EXTREME Overclocking Forums
Home | Reviews | Forums | Downloads | RealTime Pricing Engine | Folding Stats Contact Us


Go Back   EXTREME Overclocking Forums > General CPU Cooling > Phase Change, Dry Ice, LN2, Other Crazy Stuff

Welcome Guest Visitor! Please Register, It's Free and Fun To Participate!
The EXTREME Overclocking Forums are a place for people to learn how to overclock and tweak their PC's components like the CPU, memory (RAM), or video card in order to gain the maximum performance out of their system. There are lots of discussions about new processors, graphics cards, cooling products, power supplies, cases, and so much more!

You are currently viewing our boards as a "guest" which gives you limited access to view most discussions. You need to register before you can post: click the register link to proceed. Before you register, please read the forum rules. By joining our free community you will have access to post topics, communicate privately with other members (PM), respond to polls, upload your own pictures, and access many other special features. Registration is fast, simple, and absolutely free! To start viewing messages, select the forum that you want to visit from the selection below.

After you have registered and read the forum rules, you can check out the FAQ for more information on using the forum. We hope you enjoy your stay here!

Note To Spammers: We do not allow unsolicited advertising! Spam is usually reported & deleted within minutes of it being posted, so don't waste your time (or ours)!


Please Register to Post a Reply
 
Thread Tools
Old 07-23-2012, 11:26 PM   #1
Names_Lucky
Running System Stock
Regular Member
 
Posts: 32
Last Seen: 10-10-2016
iTrader: 0 / 0%
Heatpipes=Bad?

I have a (long sorry) question. I will post a build log when all done, but I'm in process of finishing my submerged oil AC cooled computer. I can reach sub-zero temps (-10C so far, it has 50l of oil, so it's still coming down after 3 days of being on 24/7). At stock speeds/voltages, when I stress test the GPUs (2x 670s) with furmark I get an increase of about 20-30 degrees. However, when I stress test my CPU (also stock) with prime I get a very rapid climb from -10C to 70C. Before submerging both GPU and CPU were stress tested at stock speeds/voltages, and the CPU hovered at around 50C during a prime test. This makes me believe that it is not a seating problem.

My CPU heatsink (Hyper 212+) is a direct contact heatpipe design, while my GPUs are not. The GPUs have heatpipes but are not direct contact. My thought is that the environment (-15C oil), and the temperature of the surrounding heatsink, means that the heatpipes will not function properly, making them not much better than air filled pipes. I would like to hear your guys' thoughts on this, especially if anyone else has witnessed this in the subzero cooling area. Also, you may be wondering about fans or oil movement having to do with it, but the GPUs have no oil movement over them (Fan Tach reads zero), while the CPU may have some movement over them of colder oil from the AC (158cfm delta fans spin slowly in -10C synthetic oil).

If I do decide to pursue a non-heatpipe heatsink, does anyone know of a good one for an i5-2500k? I would like it to have more surface area than the Zalmans that are out there. I actually have an old q6600 zalman solid copper heatsink that I may have to mod for the 1155 socket. Does anyone know a good way to do this? Thanks for any help.
-Names_Lucky
United States  Offline
    Register to Reply to This Post
Old 07-24-2012, 12:09 AM   #2
Drewmeister
The Need 2 Freeze
Drewmeister's Avatar
Senior Member
 
Posts: 2,455
Last Seen: 07-21-2017
Age: 52
From: Phobos Lab
iTrader: 22 / 96%
Hey NL,

There are two possibilities for the temp rise. The first, as you mentioned, the heat pipe isn't functioning properly due to the sub 0c temps.. the water is possibly frozen inside. Two.. the heat pipes aren't able to transfer enough heat to the fins because of low oil movement.

An easy way to rule out the frozen heat pipe would be to warm the oil up above freezing and load test again. If temps are still bad then you'll need more oil movement through the heat sink.

The gpu's are fine as they're being cooled by both the heat sink and through the card's pcb itself.

I'll need a link or pic of the heat sinks base to help you with modding it.

Here's a good solid copper heat sink that techongiant found. http://www.newegg.com/Product/Produc...82E16835118036

Looking forward to your project!
United States  Offline
    Register to Reply to This Post
Old 07-24-2012, 10:33 AM   #3
Names_Lucky
Running System Stock
Regular Member
 
Posts: 32
Last Seen: 10-10-2016
iTrader: 0 / 0%
Thanks for reply Drewmeister, I think I'm going to test the heatpipe hypothesis first, as I really don't want to wait another 3 or 4 days for the oil to heat up, then cool back down from 0C just yet. Its at about -20C and it has taken awhile to get there from room temp. I'm going to try a zalman cooler, like you and tech recommended, and see if my temps improve. I'm really hoping it does, because I'm not sure there is much that can push -10 oil quickly short of a engine powered propeller.

After some research on other forums, I have found that others with sub zero temps using heatpipes are having similar problems. This brings me to another question, if I were to open up the pipes, allowing the oil to come in, and put them in an upright position. Would convection of the oil through the pipes be rapid enough to cool effectively? I would try this without asking, but I don't want to ruin the heatsink without warrant.

Last edited by Names_Lucky; 07-24-2012 at 11:18 AM.
United States  Offline
    Register to Reply to This Post
Old 07-25-2012, 12:14 AM   #4
Drewmeister
The Need 2 Freeze
Drewmeister's Avatar
Senior Member
 
Posts: 2,455
Last Seen: 07-21-2017
Age: 52
From: Phobos Lab
iTrader: 22 / 96%
The oil would be way too thick to enter the heat pipes.. probably even too thick for any air cooler at -20c. Btw.. which oil are you using?

You might want to consider something like a Corsair H60 or so. Slappa experimented with a H50 in -15c air and it seemed to perform just fine... awesome results too on his 1090T @5Ghz. Checkout the video here: http://www.youtube.com/watch?v=Ue1Ri...feature=fvwrel

In this situation I'd go with Technogiant's idea of butting the radiator right up against the evaporator fins. Oil movement/flow won't be an issue then, the thick oil will be acting more as a thermal interface material between the evap and radiator.

Glad to help out.. thanks for posting.

-Drew
United States  Offline
    Register to Reply to This Post
Old 07-25-2012, 10:46 AM   #5
Names_Lucky
Running System Stock
Regular Member
 
Posts: 32
Last Seen: 10-10-2016
iTrader: 0 / 0%
Drew,

At -20Cish is there any worry about the internal liquids of the h60 freezing? Or do they use an alcohol mix or something? I installed a 100% copper heatsink and and got about a 20C better load, however, if you think the corsair will work in -20C I'm up for trying it since my temps are still not very impressive. Also the oil is synthetic motor oil, I though it would be best given mineral oil freezes at around -30C, and this synthetic's pour point is below -50. It is viscous, but I think it was the best (affordable) oil I could have picked for this temp range.

Update: I asked Corsair directly if the H60 could handle temperatures of -20C. I got the response "Hello ,its a 60/40 mix the same that is in cars". This means that it should have a freezing point south of -40C correct? Although, Drew, you are the master of mixtures, do you think the 60/40 mixture will be too viscous at -20ish for the (most likely cheapo) H60 pump? Would it be better (and under $100) to get a water/"Drew mixture" loop solely for the CPU that can handle being submerged in -20C oil. I may be able to get 1 or 2 ehiem 1250s for cheap from a friend.

On another note: I may as well of went with the chilled liquid loop by now (I decided against it from a funding standpoint, but after eveything I think I ended up spending more on this build), I really thought the submerged build would perform better than it is. I have load temps of 50C with the full copper heatsink, same as my air temps. Something just seems wrong, heat conduction cannot be that bad without oil movement. I did have to mod my old 775 heatsink to make it fit, maybe something went wrong in that process. The GPUs work so wonderfully under load, and they are furthest away from the AC.

Last edited by Names_Lucky; 07-25-2012 at 04:34 PM.
United States  Offline
    Register to Reply to This Post
Old 07-28-2012, 12:43 AM   #6
Drewmeister
The Need 2 Freeze
Drewmeister's Avatar
Senior Member
 
Posts: 2,455
Last Seen: 07-21-2017
Age: 52
From: Phobos Lab
iTrader: 22 / 96%
Expect the temp of the coolant to be quite a bit warmer under load. With -20c oil.. my guess is you would be seeing -15c or warmer coolant temp in the H60. Slappa had his H50 at -15c so should be fine.

Of course a quality water cooling system would be the best route. You could pick up something like a d-tek fusion for under $50 and run some de-icer fluid for some excellent core temps and oc potential.

Yes.. heat conduction can be that bad without oil movement. At -20c the oil is like molasses and can't transfer the heat fast enough, it becomes heat soaked right around the fins of the heat sink. The gpu's have a much larger surface area to reject the heat.
United States  Offline
    Register to Reply to This Post
Old 07-29-2012, 05:37 PM   #7
Names_Lucky
Running System Stock
Regular Member
 
Posts: 32
Last Seen: 10-10-2016
iTrader: 0 / 0%
Thanks Drew for all the help. I have bought a h100 (at 75 bucks nib off craigslist I couldn't help it) I haven't been able to get time to install yet. I also bought some 850 cfm fans, they arrive Wednesday....I hope these have the power needed to get the oil moving again. (on a side note, are you familiar with fans? Would cfm or maximum static pressure be a better statistic to determine fan torque?) I bought high cfm fans with as high static pressure I could get with AC plus for under $100. I will need a separate PSU if I decide to go DC fans, which can have high static pressure for under $100.

Lastly, I noticed that after stress testing my GPUs while overclocking that the oil temp had risen to 20-30C (at least the top-most oil). Unfortunatly, I didn't reach my hand far down to see if the bottom oil was still cool before I shut the computer off and left it to cool overnight. Am I incorrect to think that a 5000btu (1400+ watt) AC is not sufficient in keeping the oil cool while gaming/stress testing, or is it that the oil is just not moving? Ice develops along the lines of the AC, does this degrade performance significantly? Or could I be running out of charge? It is a new AC, but it has been running constantly for weeks.
United States  Offline
    Register to Reply to This Post
Old 07-31-2012, 11:17 PM   #8
Drewmeister
The Need 2 Freeze
Drewmeister's Avatar
Senior Member
 
Posts: 2,455
Last Seen: 07-21-2017
Age: 52
From: Phobos Lab
iTrader: 22 / 96%
A higher static and wattage rating would probably lend itself to higher torque vs cfm. I've heard of ppl clipping the blades to help them spin faster in the oil. Keep in mind that all the heat from the fan motors will be transferred to the oil. You really need an Electric Scavenge Gear Pump but unfortunately they are expensive and loud.

A 5k btu a/c has plenty capacity for your heat load.. the lack of oil flow/movement is the problem. You should just raise oil temps to where you see the best core temps on the gpu's.

The ice on the suction line is normal at that temp range. Be careful of that ice melting and getting into the oil. Insulate the lines all the way below the oil level and make sure they're sealed airtight.

Try and mount that H100 radiator right on the evap coil.. maybe secure it with long tyraps.
United States  Offline
    Register to Reply to This Post
Old 08-02-2012, 11:38 PM   #9
Names_Lucky
Running System Stock
Regular Member
 
Posts: 32
Last Seen: 10-10-2016
iTrader: 0 / 0%
Well, crappy news and even more crappy news. New 850 cfm 1.2 inch H20 fans barely turn, and don't produce any oil movement over the components. The H100 is great in air, but submerged, even ziptied tight to the evap, is horrid. There is just not enough oil movement, and the cooling rate of the evap on the radiator must be very insufficient. I also asked an oil pump guy about getting my oil moving through lines (like with a scavenge pump you suggested) but he said to get it moving would require ALOT ($$$) of power to get the flowrate to a usable 1gpm. I think I'm going to take a different route and start working towards a closed chilled liquid line. With all the wisdom you've gained over the years, any recommendations on a pump that can handle that di-icer windsheild washer at -30ish through a CPU and 2xGPU blocks? Would my mpc35x I bought work, or would the plastic housing break over time?
United States  Offline
    Register to Reply to This Post
Old 08-03-2012, 12:39 AM   #10
Drewmeister
The Need 2 Freeze
Drewmeister's Avatar
Senior Member
 
Posts: 2,455
Last Seen: 07-21-2017
Age: 52
From: Phobos Lab
iTrader: 22 / 96%
What temp was the oil? Warm it to 0c and try. That size a/c will only do around 600w continuous at 0c anyways. If you want to pull it down to sub -20c for benching then go chilled liquid.

If you still can't get any reasonable flow and core temps at 0c then just covert to a liquid chiller and use the oil as a dielectric. Run de-icer through the water blocks and maybe some coils of copper to cool the oil slightly.

The mpc35x is machined from black POM so should be fine with the methanol and temps. Flowrate looks good for quite a few low restriction blocks.. stay away from the pin matrix designs. You don't need full coverage gpu blocks.. just make sure the vrm's have a heat sink.. the vram chips don't need anything and will be cooled by the oil.
United States  Offline
    Register to Reply to This Post
Old 08-03-2012, 08:31 PM   #11
Names_Lucky
Running System Stock
Regular Member
 
Posts: 32
Last Seen: 10-10-2016
iTrader: 0 / 0%
This thread, originally about heatpipes is getting pretty long, should I rename to " Lucky's not-so-lucky failure at phase change + oil" I do believe I am going the chilled liquid (de-icer) route now. I am not even sure I want to use ANY oil. I woke up this morning to water contaminated oil. I insulated the suction/other lines, but all the pulling computer/fans/ect out of the oil made me lose just enough oil for the top of the evap to be close enough to the surface to build ice. Additionally, even the surface of the oil forms condensation even with fans blowing on it. While water emulsion in oil is still non-conductive I really don't want to take any chances. After extensive tests I may use the oil to protect against condensation IF it proves non-conductive, but I have thrown my towel in for chilled oil cooling. I hope this doesn't deter any of the other smarter chiller builders out there that are thinking of this. Just learn from my mistakes so my attempt wasn't in vain.

So, Drew (or anyone else) few questions for new chilled liquid route. I am going for full GPU blocks, as I'm not sure I"m going to use oil, would the mcp35x still do the job, or do you recommend the little giant you have? I can still return the mpc35x, and the price difference is only 40-50 bucks. I recall you had to epoxy/insulate your little giant for the cold, how exactly is that done?

Update: Well I discovered I can't get full cover blocks for my msi 670 power editions. I thought they had same pcb as 680 but I guess it is custom (I still love them and will not exchange). So I"ll have to go with a universal design. So with this in mind, will sufficient airflow and some heatsinks keep them cool (possibly cooled some from cold block?) if the card is heavily insulated, or would I benefit from running these in room temp oil as I can get it moving pretty well at room temp. Is there some kind of mod I can do to get good heat transfer from the VRMs to the block? If I do decide to stay with oil submersion, would leaving the copper GPU blocks uninsulated result in enough cooling power to keep the oil from rising? Would this harm my GPU temps drastically. Sorry about SO many questions, just really don't want to make many more costly mistakes.

Condensation proofing: I did this awhile back for a cold AC air build I had, and I must of did a decent enough job since I didn't fry anything. However, this time around I have oil coated parts. Anyone heard of a good way to remove the oil, OR instead of the initial nail polish/conformal spray can I put a good layer of dielectric grease over the oil on everything then insulate over that,leaving chips/cap tops uncovered? I will most likely do the whole board/GPUs, as I am very paranoid about frying my stuff. I"ll also be using good fans to keep constant air movement over the system to cool the chips and help against condensation.

Also, with cold temps should acetal be avoided? I thought I saw a post by you confirming it is ok to -30 to-40ish, but I cannot find it, maybe it was a dream (yes, unfortunately I have dreams and NIGHTMARES about this **** computer)

As always, thanks for the help

Last edited by Names_Lucky; 08-04-2012 at 02:02 AM.
United States  Offline
    Register to Reply to This Post
Old 08-05-2012, 10:47 PM   #12
Drewmeister
The Need 2 Freeze
Drewmeister's Avatar
Senior Member
 
Posts: 2,455
Last Seen: 07-21-2017
Age: 52
From: Phobos Lab
iTrader: 22 / 96%
Well at least you gave it a good shot there NL. I'm sure it would have worked decent at +0c temps but then you would be wasting the cooling power you have. Then of course there are the many issues associated with using and maintaining an oil cooled system, and as you know, are greatly compounded when going sub 0c. I would just avoid it all together.. even as a dielectric.

So the next step is to get all the oil off your hardware so the condensate proofing material will adhere to it. I would probably start with mineral spirits.. get a gallon of it and use a soft brush to clean most of the oil off first. Then follow that up with a good electronics degresser like CRC Electronics Cleaner or NOVEC DEGREASER. I believe you can get the CRC local at places like Autozone. http://www.autozone.com/autozone/acc...er=119711_0_0_ Put a fan on it right away to dry it out. I could probably help you out with the insulation materials.. pm me.

The VRM's only need a heat sink and small fan.. 20-30cfm is plenty. The vram doesn't need anything as it will get plenty of cooling from the cool pcb via the chilled waterblock.. the fan from the vrm's will help cool them too.

The Little Giant I use has performed flawlessly at extreme temps. The only pump I had trouble with was the Danner Mag pond pump.. those have very thin plastic volutes and become brittle over time. They're ok if you leave then alone.. mine cracked when I was changing the barb ftgs. My recommendations are the LG's or Iwaki's. Your pump should be fine with non restrictive blocks. Remember.. stay away from pin matrix. If you want to step up to an Iwaki then you're looking at $160-$250. MD 20, 30 or 40 series depending on water blocks used.

Acetal Properties: http://www.dynalabcorp.com/technical_info_acetal.asp

The design of the block's mating surfaces and type of o-ring used will have more impact on it's resistance to temps.
United States  Offline
    Register to Reply to This Post
Sponsored Links:
Please Register to Post a Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -6. The time now is 08:43 AM.

Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.
Powered by vBulletin
Copyright ©2000 - 2016, EXTREME Overclocking