Overclock.net › Forums › Cooling › Specialized Cooling › Phase Change › The 24/7 Sub Zero Liquid Chillbox Club
New Posts  All Forums:Forum Nav:

The 24/7 Sub Zero Liquid Chillbox Club - Page 69

post #681 of 788
Thread Starter 
Yeah the OC % gain vs power consumed etc, not really something i want to look at lol.

Dice for benching is simpler and colder yes , still i like my gaming sessions.

The TXs have been very successful , especially considering they were not custom PCB like the 780 tics and the 7970 lightnings i had prior. I most likely will keep them in a nvidia rig for a while. 3d vision is the only reason i won't completely move it on. The odd game that works in 3D vision is pure worth it.

Next rig i'm hoping its an AMD one , if Vega is just onpar or slightly under NV greatest at the time i'd like a change. I also think it will be more fun in the OC department as GPU boost 3.0 is really pushing those pascals hard and water cooling + has little affect (so far) as there's too many hard limiters in play (voltage caps etc).
post #682 of 788
Thread Starter 
Well i must admit its taken some time ... anyone would think this is hard and cost heaps to do or something (building a 24/7 subzero liquid chilled rig lol) ... but i'm pleased to announce our 5th club member , please welcome Toolmaker03 biggrin.gif

Now Toolmakers rig meets all the rules for entry , he uses liquid for cooling , the liquid is subzero - about -10c under some serious loading (1x3930k and 2xGTX 580) and the setup can operate 24/7 (its currently in the final stages of the build). So that's all 3 club rules met. Whats really interesting about this rig though is the cooling is Pelter / Tec based !!. At the heart of the system lie 3x410 watt tecs , with 6x360 rads and fans cooling the hotside. I just hope electricity is cheap where he lives , although those bitten by the subzero cooling bug could care less about the power sacrifice to the overclocking gods we all make , sacrificed plenty of watts myself in the search for mhz wink.gif

Instead of a fan chilled chillbox , we have a vacuumed chamber , the next stage is to fill with nitrogen to relieve the need to have the chamber under vacuum. This is one really interesting build thumb.gif

A serious amount of work has gone into this so i encourage you guys to check the thread out .. i'll just leave some nuggets here biggrin.gif

Having some issues modifying the 1st post so until i get that sorted here is the build link :
http://www.overclock.net/t/1584867/tec-vacuum-chamber-build-log

Schematic of system :



Implementation :



Peltier HX.


Edited by Orthello - 7/10/16 at 10:22pm
post #683 of 788
well, it took me three tries, and about two months, to actually get my box air tight. you are very correct in the fact that it is the cheapest section, of these builds to create. thumb.gif
lets see here, I had the box, PC case, radiator, fans, leak detector, temperature/humidity gauge, wiring extensions, tubing, sealants, liquid rubber, liquid plastic, and extra insulation to go around the exterior of the box. so yea, even with all of that, it was still the cheapest section of the build.
post #684 of 788
Quote:
Originally Posted by toolmaker03 View Post

well, it took me three tries, and about two months, to actually get my box air tight. you are very correct in the fact that it is the cheapest section, of these builds to create. thumb.gif
lets see here, I had the box, PC case, radiator, fans, leak detector, temperature/humidity gauge, wiring extensions, tubing, sealants, liquid rubber, liquid plastic, and extra insulation to go around the exterior of the box. so yea, even with all of that, it was still the cheapest section of the build.

Could start with one of these great base to start with especially the double glazed glass door.
I figured that there is no real reason not to also mount the PSU and SSD inside the chamber to reduce the amount of wiring required to be sealed.
CREATOR: gd-jpeg v1.0 (using IJG JPEG v62), default quality

I might add that the existing refrigeration gear would have to be removed.
I had a 1HP condensing unit ready for this but in the end I decided it was going to be too bulky and expensive for the actual gains netted.
Edited by feznz - 7/16/16 at 4:25pm
kamikaze
(17 items)
 
  
CPUMotherboardGraphicsGraphics
3770k @ 4.8 asus maximus v extreme Asus Strix 1080 Ti  Asus GTX 770 DCUii 
RAMHard DriveCoolingOS
GSkill 2400Mhz Sandisk ultra pro xspc 240 raystorm Windows 8.1 Pro 
MonitorMonitorMonitorKeyboard
LG UC9938-w AOC 24"  AOC 24" G19 
PowerCaseMouseAudio
enermax maxrevo 1500w nzxt 820 Phantom saitik rat 9 Asus Xonar DX 
  hide details  
Reply
kamikaze
(17 items)
 
  
CPUMotherboardGraphicsGraphics
3770k @ 4.8 asus maximus v extreme Asus Strix 1080 Ti  Asus GTX 770 DCUii 
RAMHard DriveCoolingOS
GSkill 2400Mhz Sandisk ultra pro xspc 240 raystorm Windows 8.1 Pro 
MonitorMonitorMonitorKeyboard
LG UC9938-w AOC 24"  AOC 24" G19 
PowerCaseMouseAudio
enermax maxrevo 1500w nzxt 820 Phantom saitik rat 9 Asus Xonar DX 
  hide details  
Reply
post #685 of 788
well I would not put the power supply inside the chill box, as it produces heat, and the idea of a chill box is to keep the inside of the box cool. the more components placed in the chill box that has to be cooled, the harder it will be for the chiller system to do that. in a worse case scenario the amount of heat being produced is more than the amount of coldness being produced, with no way for the heat to escape from the chill box, the out come would be a hot box instead.
post #686 of 788
Thread Starter 
Quote:
Originally Posted by toolmaker03 View Post

well I would not put the power supply inside the chill box, as it produces heat, and the idea of a chill box is to keep the inside of the box cool. the more components placed in the chill box that has to be cooled, the harder it will be for the chiller system to do that. in a worse case scenario the amount of heat being produced is more than the amount of coldness being produced, with no way for the heat to escape from the chill box, the out come would be a hot box instead.

A 90% efficient Power supplies is going to dump at least 10% of its total power consumption into the chillbox , if located there. You would only consider this really if you had bad cable length issues or could not locate the PSU close enough to the chillbox. Lengthening cables reduces voltage on those cables due to increased resistance , unless you run heavier gauge. Beyond 90cm its starting to become too far for the cable runs. I extended my AX1200i cables up to 90cm by going up a gauge and making sure they were CU cables.
Edited by Orthello - 7/17/16 at 1:04pm
post #687 of 788
Quote:
Originally Posted by Orthello View Post

A 90% efficient Power supplies is going to dump at least 10% of its total power consumption into the chillbox , if located there. You would only consider this really if you had bad cable length issues or could not locate the PSU close enough to the chillbox. Lengthening cables reduces voltage on those cables , unless you run heavier gauge. Beyond 90cm its starting to become too far for the cable runs. I extended my AX1200i cables up to 90cm by going up a gauge and making sure they were CU cables.

that is also what I did, standard power supply cables are 18AWG wire, all of my wire extensions are 14AWG wire. the standard power supply wire are 18" in length, I added a 12" section of extensions for going through the wall of the chill box, they are 14AWG wire, and I added 16" sections that goes from the exterior of the chill box to the power supply. these wire extensions are also 14AWG wire. the 14AWG wire extensions did cost me a little more, but they where worth it for this project.

Edit on the lengths
Edited by toolmaker03 - 7/20/16 at 10:05am
post #688 of 788
but the MB does not need to be TB, you just use a TB "hub"; plug as much as possible from the MB into the hub, run a single TB cable out of the box to another hub sitting on users desk.

Linus did something similar when he did rack mounted PCs for himself & his wife at home.

Of course, these "multifunctional" TB hubs that provide mouse, keyb, usb & graphics ports are not cheap frown.gif
post #689 of 788
about multi GPU.. besides mining... the future is unclear. It seems that SLI as we know it is possibly dead and even then, for +4K & 8K you need lotsa G-RAM, which you dont get from SLI anyways. The future is "piggyback boards". Pascal already is ready for that, but not implemented as such yet.
Want to double your GPU? Just plug an extra module on top of your existing one smile.gif But ideally it needs propriatary sockets on a MB... it all depends if & when AMD & Intel are going to play ball.
Just stack module upon module onto the same "base" pcb. A bit like Arduino smile.gif

Either way, as it stands now, a mixture of connectivity is inevitable. Some TB, some USB, etc.
However, from a business point of view, segmentation is key. A base model with only the basic connectivity with an USB or two; a midrange model which comes with a TB hub; a superdeluxe which comes with every (legacy) connector known to man - including firewire & centronics smile.gif

Something with centronics ports can also be targeted at "hot & dusty" industry environments... which is a segment currently ruled by Siemens, IBM & Lockheed. It you could beat their pricetag, you'ld have to scramble to get a production line going smile.gif
post #690 of 788
Quote:
Originally Posted by Orthello View Post

A 90% efficient Power supplies is going to dump at least 10% of its total power consumption into the chillbox , if located there. You would only consider this really if you had bad cable length issues or could not locate the PSU close enough to the chillbox. Lengthening cables reduces voltage on those cables due to increased resistance , unless you run heavier gauge. Beyond 90cm its starting to become too far for the cable runs. I extended my AX1200i cables up to 90cm by going up a gauge and making sure they were CU cables.

Probably a fair call for 9xx they were still a bit hungry for power, now with the 10xx series according to this and maxing out the OC
http://www.coolermaster.com/power-supply-calculator/
probably talking an extra 80W or 90W of heat, I think is a minimal amount to counteract.
kamikaze
(17 items)
 
  
CPUMotherboardGraphicsGraphics
3770k @ 4.8 asus maximus v extreme Asus Strix 1080 Ti  Asus GTX 770 DCUii 
RAMHard DriveCoolingOS
GSkill 2400Mhz Sandisk ultra pro xspc 240 raystorm Windows 8.1 Pro 
MonitorMonitorMonitorKeyboard
LG UC9938-w AOC 24"  AOC 24" G19 
PowerCaseMouseAudio
enermax maxrevo 1500w nzxt 820 Phantom saitik rat 9 Asus Xonar DX 
  hide details  
Reply
kamikaze
(17 items)
 
  
CPUMotherboardGraphicsGraphics
3770k @ 4.8 asus maximus v extreme Asus Strix 1080 Ti  Asus GTX 770 DCUii 
RAMHard DriveCoolingOS
GSkill 2400Mhz Sandisk ultra pro xspc 240 raystorm Windows 8.1 Pro 
MonitorMonitorMonitorKeyboard
LG UC9938-w AOC 24"  AOC 24" G19 
PowerCaseMouseAudio
enermax maxrevo 1500w nzxt 820 Phantom saitik rat 9 Asus Xonar DX 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Phase Change
Overclock.net › Forums › Cooling › Specialized Cooling › Phase Change › The 24/7 Sub Zero Liquid Chillbox Club