
v_lestat said:
the NV 5 is crap frankly. your lucky if you get 5 deg cooler. which dont help at all.
this is the BFG new Silver dual fan and heatsink. and i have it plugged into my fan controller so the fans are on all the time. 100%.

I purchased my card from Best Buy, so it came with the copper-base heatsink you're speaking so fondly of (the one pictured above, with the two blue LED fans, correct?) It's one of the best looking heatsinks I've seen, but it didn't really work that well in my case.

regarding the NV 5, it depends on your setup. I've got very good airflow running across the NV 5 (and all my PCI slots), which in turn blows the air out of the case rather than recirculating it after it is warmed. I saw my temps go from high 60's C idle and 83C load to 54C and 68C load (after the AS5 cured a bit for about a week - immediate temp drop after application was to 73C during load). 15C drop in temps is good in my book.
 
GT is 1.3v
Ultra is 1.4v
Ultra Extreme 1.5v

With 1.5v I can now get 450/1290. Not sure if many people have used 1.5v on a GT.
 
the NV 5 is crap frankly. your lucky if you get 5 deg cooler. which dont help at all.

Then explain how I managed to drop my load temps by 10C....and why I've been seeing a good number of ppl around a few different forums installing them on their X800's and dropping their load temps by over 15C. If someone has a 6800Ultra they're really not going to gain a lot in the OC department with the NV5....just a quieter running fan. The GT's reference cooler however leaves a LOT to be desired, which is where the NV5 is an excellent choice. It's funny, it seems that a lot of ppl with GT's that're bashing the NV5 have cards that are NOT running on a reference cooler, like the new BFG's, Leadtek, or Gainward. I've heard mixed reactions about the newer BFG cooler though, but you're out of your mind to say that the NV5 is crap and/or not worth the cash if you've got a GT.

With 1.5v I can now get 450/1290. Not sure if many people have used 1.5v on a GT.

I still want to give it a go though. What kind of cooling are you running on your Ultra, and what were your temps and overclocks you were running at 1.4v as compared to now? I'm wanting to take a step up to 1.5v to see if I can push a little higher on the OC, because 450MHz for me is fine, yet 460 doesn't really artifact....it just hardlocks....and my temps were only in the high 60's under load at those clocks.
 
Im using BFG's massive stock cooler, I really doubt that the NV5 would make a difference. I'll probably go watercooling if I change anything.

The idle temps stayed the same about 56C. Full load went from 76C to 81C. Increase of 5C.

I run my computer with the case open though. Until I get a decent case.

I've been thinking of getting watercooling and upping the voltage to 1.6v I wonder if anyone has tried that :D
 
hey btf email the bios you use man. be sure to rar or zip it or it might reject it as a virus.
lestat@mchsi.com
screw it ill give it a try.
i can mod mine but it seems that anything higher than what i have just locks or does the 3 second stall.
not sure with nvidia but with my ati cards ive seen different bios's definately make a difference in overclocking.

but redoing the thermal paste defiantely helped.
doom 3 1600x1200 high quality played thru 1 level and ran time demo 5 times.
temps were 68 deg(i have the nvidia control panel running open in the background and i very quickly quit the game and exit to the desk top where i immediately see the temps.. now thats a helluva difference from the 75+ i was getting.
was going to install a probe but the little electrical thingamajiggers around the gpu die kind of get in the way and i dont wanna have them pink the probe wire in any kind of a way. so i have to rely on the cards built in probe.

anyone seen a taskbar tool that displays the temps in your task bar for this card.
if not someone make one lol ;)

and the ati cooler is different and KNOWN to be better than the nvidia one, infact look around at websites. many webistes reported a temperature INCREASE when using that thing so,, yeah they arent very good in my eyes.

and loud ya know it amazes me how people call video cards loud.

i run with my case lid open. and my stuff is loud, butif i turn all my fans down to min. i can BARELY hear the nvidia fan,, and thats full on 100%. and this again is the big chrome, dual fan copper heatsink from BFG not the old black one.
so loud ? sheesh id hate to have you around when i have my stereo on, you'd go deaf.
 
v_lestat said:
but redoing the thermal paste defiantely helped.
doom 3 1600x1200 high quality played thru 1 level and ran time demo 5 times.
temps were 68 deg(i have the nvidia control panel running open in the background and i very quickly quit the game and exit to the desk top where i immediately see the temps.. now thats a helluva difference from the 75+ i was getting.
was going to install a probe but the little electrical thingamajiggers around the gpu die kind of get in the way and i dont wanna have them pink the probe wire in any kind of a way. so i have to rely on the cards built in probe.

in the one second it takes you to go from doom 3 to your desktop, your temp can drop nearly 10C. if you want a more reliable reading of your temp, run the rthdribl demo and have it opened up to nearly full screen, with just enough of the window not maximized to see the temp in the nvidia control panel. close the demo and watch how your temp drops nearly (in my case 10C exactly) in about a second.
 
doom 3 1600x1200 high quality played thru 1 level and ran time demo 5 times.
temps were 68 deg(i have the nvidia control panel running open in the background and i very quickly quit the game and exit to the desk top where i immediately see the temps.. now thats a helluva difference from the 75+ i was getting.

Still not entirely accurate bro....why are ppl still not doing this right? Run rthdribl, turn off multisampling....and stretch the window to fill almost the entire screen. Open up rivatuner or your Forceware control panel and watch the temps.....let it run for a half hour.....THAT is your load temp. Even during the split second transitions between scenes in rthdribl my core temp drops 2-3C...and then shoots back up, and closing a game takes about the same amount of time.....and the time it takes to get to the menu and exit the game IS going to lower your temps compared to what you'd see for results using rthdribl.

I realise that the coolers are bulit differently, but with the Silencer(s) installed PROPERLY I have never seen someone's temps go UP from using the stock cooler. You do have to use quite a bit more paste with the Silencer to get it to make proper contact with the core/ram.....but once you know what you're doing with the installation you will NOT see temp increases....whether you were using the stock X800 or stock 6800GT cooler.

It's such a shame that so many ppl are misinformed about the NV and ATI Silencers, as they ARE an excellent buy, and the ridiculously simply way to drop your load temps and usually get a decent OC boost out of your card. There are still quite a few people complaining about higher temps.....but the thing is that there are also quite a few ppl sticking to the old paper thin layer of arctic silver rule with these things....which doesn't work in this case.

I'll say it again, the Silencer if installed properly WILL lower your temps if you are running a reference GT or reference X800 cooler....by how much is up to your case temps and specific setup.....but the cooler is inferior in no way to either reference cooler, other then the fact that it's a bit tricky to get it installed properly.

close the demo and watch how your temp drops nearly (in my case 10C exactly) in about a second.

With me/my card it'll instantly drop by only 4C....but that is something that can/would be effected by your case temps and cooling setups. Whether someones temps are going to take a nosedive like that immediately after closing the demo is irrelevant....what does matter though is the fact that the temp WILL instantly jump down at least a few degrees....and any reading you get after closing a game is innacurate.
 
I got the bios from a thread on this forum. It's a BFG ultra bios moded to default to 450/1200 2d/3d and add .1v on ram/gpu.

I wouldn't use this bios unless your sure you can do 450/1200 stable or else you'll need a vga card to get you out of that problem :eek:

I would suggest using EVGA Extreme Bios it defaults to 400/1100 2d and 450/1200 3d I believe.

This is the best website for bios's:

http://www.mvktech.net

Go into the download section and you'll find every bios ever made for 6800's


Edit: I don't know if you can flash the GT to an ultra bios. I think you would be better off finding a modded 6800GT bios that adds .1v then find your most stable overclock.
 
btf said:
Edit: I don't know if you can flash the GT to an ultra bios. I think you would be better off finding a modded 6800GT bios that adds .1v then find your most stable overclock.

You can, but it is easier to just mod your own BIOS and simply bump up the voltage.
 
Looking at the nvidia control panel for temps is useless. In my case it will drop 10-15C before I check it.

Best way is to run riva tuner in the background while playing a game or running 3dmark then check the peaks.
 
btf said:
Best way is to run riva tuner in the background while playing a game or running 3dmark then check the peaks.

True, that'll also work.....at least that way you can see what it WAS at on the graph before it takes the nosedive.

Oh BTW, I've also got my own thread up having to do with the BIOS flash when I originally did it to my Inno3D, a bit more straight forward IMO then a couple other threads I've seen.
 
hey i dont doubt for one minute the ability of the nv silencer.
ive had one on both my 9700 pro and my 9800 xt.
but they are KNOWN to cool better on ati cards then nvidia cards.
rthdribl isnt the endall and again is synthetic.
just like my temps with 3dmark and doom3 ,, which do i go with ?? the doom 3 temps,, why ? cuz its a real game not a synthetic benchmark beating the holy crap out of the card.

ive like the silencers in the past but like i said read around before spouting and kind of belittling people. which i know wasnt meant to bash people ,, so.. dont take what i said as a bash.
but go read ,, read and read. site after site ,, forum after forum are reporting a truly mixed bag with the nv 5. and too many are reporting a whimpy 5 deg., and there are those who are reporting temp increases or no change at all.
i for one believe if they would make a model with a faster fan it would cool better.

or if i could find that same fan somewhere and it was faster i would get one.
but put your hand beind the pci slot. if its anything like my vga silencer Rev 3 you wont feel hardly any air at all, and thas with fan on high. it just flat out isnt moving enough air.
and the heatsink, although it is copper, isnt large enough to handle the heat of all the ram sinks and the gpu.

personally if i can hold the temps the way i am,, then why go spend 30 bucks for the same results.
 
v_lestat said:
rthdribl isnt the endall and again is synthetic.
just like my temps with 3dmark and doom3 ,, which do i go with ?? the doom 3 temps,, why ? cuz its a real game not a synthetic benchmark beating the holy crap out of the card.

The thing is that some games will effect your card/temps/overclocks in different ways, and an overclock or temperatures that are "tuned" for a certain game might not prove very usefull in the future. "beating the holy crap" out of your card with rthdribl is still the best surefire way to find your max load temps and/or overclocks. It doesn't matter in this case that rthdribl is asynthetic test and not a real game, because it still does its job very well.....raising the cards temps.....we're not getting a 3DMark score or anything here, we're talking about temperatures, and if it raises your temps it raises your temps.....and shouldnt matter if its synthetic.

I can understand people finding their max overclocks that are stable/artifact free in the games that they play, because it kind of makes sense.....but if you're talking about finding your load temps.....using a game to find the temperatures rather then something that REALLY brings out the true capabilities of your cooler is kind of silly to me. Think about it....why not use rthdribl? If you find your max oc and/or temps in erthdribl you've got a hell of a lot more re-assurance that things will hold up in your games. I'd personally rather have my card "tuned" to a higher temperature using rthdribl then I would using the games I play.....as I hate running into quirky little things playing my games due to not thouroghly testing my overclocks and/or temps.

I'm fine doing things my way and it sounds like you've got your own methods. I'm not going to try and convince you anymore then I already have, just trying to let you know the pro's and cons of using games as opposed to rthdribl for getting your temps recorded.
 
Are you going to flash it with 1.4v or 1.5v? I'd be curious to see how far a GT could go with 1.5v.
 
heheh wow what a card cooker damn,

lol i used to mess with rthdribl but never had a temp probe or card with temp readings except my 9800 xt and holy cow.. lol wow

83 c peak. and yeah your right,, i didnt believe you but it drops an easy 10 deg the minute you close the program.
wow.
 
ive read that it can handle the 1.5v cuz come card makers are using the 1.5v on their ultras. i gotta do some more readiong before i go up ,, its at 1.4 now so...
 
v_lestat said:
ive read that it can handle the 1.5v cuz come card makers are using the 1.5v on their ultras. i gotta do some more readiong before i go up ,, its at 1.4 now so...

Same here, I'll probly end up waiting 'till I get my water cooling on the GT, but i will eventually be wanting to move to 1.5v.....just wanna/gotta do some more looking into the results ppl are seeing with 1.5v GT's first.
 
Hey I'm waiting for the same thing except with 1.6v :) I don't want to be the first to do it and fry my card.
 
i just dont know how one is expected to cool the ram properly unless you spend 150 bucks on the danger den NV63 or whatever its called.
ya know i would do it in a heart beat but gimmie a break 125-150 bucks for a bloody water block.,, no flippin way.
 
v_lestat said:
i just dont know how one is expected to cool the ram properly unless you spend 150 bucks on the danger den NV63 or whatever its called.
ya know i would do it in a heart beat but gimmie a break 125-150 bucks for a bloody water block.,, no flippin way.

Hehe, that's $125 american....I live in Canada.....try $157 for me......where my TDX for my P4 cost me like $70.

A much more cost effective solution is to pick up the Swiftech block /w adapter for the 6800.....ppl are getting lower core temps with it as opposed to the NV68, simply because the swifty does not cool the memory. Next, pick up some Monster BGA Ramsinks from FrozenCPU to take care of the memory. Combined with a side panel intake fan and/or a Vantec Fancard you're going to have both the best "plain jane" storebought h2O setup for your 6800, as well as some VERY capable memory cooling. It's what I plan on doing, those ramsinks are just too tough to turn down. :p
 
I guess mine was the opposite (still a shitty job on the paste) this is from another post I had @ hardnsoft.net of when I added my NVSIlencer 5

bfg3.jpg


Wow, im really surprized at the CRAP BFG uses to put their cards together, there were some white "spots" of sillicon, look for your self:

what your card might looks like too (if it was me I'd reapply thermal paster regardless of keeping stock HSF or not)

goo.jpg


stock cooler + goo

stockgoo.jpg


BFG 6800GT core:

corecloseup.jpg


here are the stock cooler vs the NV5 side by side:

comparationcoolers.jpg


and here is everything to give you an idea of size:

comparationgeneral.jpg


finally, how the beast looks: (watch out for clearance, I had to move on of my HD's lower cause the cooler woun't fit, its massive)

victory.jpg


hope that gives you guys a good idea how this thing works :)

--NEO

after NV5 my card runs @ 49 idle / 70 Load (load temps taken after 30 min run of rhtbdl @ 1280x1024 while its still running & clocks of 400/1100)
 
Back
Top