|
Rexxed posted:Going off of this post it doesn't look like using a fan to pull air through it instead of pushing has very much difference in overall cooling:
|
#
?
Jun 11, 2014 07:51
|
|
|
|
| # ? Jun 11, 2024 20:45 |
|
|
Cardboard Box A posted:Thanks, is there any danger of dust being an issue? The case in question has a filtered front fan intake, if that mitigates the issue. Dust build up shouldn't be much different. I try to dust mine out every spring since the ambient temperatures go up in the summer.
|
|
#
?
Jun 11, 2014 08:02
|
|
|
Athlon x4 620 I'm running into some overheating issues (when encoding etc, not normal use).. My PC finally shut off the other day so I ran CPU Temp and it was getting over 85 and the max is meant to be 70? Anyway, the "passive" setting in Windows CPU managment (reduce multiplier before fan) doesn't seem to do anything. I can "fix" the problem by reducing max state to 99% but it's disappointing because it reduces the multiplier from 13x down to 9.5% (last I checked 9.5 wasn't 0.99X13 but oh well) Anyway I just got a jist of an idea, of lowering the core voltage (slowly).. it's at 1.4v at the moment. It'd be nice to not have to drop the multiplier so much just to get it to reasonable temperatures. Is this a valid way to do it? Is there an in-windows way to do this or do I have to keep going to bios?
|
|
#
?
Jun 11, 2014 21:02
|
|
|
echinopsis posted:Athlon x4 620 Are your fans and things clean?
|
|
#
?
Jun 11, 2014 21:03
|
|
|
Sortof.. I don't have a decent way to clean them right now, that would involve spending cash on canned air.. I just dropped voltage by 0.1v and so far it's doing a marvelous job. I'll leave it for 10 more minutes and see what temp it gets to.
|
|
#
?
Jun 11, 2014 21:14
|
|
|
echinopsis posted:I just dropped voltage by 0.1v and so far it's doing a marvelous job. As in 20 degrees centigrade drop.. holy poo poo I have struck gold. Until I BSOD of course 
|
|
#
?
Jun 11, 2014 21:17
|
|
|
echinopsis posted:Sortof.. I don't have a decent way to clean them right now, that would involve spending cash on canned air.. Gotta pay to play.
|
|
#
?
Jun 11, 2014 21:19
|
|
|
echinopsis posted:Athlon x4 620 Have you put it under that workload before and not overheat? It's been a while since I've overclocked AMD but I believe ~70 is as hot as they should get like you said, and 1.4 volts is fine for these, pretty moderate. I'd do what Don Lapre said. I had a stock AMD cooler that would get very dusty very quickly and make my temperatures runaway. Redo thermal paste, consider a cheap cooler if its feasible (assuming stock cooler), etc. In any case, yes lowering vcore is what you need to do. If it gets unstable, you have to lower your multiplier. If you're overclocked on some sort of autotune software you might actually have a lot of voltage you can drop here. But frankly it sounds like a fan/heatsink issue more than anything else
|
|
#
?
Jun 11, 2014 21:20
|
|
|
On a related note I think everyone in this thread should know how to re-oil a fan bearing. Its saved me from having to source replacements for expensive/exotic fans.
|
|
#
?
Jun 11, 2014 21:23
|
|
|
Yeah, so far the vcore solution is working, better than I hoped, but yeah.. I don't remember it getting this hot back in the day when I got it. Gotta get that canned air. smdh at paying for air
|
|
#
?
Jun 11, 2014 21:24
|
|
|
Shaocaholica posted:On a related note I think everyone in this thread should know how to re-oil a fan bearing. Its saved me from having to source replacements for expensive/exotic fans. I don't know how to do this
|
|
#
?
Jun 11, 2014 21:24
|
|
|
echinopsis posted:I don't know how to do this http://www.overclockers.com/em-spinning-lubricate-pc-fans/ Not all fans are designed like this but the bigger ones are. Smaller low power GPU fans are different.
|
|
#
?
Jun 11, 2014 21:26
|
|
|
echinopsis posted:Sortof.. I don't have a decent way to clean them right now, that would involve spending cash on canned air.. The main issue is right between the fan and the heatsink. You can remove the fan and just use a damp rag and wipe the top of the heatsink to get any buildup that is blocking air from flowing down into the fins.
|
|
#
?
Jun 11, 2014 21:56
|
|
|
I didn't know how either and just wondered "do you just squirt oil on it" and apparently it is so
|
|
#
?
Jun 11, 2014 21:58
|
|
|
I clean unused paint brush works really well too to get out stubborn dust and its not a consumable like compressed air so it'll last forever but it still costs money.
|
|
#
?
Jun 11, 2014 21:59
|
|
|
I use an A/C airbed compressor for that 
|
|
#
?
Jun 11, 2014 22:26
|
|
|
Don Lapre posted:The main issue is right between the fan and the heatsink. You can remove the fan and just use a damp rag and wipe the top of the heatsink to get any buildup that is blocking air from flowing down into the fins. Thanks Don! One stand up guy. Ended up forking out for can of duster. $20! gently caress me
|
|
#
?
Jun 11, 2014 22:29
|
|
|
Shaocaholica posted:http://www.overclockers.com/em-spinning-lubricate-pc-fans/ also on page http://www.overclockers.com/add-leds-fans/
|
|
#
?
Jun 11, 2014 22:30
|
|
|
Don Lapre posted:The main issue is right between the fan and the heatsink. You can remove the fan and just use a damp rag and wipe the top of the heatsink to get any buildup that is blocking air from flowing down into the fins. I blew out that poo poo you're talking about and I'm down another ~20 So that plus dropping the vc 0.1v took me down from 85+ on full load to 41.. epic moves dudes
|
|
#
?
Jun 11, 2014 22:36
|
|
|
Good god. Now crank it back up manually until you hit 70 again lol Also $20 for a can of compressed air? pfffff
|
|
#
?
Jun 11, 2014 22:44
|
|
|
It's a big can... Yeah I was pissed off but what else could I do? Ha! Why crank up vcore? I'm not trying to overclock at all, I just thought this was the most appropriate place for talking about underclocking! 
|
|
#
?
Jun 11, 2014 22:46
|
|
|
Oh, because I figured at 1.4 volts. Well, you could, now you know how. If you do encoding it will be a direct benefit ie- turn up your multiplier until its unstable, then turn up vcore until its stable, and back and forth until your thermal limit it reached  enjoy your faster computer
|
|
#
?
Jun 11, 2014 22:53
|
|
|
echinopsis posted:It's a big can... Yeah I was pissed off but what else could I do? Maybe too late, but something like this: http://www.amazon.com/Metro-Vacuum-ED500-500-Watt-Electric/dp/B001J4ZOAW
|
|
#
?
Jun 11, 2014 22:53
|
|
|
teh_Broseph posted:Maybe too late, but something like this: http://www.amazon.com/Metro-Vacuum-ED500-500-Watt-Electric/dp/B001J4ZOAW I can see this can lasting me half a decade. I don't think I need to invest in 15 years worth of duster all at once.. But thanks! I appreciate the post. Also how don't you know I live in NZ? Ignoarints posted:Oh, because I figured at 1.4 volts. Well, you could, now you know how. If you do encoding it will be a direct benefit Hmm perhaps.. I was always unhappy with it being so hot but I'd be content with low 60s.. I didn't know I could increase the multiplier anymore.. Maybe I can ;')
|
|
#
?
Jun 11, 2014 23:07
|
|
|
Radiator fans driving me crazy. I thought one was going bad because it really sounded like poo poo under load, but after dropping $35 on 2 jetblo 120 fans (for this $50 cooler..) they sounded perfectly fine for one IBT run then one fan started sounded like poo poo. It was the same fan location - the one touching the case. I thought perhaps it was because the long screw was going through the radiator that I was hearing the pump noise resonating through the case. When I remove the power connector for just that fan it immediately stops. These new fans are coated in rubber I figured that would at least rule that out, so I was starting to think I had yet another bad fan. After a bunch of swapping them around I determined it was in fact just that spot. So looking closer I saw that the fans had a raised rubber grommet area around the metal sleeve for the screw on one side and it was more or less flat on the other. I turned them around to have the thicker rubber side touching the case and the sound is virtually gone. So the sound was just the fan vibrations resonating through this new case (h440) via the screw, touching the metal sleeve.  Now my fans are drawing air inside the case through the radiator which I don't want. I figured this might actually cool my CPU better, but at the expense of making my whole case hotter. In reality it's a whole 1 degree cooler so well within margin of error, but hitting 77 degrees on average in IBT with a spike to 81 and 82 degrees on two cores for one run meaning I gained 3-4 degrees with just these fans over the stock ones. Pretty rad! I was staring at a kraken x40 when I was looking at fans and just considering buying that instead and saving the old AIO for my GPU. I reasoned that I'd still need a new fan since I thought one was broken so I got the fans instead. Should have just bought the x40... doubt I'm going to return the fans though because they are in fact better. (im just keeping them because theyre red) echinopsis posted:I can see this can lasting me half a decade. I don't think I need to invest in 15 years worth of duster all at once.. But thanks! I appreciate the post. Also how don't you know I live in NZ? There's little to lose honestly if you just keep it moderate.
|
|
#
?
Jun 12, 2014 05:21
|
|
|
Well I can't increase the multiplier anymore so can I increase the 200mhz clock? what kind of increment does one recommend?
|
|
#
?
Jun 12, 2014 05:27
|
|
|
echinopsis posted:Well I can't increase the multiplier anymore so can I increase the 200mhz clock? what kind of increment does one recommend? Ah drat sorry man, forgot its not a black edition. But yes, basically, increase your FSB (200mhz clock) in increments of 10 or 5 until you are unstable. I like to test stability with a variety of tools, but by far the easiest quick way for me is IntelBurnTest. Once it's unstable, increase your vcore voltage in increments of 0.01v. Keep doing this until you hit 60 degrees with IBT. Don't worry, its really difficult to actually hit this temperature with any real workload. Once you are comfortable with some settings, run a longer stress test like Prime95 small fft for as long as you like. Because you are adjusting your FSB value, there can be a lot more to this. You might not be able to stabilize with just vcore alone, even if you have the thermal room to do so and even if your chip is technically capable. At that point you have to increase the CPU north bridge frequency and/or voltage. But this is more complicated obviously and the first thing is much easier. If you do get interested, this is probably relavent
|
|
#
?
Jun 12, 2014 05:39
|
|
|
Hmm I'm not sure if that's for me. 98% of the time I'm just browsing or whatever.. Thanks for the potential advice though.
|
|
#
?
Jun 12, 2014 06:30
|
|
|
Is a 550W XFX PSU (Seasonic design) enough to overclock a 280x? I plan on keeping it at stock voltage but seeing how far I can go with core and memory clocks. It's this model with the Windforce 3X cooler. My current 7850 is stable at 1130/6000, but voltage is locked at 1.225 in GPU Tweak and that card has a much lower TDP.
|
|
#
?
Jun 15, 2014 00:28
|
|
|
MondayHotDog posted:Is a 550W XFX PSU (Seasonic design) enough to overclock a 280x? I plan on keeping it at stock voltage but seeing how far I can go with core and memory clocks. It's this model with the Windforce 3X cooler. My current 7850 is stable at 1130/6000, but voltage is locked at 1.225 in GPU Tweak and that card has a much lower TDP. Yes, and with 100W+ or so to spare.
|
|
#
?
Jun 15, 2014 00:52
|
|
|
Gotta love DIY pc builders That is an h100 installed with no back bracket. Just the 4 posts backwards tightened as hard as possible into the board.
|
|
#
?
Jun 16, 2014 21:37
|
|
|
Don Lapre posted:Gotta love DIY pc builders How hard is it to read a goddamn manual? I mean it's mostly loving pictures anyways, it even has separate sections iIntel vs. AMD so it's less confusing. 
|
|
#
?
Jun 16, 2014 23:26
|
|
|
i remember once as a kid i didn't understand how evenly thermal compound could spread, so I treated the heat spreader on my athlon like it was an acrylic painting
|
|
#
?
Jun 16, 2014 23:44
|
|
|
So this is interesting: http://www.tomshardware.com/news/asus-overclocking-h87-h97-b85,27076.html Apparently Asus decided that everyone should have the right to OC.
|
|
#
?
Jun 17, 2014 22:42
|
|
|
Krailor posted:So this is interesting: Asus getting big enough they are starting to push their weight around.
|
|
#
?
Jun 17, 2014 23:05
|
|
|
It does make a good point about sales. It wouldn't be unthinkable for Intel to collaborate or support this decision for the Pentiums. Hooray for fewer "locked" features
|
|
#
?
Jun 17, 2014 23:11
|
|
|
I did it you guys. I had my PC open today so I figured I'd finally change the paste on my chip. Delidded 4770k overcocked to 4.5ghz and overvolted by .01 or whatever one increment is. With my previous AS5 paste: Idle 50c, Load (Prime95 Blend) 95c, causing the PC to reboot. Everything the same but with a fresh coat of Noctua NT-H1: Idle 42c, load 63c. Yeah. Did I mention this is on air? Also pretty weird, I used a rice grain for all applications of both pastes, but when I removed the chip, all of the AS5 was still liquid after a month use. I noticed when I first put the AS5 on that it has some clear liquid along with the paste, kind of like that bit of water in a ketchup bottle, proportionately. The NT-H1 was better mixed, and I got the new temps instantly. Finally, two R9 280X in crossfire stock clocks on Saints Row IV with uncapped framerate with Prime95 Blend running in the background, pulls 550 watts max from my 620 watt PSU. Whee!
|
|
#
?
Jun 18, 2014 06:16
|
|
|
Two things: - heh, overcocks - how did you measure the 550W? Also Arctic makes lovely TIM, news at 11. Their best/competitive TIM is their two part adhesive epoxy which trades blows with the best of them.
|
|
#
?
Jun 18, 2014 07:53
|
|
|
deimos posted:Two things: I measured the 550w with my trusty Kill A Watt: http://www.amazon.com/gp/aw/d/B000RGF29Q Actually I forgot to subtract 15w for my monitor since I have the whole surge protector routed through it.
|
|
#
?
Jun 18, 2014 13:33
|
|
|
|
| # ? Jun 11, 2024 20:45 |
|
|
Zero VGS posted:I did it you guys. I had my PC open today so I figured I'd finally change the paste on my chip. Original temps sure were terrible. After temps are nearly incredible, but if you only had to bump 0.01 it makes sense. I have a hard time believing that's the only voltage increase you are going to need in the end, but the fact that it even boots with 0.01 over and +1.0 ghz you're already in luck. I'd run IBT to get some hot temps, or just small FFT in P95 for 15 minutes. I'm pretty much a firm believer in quality of paste - I mean it's not exactly a lot of money either - but its also true that it simply isnt important for most people
|
|
#
?
Jun 18, 2014 14:47
|
|