Overclock.net › Forums › Cooling › Water Cooling › Getting shut down at low temps?
New Posts  All Forums:Forum Nav:

Getting shut down at low temps?

post #1 of 13
Thread Starter 
Ok, so this is my first build to include GPU water cooling.

Relevant hardware:

4770k w/ Coollaboratory Liquid Pro and Swiftech Apogee block
EVGA 780 Ti Classified with CL Liquid Pro and EK-FC780 Classy block
Swiftech mc355 pump
5/8x3/8 tubing
2x 240mm Black Ice Stealth Rads w/ 4x 120mm Aeroshark high static pressure fans
EVGA 750W PSU
Bitfenix Prodigy Case

I think that's all that matters but let me know if more info is needed. (I know the case doesn't matter lm just quite proud of my mITX powerhouse lol)

Anyway, just got the rig up and running about 2 days ago, and today went for a BIOS overclock/voltage mod (1.21V) on the GPU. Unfortunately when put under load it shuts down and powers back on. From what I've read this would usually indicate high GPU temps; I only ever get to 45-50C before the power goes. So I've switched back to the ln1 (stock) BIOS. Now, I don't need to have it overclocked but why else would I invest so much into the cooling?.

The CPU is OC'd stable at 4.1 with no voltage mods and I maintain about 70-75C under full load.
On the stock BIOS with the GPU I top at about 55-60C under stress testing. (Checking the VRM in GPU-Z I hit similar temps as well +/- 5C)
When I flash the BIOS (I've tried 2 pre-built, basically a variation on the skyn3t rev.2 as well as the skynet rev.2 itself) I hit about 45-50 before I lose power, this is about 5-10 seconds into stress testing. I even tried using the Kepler BIOS Tweaker to lower the clocks on the skyn3t BIOS, and although I did maintain power maybe a little longer it still crapped out on me pretty quick,

This is my first time overclocking a GPU.

So what the "F" is going on?
post #2 of 13
sounds like a power issue. is your clock speeds to high on the gpu? sounds like you need to play with it more. smile.gif

edit: i wouldn't think that psu would have problems with an OC on the gpu/cpu but you never know, do you have another psu you could try? if not that, try a different smaller oc on the gpu and see if the same problem happens.
Edited by dissonance - 3/9/14 at 6:31pm
post #3 of 13
Thread Starter 
I dunno if it's a PSU issue... It's brand new PSU. Unfortunately the reason I bought this PSU is my 600W doesn't have 2x 8-pin for the card. Actually scratch that, I can yank the PSU from my full tower, it's something like an 850W or 950W.

I don't however think it is the PSU. I've never OC'd a GPU before so I don't know exactly what the issue is, but the PSU is about furthest from my mind. Ot's just pretty disappointing when I trying to apply a BIOS OC that most people are getting stable result on the stock ACX cooler frown.gif

Another piece of the puzzle is that when I did apply the BIOS OC I was able to game on Bioshock Infinite for over an hour, I had alt-tabbed out several times to find where a particular Infusion was (in game) I got a shutdown maybe the 3rd or 4th time alt-tabing out, it shut down while I was watching a youtube vid. Alt-tabbing out of the game back into the OS seems to put the GPU under higher load, which is strange bc on my AMD full tower when I alt-tab it takes most of the stress off. Yeah I'm gonna have to play with it more, just wish it were more friendly like my CPU that handles it's OC like a champ, and I haven't even pushed it yet.
post #4 of 13
I may be wrong but I always thought that a bad overclock on a gpu will just cause a system crash or gpu driver crash but not a shut down.
I think dissonance might be onto something with possible power issue.
Old X58
(17 items)
 
  
CPUMotherboardGraphicsRAM
i7 970 6 core ASUS Sabertooth X58 MSI GTX 1070 Gaming X GSkill 24 gigs 
Hard DriveHard DriveHard DriveCooling
Crucial MX100 256 gig Crucial M500 120gig WD VelociRaptor 1TB Noctua 
OSMonitorMonitorKeyboard
Windows 7 Pro 64bit LG 29inch widescreen DP LG 29inch widescreen DP Corsair K65 
PowerCaseMouseAudio
EVGA SuperNova G2 1000w Corsair C70 Corsair M65 Sound Blaster Z 
Other
Custom sleeved psu cables 
  hide details  
Reply
Old X58
(17 items)
 
  
CPUMotherboardGraphicsRAM
i7 970 6 core ASUS Sabertooth X58 MSI GTX 1070 Gaming X GSkill 24 gigs 
Hard DriveHard DriveHard DriveCooling
Crucial MX100 256 gig Crucial M500 120gig WD VelociRaptor 1TB Noctua 
OSMonitorMonitorKeyboard
Windows 7 Pro 64bit LG 29inch widescreen DP LG 29inch widescreen DP Corsair K65 
PowerCaseMouseAudio
EVGA SuperNova G2 1000w Corsair C70 Corsair M65 Sound Blaster Z 
Other
Custom sleeved psu cables 
  hide details  
Reply
post #5 of 13
Quote:
Originally Posted by jlhawn View Post

I may be wrong but I always thought that a bad overclock on a gpu will just cause a system crash or gpu driver crash but not a shut down.
I think dissonance might be onto something with possible power issue.

+1. My system gives me a blue screen whenever I have a bad OC (CPU or GPU).
Gaming Rig
(25 items)
 
HTPC Rig
(21 items)
 
 
CPUMotherboardGraphicsRAM
Intel 5930k Asus X99 Deluxe ii 1080 GTX FE G. Skill F4-3200C16Q-16GVK 
Hard DriveHard DriveHard DriveHard Drive
Intel 530  Samsung EVO 850 Toshiba DT01ACA300 Toshiba DT01ACA300 
Hard DriveHard DriveHard DriveHard Drive
Toshiba DT01ACA300 Toshiba DT01ACA300 Toshiba DT01ACA300 Toshiba DT01ACA300 
CoolingCoolingCoolingCooling
MCP35x x2 Alphacool 480mm UT60 Alphacool 480mm UT60 Gentle Typhoon AP15 
CoolingOSMonitorMonitor
NB eLoops B13 Windows 10 Pro Hisense 4k 40" TV 24" BenQ ET-0036-EB 
KeyboardPowerCaseMouse
Varmill VA87M EVGA 1300 G2 Caselab STH10 Logitech G9x 
CPUMotherboardGraphicsRAM
Intel G3258 ASRock Z94 Pro 4 Asus 290 G. Skill 8GB 
Hard DriveHard DriveHard DriveHard Drive
Toshiba 5 TB Toshiba 3 TB Samsung spinpoint 2 TB Western Digital Caviar Green 
Hard DriveHard DriveHard DriveHard Drive
Western Digital Caviar Green Western Digital Caviar Green Western Digital Caviar Green Western Digital Black 
OSMonitorPowerCase
Windows 10 Pro Panasonic TH-50PZH80U Corsair HX620 White Enthoo Primo 
AudioAudioAudioAudio
PSB Image 4T Onkyo TX-NR636 PSB Image 4T JBL Loft Series 
Audio
Nuance Sub 
  hide details  
Reply
Gaming Rig
(25 items)
 
HTPC Rig
(21 items)
 
 
CPUMotherboardGraphicsRAM
Intel 5930k Asus X99 Deluxe ii 1080 GTX FE G. Skill F4-3200C16Q-16GVK 
Hard DriveHard DriveHard DriveHard Drive
Intel 530  Samsung EVO 850 Toshiba DT01ACA300 Toshiba DT01ACA300 
Hard DriveHard DriveHard DriveHard Drive
Toshiba DT01ACA300 Toshiba DT01ACA300 Toshiba DT01ACA300 Toshiba DT01ACA300 
CoolingCoolingCoolingCooling
MCP35x x2 Alphacool 480mm UT60 Alphacool 480mm UT60 Gentle Typhoon AP15 
CoolingOSMonitorMonitor
NB eLoops B13 Windows 10 Pro Hisense 4k 40" TV 24" BenQ ET-0036-EB 
KeyboardPowerCaseMouse
Varmill VA87M EVGA 1300 G2 Caselab STH10 Logitech G9x 
CPUMotherboardGraphicsRAM
Intel G3258 ASRock Z94 Pro 4 Asus 290 G. Skill 8GB 
Hard DriveHard DriveHard DriveHard Drive
Toshiba 5 TB Toshiba 3 TB Samsung spinpoint 2 TB Western Digital Caviar Green 
Hard DriveHard DriveHard DriveHard Drive
Western Digital Caviar Green Western Digital Caviar Green Western Digital Caviar Green Western Digital Black 
OSMonitorPowerCase
Windows 10 Pro Panasonic TH-50PZH80U Corsair HX620 White Enthoo Primo 
AudioAudioAudioAudio
PSB Image 4T Onkyo TX-NR636 PSB Image 4T JBL Loft Series 
Audio
Nuance Sub 
  hide details  
Reply
post #6 of 13
I'd be interested to hear what the result is here as I'm suffering a very similar issue with my AMD rig.

FWIW, specs are:

AMD FX6300 CPU
8GB RAM
XFX R9 290
CoolerMaster Thunder 700W

Difference is, i'm checking out some closed loop solutions, with an Antec Kuhler 650 on the CPU and a Corsair H105 paired with an NZXT Kraken G10 adapter on the GPU (and heatsinks on the RAM and VRMs) - running idle it's fine, loading up Metro Last Light, the intro plays, and menu will just about show, before I noted a drop on the 12V rail to 11.8V and the screen goes blank, system beeps once, and that's it- gotta cold boot to get it back. Temps aren't going any higher than 40.C before giving out (and the heatsinks are all cool to the touch too). Overclocked, or not.

I'd be curious to know what the current draw is on the closed loops anyway; the Corsair H105 says 0.34A - but doesn't specify if that's the pump, fans, or all of them... And nothing at all to be found for the Antec Kulher!

--

Edit: Managed to pull some stats from GPU-Z with 0.1s resolution! https://dl.dropboxusercontent.com/u/11053908/GPU-ZSensorLog-10042014.xls
Edited by dalgibbard - 3/10/14 at 10:07am
post #7 of 13
Are you OC right now? What happens when you leave everything to default and no OC or tweaks?

Does it behave the same way?

If it does, do you have a spare PSU or borrow a PSU to troubleshoot?
Gaming Rig
(25 items)
 
HTPC Rig
(21 items)
 
 
CPUMotherboardGraphicsRAM
Intel 5930k Asus X99 Deluxe ii 1080 GTX FE G. Skill F4-3200C16Q-16GVK 
Hard DriveHard DriveHard DriveHard Drive
Intel 530  Samsung EVO 850 Toshiba DT01ACA300 Toshiba DT01ACA300 
Hard DriveHard DriveHard DriveHard Drive
Toshiba DT01ACA300 Toshiba DT01ACA300 Toshiba DT01ACA300 Toshiba DT01ACA300 
CoolingCoolingCoolingCooling
MCP35x x2 Alphacool 480mm UT60 Alphacool 480mm UT60 Gentle Typhoon AP15 
CoolingOSMonitorMonitor
NB eLoops B13 Windows 10 Pro Hisense 4k 40" TV 24" BenQ ET-0036-EB 
KeyboardPowerCaseMouse
Varmill VA87M EVGA 1300 G2 Caselab STH10 Logitech G9x 
CPUMotherboardGraphicsRAM
Intel G3258 ASRock Z94 Pro 4 Asus 290 G. Skill 8GB 
Hard DriveHard DriveHard DriveHard Drive
Toshiba 5 TB Toshiba 3 TB Samsung spinpoint 2 TB Western Digital Caviar Green 
Hard DriveHard DriveHard DriveHard Drive
Western Digital Caviar Green Western Digital Caviar Green Western Digital Caviar Green Western Digital Black 
OSMonitorPowerCase
Windows 10 Pro Panasonic TH-50PZH80U Corsair HX620 White Enthoo Primo 
AudioAudioAudioAudio
PSB Image 4T Onkyo TX-NR636 PSB Image 4T JBL Loft Series 
Audio
Nuance Sub 
  hide details  
Reply
Gaming Rig
(25 items)
 
HTPC Rig
(21 items)
 
 
CPUMotherboardGraphicsRAM
Intel 5930k Asus X99 Deluxe ii 1080 GTX FE G. Skill F4-3200C16Q-16GVK 
Hard DriveHard DriveHard DriveHard Drive
Intel 530  Samsung EVO 850 Toshiba DT01ACA300 Toshiba DT01ACA300 
Hard DriveHard DriveHard DriveHard Drive
Toshiba DT01ACA300 Toshiba DT01ACA300 Toshiba DT01ACA300 Toshiba DT01ACA300 
CoolingCoolingCoolingCooling
MCP35x x2 Alphacool 480mm UT60 Alphacool 480mm UT60 Gentle Typhoon AP15 
CoolingOSMonitorMonitor
NB eLoops B13 Windows 10 Pro Hisense 4k 40" TV 24" BenQ ET-0036-EB 
KeyboardPowerCaseMouse
Varmill VA87M EVGA 1300 G2 Caselab STH10 Logitech G9x 
CPUMotherboardGraphicsRAM
Intel G3258 ASRock Z94 Pro 4 Asus 290 G. Skill 8GB 
Hard DriveHard DriveHard DriveHard Drive
Toshiba 5 TB Toshiba 3 TB Samsung spinpoint 2 TB Western Digital Caviar Green 
Hard DriveHard DriveHard DriveHard Drive
Western Digital Caviar Green Western Digital Caviar Green Western Digital Caviar Green Western Digital Black 
OSMonitorPowerCase
Windows 10 Pro Panasonic TH-50PZH80U Corsair HX620 White Enthoo Primo 
AudioAudioAudioAudio
PSB Image 4T Onkyo TX-NR636 PSB Image 4T JBL Loft Series 
Audio
Nuance Sub 
  hide details  
Reply
post #8 of 13
Quote:
Originally Posted by Ironsmack View Post

Are you OC right now? What happens when you leave everything to default and no OC or tweaks?

Does it behave the same way?

If it does, do you have a spare PSU or borrow a PSU to troubleshoot?

For me, it behaves/behaved that way no matter what. I think I may have made things significantly worse though... the only other PSU I had laying around is a rubbish 500W cheapo job (which would never have been up to the task solely by itself...), so I opted to side-power the graphics card from that, with it running just the 6 pin PCI-E, whilst the main system does the rest (including the 8pin PCI-E)- powered them both on simultaneously and everything... but now I appear to have major graphic corruption, even on the BIOS screen. I'm pretty much ready to throw in the towel; as the next tasks would be:
1) Buy a new PSU, hope it works
2) Buy a new Mobo, hope it works
3) Hit the price of a new graphics card and give up

frown.gif It's seemingly pointless to go through points 1 and 2 with the card in the state it is... I'm hoping I can try running the GFX card on a spare mobo, but it's a fairly massive task as it involves breaking apart my storage server! But I guess, if i can prove the GFX Card to be OK, then a replacement motherboard wouldn't be quite so bad. Oh *sigh*. Sorry about the thread hijack too!
post #9 of 13
Thread Starter 
It was my PSU. I have an 850W in my full tower AMD rig, swapped that out, no shut down. I just purchased a Corsair HX650 and am hoping that'll be enough. Just so everyone doesn't call me an idiot for buying a 100W less than what I currently have, it's bc the mITX case that I'm using has a small bay area for the PSU, as it was the 750W I had was actually sticking out of the back of the case about 13mm. So I purchased a higher-end 650W and if I still get the shut-down issue there's no way I'm going to fit an 850W in there so I'll clock back my CPU and try to find a balance between CPU/GPU clock speeds and power consumption.

@dalgibbard No prob with the hijack. TBH I thought about running just the GFX card from the 850W so I wouldn't have to ruin all my cable management in that tiny case just a day after I finished it. With everything I have stuffed in that case It is a really tight area to work in, a lot tighter than I initially thought. Luckily though out of fear of damaging my precious 780Ti I did just that, ripped everything out and ran the 850W. No more shutdowns. Now my mITX rig is sitting unused since I returned the EVGA PSU and am waiting to receive the HX650. Still got my AMD rig though.

Second, you can just remove your card and run the onboard graphics to see if the problem is you card or something else that's causing the issue. I'd wager it's the card.

BTW for a closed loop I'd HIGHLY recommend the Silverstone Tunda 240mm. I felt the quality of everything on that was better (by a significant margin) than any other closed loop I've used (Corsair H80, H110, and the Thermaltake 240mm). Better temps too. Get that and some Coollaboratory Liquid Pro. Incredible stuff.

Edited for spelling
Edited by cjt0 - 3/17/14 at 8:38am
post #10 of 13
FWIW, the card is knackered frown.gif hoping to get a replacement soon though, will wait and see. For the meantime though I've also bought a replacement PSU which arrived earlier this week- went with the XFX 750w Pro Series Black- it's had stellar write ups, even VS the seasonic on which it's based!

Due to using an NZXT G10 adapter on the GPU, I'm restricted to only using asetek design closed loops- but the H105 seems more than sufficient anyway! Am looking at sorting some additional heatsinks for the VRMs to be safe though. Should have just bitten the bullet and got a full block, custom loop... lessons learnt and all that!
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Water Cooling
Overclock.net › Forums › Cooling › Water Cooling › Getting shut down at low temps?