TEC/Peltier CPU Chilled Water Cooling

[CompuTronix]

Hi Ry,

The answer is C: The coolant temperature continued to decrease while IBT was running.

Cheers!

CT

 

[4Ryan6]

Hi Ry,

The answer is C: The coolant temperature continued to decrease while IBT was running.

Cheers!

CT

I ran some additional tests I thought you might be interested in, when we ran that test the CPU (3770K) was clocked at 5ghz using 1.375v with Hyper Threading disabled and as you said the coolant temperature was continuing to drop while Intel Burn Test was running.

When Hyper Threading is enabled it takes 1.395v to stabilize the 3770K at 5ghz and when running Intel Burn Test with Hyper Threading enabled, the coolant temperature holds it's reading at what the test is started at, it does not continue dropping.

Ry

 

[Joe1776]

That some good advice right there. maybe ill have to step up to a 4 peltier system and put all four on switches so i can pick a chose which ones are turned on.

How about using pwm controll with arduino or something and a power mosfet. This way you can get 255 levels of cooling with any number of peltiers.

 

[Iamsoda]

Could I just stick this right on my CPU and have a massive high rpm fan plus heatsink on it?

 

[4Ryan6]

I just went through an extreme experience with my GTX Titan Hydro Copper graphics card, I had already mentioned the sorry mounting of the water block to the card with the GPU die itself only covered with thermal compound about 85%, with one corner completely bare of TIM.

Discovering that made me extremely grateful I had never pushed it too hard overclocking it, I was also thoroughly disgusted at the gelatinous crap thermal padding that was used between the memory chips and VRs, I had never even seen thermal padding that crappy, it was like 1.5mm thick gel, is my best description of it.

I would have known right then the problems I was looking at if I had investigated the memory chip and VR modules impression depth and contact to the thermal padding, as I would have discovered the serious discrepancies of the memory chips out of a flat level plane.

The memory chips were hardware mounted to best describe it as a roller coaster, of course I did not discover this issue until I bought a Watercool Heatkiller water block to replace the Swiftech Komodo that actually comes on the GTX Titan Hydro Copper.

When mounting the Heatkiller and testing for fit my initial goal was excellent GPU die contact so I wasn't even concerned with memory chip issues until I got the dies coverage I was after, it was then I discovered the hidden in plain sight nightmares the graphics card was hosting.

Before I continue I want you the reader to understand that the Heatkiller water block was excellent, however you cannot get good contact when a flat surface is being mounted to a roller coaster, Heatkiller provided .5mm thermal padding for memory chips and 1mm for the VRs.

The VRs were not a problem but the memory chips not being manufactured level and using the supplied using .5mm thermal pad 6 were squished flat and 6 barely even touched the thermal padding at all, so 6 of the 12 memory chips on the card were barely touching the thermal padding at all.

So the initial fitting of the water block to the GPU die was literally a total waste of time as when the thermal pads were added the GPU contact was lifted and not making good contact, that's when it all turned into a nightmare!

One of the worst things about me is I am a perfectionist, and I could not stand what I was looking at, I fully understood why the card was limited in it's performance, because full contact with everything that needed cooling just wasn't happening.

There are 24 memory chips on this graphics card 12 on the front side with active cooling and 12 on the backside with passive cooling, I had already decided with the Heatkiller block to bypass the cards back plate and just mount a set of 12 ram sinks and cool the backside with air.

Even though we are all at some point of buying a graphics card especially one the cost $1300 US dollars after performance but it seems we're getting looks and fluff over raw cooling for best overall card performance and overclocking.

Chock this experience up to another lesson learned the hard way!

So is that the end of the story?

No!

I was extremely discouraged at this point of attempting to get excellent coverage and it forced me as a resolution to run regular TIM on the 6 memory chips that made contact and the .5 thermal padding on the 6 chips that did not, that achieved excellent contact with the GPU die again but but revealed another problem.

The outer 2 memory chips across the top only had about 2/3rds partial contact and required a lot of TIM to touch.

Pissed Off as Hell, comes to mind as I knew I was just not going to be satisfied with what was happening, and this was all because EVGAs manufacturing process of the graphics card in the first place was quality controlled by monkeys!

I am saying this because both of my past 580GTX EVGA cards were memory chip flat, and I had mounted Heatkiller blocks to them with just TIM only and had excellent contact, that's why I stated the Heatkiller water block was not the problem.

So what now?

I had zero intentions of running the card like that, so I had considered getting a universal fit GPU water block and thermal bonding Ram Sinks and cooling the memory and VRs with air.

However I had an additional water block an Alphacool that I had intended to flow modify for the peltier setup, the only problem was it was a clear 1366 socket mount acrylic top CPU water block, and we all know you cannot mount a 1366 socket on a 1155 socket, much less to a graphics card, so some serious modifying was in order.

So end result that I am happy with as for the first time since the GTX Titan left the EVGA factory door the GPU die is fully covered, and every memory module and VR has it's own Ram Sink..



Leak Testing:



Back in Business:



This is my resolved issue, and I shudder to think what may be under some of your graphics cards, thus the reason for posting this, as I thoroughly recommend checking it out for yourself, (if you of course have the hands on ability to do so!), to see what's actually going on under the hood of that graphics card you are running.

Edit: Gaming load has dropped to 28c with a 24c ambient.

Edit: 1, Ran an hour of benchmarking various programs Uningine Heaven, and Valley (let Valley just run for 30 minutes before starting the benchmark), 3DMark 11, etc.the highest temperature reached was 33c at a 26c ambient.

Memory chips on the back of the card are 26.4c, memory chips on the front of the card are 26.2c, (obviously the copper hedgehog memory sinks on the front are the best, taking into consideration they are thermally bonded and the ones on the rear are aluminum and just stuck on).

The VRs are 25c with a 26c ambient? (I have to admit I did not expect that, nor do I fully understand why but I did change out one of the front intake fans to a low speed San Ace 120mm x 38mm, which is a little louder but outputs quite a bit more air volume.)

Heat sink temperatures were taken with a point and shoot laser thermometer.

 

[tech-wreck]

nice job with the copper hedgehog

i have had a thought, what would happen if i re-wired the peltiers so they were all parallel, and have two power connectors instead of two parallel pairs on one connector?

 

[4Ryan6]

nice job with the copper hedgehog

i have had a thought, what would happen if i re-wired the peltiers so they were all parallel, and have two power connectors instead of two parallel pairs on one connector?

Thanks it does look like a copper hedgehog!

To your thought, I do not know what would happen, you'll just have to try it and see?

 

[4Ryan6]

It is a good feeling to know I have full coverage over the entire GPU Die, and also that all the Memory and VR chips have equal cooling and are all being cooled.

The thermal compound I used on the GPU Die is Artic Silver 5, after much in depth research I discovered that Artic Silver 5 actually does contain 99% silver particles as they claim and since Silver IE. Ag has the best Thermal Conductivity of all the elements on the periodic table, you really cannot ask for any better and actually get it. (Check out the periodic table sorted by thermal conductivity yourself, don't take my word for it!)

Most Thermal Compounds silver in color are actually a type of aluminum or carbon based, whereas aluminum is a good thermal conductor, it is just not the best.

One thing I can unequivocally state regarding Artic Silver 5 from my past usage of it, is that the thermal performance always improved at least 3c ~ 5c after it had time to cure.

When I ran the Thermal Compound Roundup back in 2011 Artic Silver 5 did very well because it was not allowed it's full cure in time, with it's cure in time allowed it would challenge the best of the best in the roundup.

No, I am not selling Artic Silver 5, however it is what it is, and Artic Silver 5, is one of the things you may buy today that you actually get what you pay for!

Have a wonderful day everyone! Ry

 

[PCNoob27]

Hi Ryan. Im new to the forum. Have been reading ok Toms for almost a decade and your post sealed the deal when finally deciding to join. I can't explain how much I admire your perseverance to achieve your goals! I do have a few questions relating to the Peltier Waterloop. With your knowledge about Peltier watercooling and the effects of heat and cooling.. I am just asking for your educated guess.

Would it be possible to cool the cpu with just a Peltier waterblock, a design such as yours, redirected back to the cpu waterblock? A smaller, more concentrated, waterloop design. Or would cpu temperatures be too much for the Peltier to effectively cool in such a small loop?

Possibly adding a small reservoir and a second Peltier? Like such.. cpu waterblock-> Peltier waterblock-> small reservoir-> Peltier waterblock-> back to cpu waterblock? Total water about .5 gallons at a rate of ? .5 - 1 gpm?
How long did it take for your reservoir to reach your desired temps from ambient? Such as, just starting up your system?

Just theoretical brainstorming your design is superb! Just wondering how the effects of temperature rise or decrease when subjected to Peltier and cpu temps in a single revolution.

Awesome titan setup btw! Love your basic, but extremely effective cooing techniques!

 

[eL3ctro]

4ryan6 I've been direct to come here from CompuTronix basically its about my water cooling I've recently installed a Raijintek Triton and I have installed red dye although after about 5 hours nothing wrong, I've noticed red dye I've mopped it of my graphics card wasn't a lot I've inspected my water block it seems like to leak from there are the tubing but I am paranoid I've put a paper sheet down to see if there is any more red dye liquid whatr is your advice about this and what should I do. I've had the sheet down for about 10 mins although there is no liquid on there although every so often there will be a drop I've noticed. I can't seem to locate it.

 

[4Ryan6]

4ryan6 I've been direct to come here from CompuTronix basically its about my water cooling I've recently installed a Raijintek Triton and I have installed red dye although after about 5 hours nothing wrong, I've noticed red dye I've mopped it of my graphics card wasn't a lot I've inspected my water block it seems like to leak from there are the tubing but I am paranoid I've put a paper sheet down to see if there is any more red dye liquid whatr is your advice about this and what should I do. I've had the sheet down for about 10 mins although there is no liquid on there although every so often there will be a drop I've noticed. I can't seem to locate it.

I am not familiar with the Raijintek Triton so before answering questions regarding your situation I am researching it first, before advising you, and in the research process I discovered this negative review at Newegg.com.

The quote below is from Newegg Raijiintek Triton customer reviews.

Proceed with caution this thing leaks! A crack developed in the reservoir and leaked down on my video card destroyed it. I was able to catch the problem preventing further damage to my Asus Crosshair board.

I would inspect that reservoir for any cracks in the reservoir itself, or leaks around the fill ports O-ring seal.

TechPowerUps Review of the Raijintek Triton, does not show them opening the access port to the reservoir and adding dye but it looks like the traditional rubber O-ring sealed G 1/4 access port, which means it requires the O-ring and needs to be finger tight but not overtightened to properly seal.

No offense but just an observance, I do not like the fact they put the reservoir directly on the pump block, and it being clear acrylic as well, a leak at the reservoir holding that much coolant could have catastrophic consequences, it could have been designed better IMO.

Acrylic has also had a past rumor history of not being a good container for some types of mixed metal coolant additives and I would suspect that is not clear distilled water that is in the loop itself but some kind of specific coolant you can add dye to. (Surely the manufacturer would have constructed the Raijintek of all compatible components and coolant, so that shouldn't be a concern.)

Inspect the reservoir for cracks and report back what you discover?

 

[eL3ctro]

4ryan6 I've been direct to come here from CompuTronix basically its about my water cooling I've recently installed a Raijintek Triton and I have installed red dye although after about 5 hours nothing wrong, I've noticed red dye I've mopped it of my graphics card wasn't a lot I've inspected my water block it seems like to leak from there are the tubing but I am paranoid I've put a paper sheet down to see if there is any more red dye liquid whatr is your advice about this and what should I do. I've had the sheet down for about 10 mins although there is no liquid on there although every so often there will be a drop I've noticed. I can't seem to locate it.

I am not familiar with the Raijintek Triton so before answering questions regarding your situation I am researching it first, before advising you, and in the research process I discovered this negative review at Newegg.com.

The quote below is from Newegg Raijiintek Triton customer reviews.

Proceed with caution this thing leaks! A crack developed in the reservoir and leaked down on my video card destroyed it. I was able to catch the problem preventing further damage to my Asus Crosshair board.

I would inspect that reservoir for any cracks in the reservoir itself, or leaks around the fill ports O-ring seal.

TechPowerUps Review of the Raijintek Triton, does not show them opening the access port to the reservoir and adding dye but it looks like the traditional rubber O-ring sealed G 1/4 access port, which means it requires the O-ring and needs to be finger tight but not overtightened to properly seal.

No offense but just an observance, I do not like the fact they put the reservoir directly on the pump block, and it being clear acrylic as well, a leak at the reservoir holding that much coolant could have catastrophic consequences, it could have been designed better IMO.

Acrylic has also had a past rumor history of not being a good container for some types of mixed metal coolant additives and I would suspect that is not clear distilled water that is in the loop itself but some kind of specific coolant you can add dye to. (Surely the manufacturer would have constructed the Raijintek of all compatible components and coolant, so that shouldn't be a concern.)

Inspect the reservoir for cracks and report back what you discover?

There are no cracks round the reservoir I've had it out and inspected and had it on yet no leak on out of the case once it there is a single drop of fluid every so often that appears on the paper towel that I placed inside. But I just can't seem to pin point it.

 

[4Ryan6]

Hi Ryan. Im new to the forum. Have been reading ok Toms for almost a decade and your post sealed the deal when finally deciding to join. I can't explain how much I admire your perseverance to achieve your goals! I do have a few questions relating to the Peltier Waterloop. With your knowledge about Peltier watercooling and the effects of heat and cooling.. I am just asking for your educated guess.

Would it be possible to cool the cpu with just a Peltier waterblock, a design such as yours, redirected back to the cpu waterblock? A smaller, more concentrated, waterloop design. Or would cpu temperatures be too much for the Peltier to effectively cool in such a small loop?

Possibly adding a small reservoir and a second Peltier? Like such.. cpu waterblock-> Peltier waterblock-> small reservoir-> Peltier waterblock-> back to cpu waterblock? Total water about .5 gallons at a rate of ? .5 - 1 gpm?
How long did it take for your reservoir to reach your desired temps from ambient? Such as, just starting up your system?

Just theoretical brainstorming your design is superb! Just wondering how the effects of temperature rise or decrease when subjected to Peltier and cpu temps in a single revolution.

Awesome titan setup btw! Love your basic, but extremely effective cooing techniques!

Hi PCNoob27, and Welcome to Toms Hardware!

Would it be possible to cool the cpu with just a Peltier waterblock, a design such as yours, redirected back to the cpu waterblock? A smaller, more concentrated, waterloop design. Or would cpu temperatures be too much for the Peltier to effectively cool in such a small loop?

I am not 100% clear on exactly what you are asking, could you be more expletive?

You start by asking, " Would it be possible to cool the cpu with just a Peltier waterblock", that could be taken as a direct mount to the CPU?

But then you continue with, "a design such as yours, redirected back to the cpu waterblock?", regarding the first part of the question the answer is yes, but not following a design like mine, that's what's kinda unclear?

Any type of direct CPU peltier water block mounting requires fully insulating the motherboard itself, the CPU socket and the surrounding area of the socket and also the backside of the motherboard around the socket area as well, because direct mounting results in sub zero cold transfer to the CPU. (Which creates ice from the moisture in the air, and when ice melts, you get water, IMO, the most commonly overlooked thing in that type of mount is the air space underneath the CPUs heat spreader.)

A smaller, more concentrated, waterloop design. Or would cpu temperatures be too much for the Peltier to effectively cool in such a small loop?

Or on the other hand are you asking if it could all be done with just one peltier/water block TEC assembly?

Possibly adding a small reservoir and a second Peltier? Like such.. cpu waterblock-> Peltier waterblock-> small reservoir-> Peltier waterblock-> back to cpu waterblock?

Can you see why I am a little confused?

So if you can clarify exactly what you are asking, I'll do my best to answer you fully.

How long did it take for your reservoir to reach your desired temps from ambient? Such as, just starting up your system?

Now that I can directly answer, I have 2 TEC assemblies automatically powered when the PC is turned on I have a 3rd TEC assembly independent switch controlled.

Normal PC power up with the 2 TEC assemblies it take 30 minutes for the coolant to drop from ambient to it's run coolant equilibrium temperature, which is normally 13c below ambient when my window AC unit is not running.

If I activate the 3rd TEC assembly that time drops to 20 minutes and does not stop at 13c below ambient, it will actually drop down into condensation danger levels if the 3rd TEC assembly is not powered down, even with the room AC off.

So I have to monitor the coolant temperature closely when the 3rd TEC assembly is energized as dropping down to an actual coolant temperature of 10c is fine which would be 16c below ambient, but going below 10c condensation begins forming at 8c which is 18c below ambient.

I normally run in the 10c coolant temperature area because in my office I have to run the AC unit, when the sun is directly on the back of my office, believe it or not in South Carolina even some days in the Winter.

I hope I have answered at least some of what you want to know?

 

[rubix_1011]

Perhaps I need to add a video to the sticky on the teardown of a GPU in order to evaluate the stock cooling solution.

 

[PCNoob27]

Thanks for the reply Ryan. I was working near Augusta, Ga all last fall/winter, let alone the summer, so I know all about the heat . Sorry about the confusion in my first couple questions. To clarify, similar to your initial setup, except without the insulated reservoir. Just 2 waterblocks total. One on the cpu and one on a Peltier/tec. Obviously one would need a small reservoir with a pump, but would a single Peltier be sufficient in cooling the cpu in a smaller setup like this?

So, to get a better understanding, the loop goes like this.

Small reservoir/pump flowing water into a waterblock (modified like yours) attached to a Peltier/tec, then continue flowing to the cpu waterblock, and finally back to the reservoir/pump.

Maybe adding a second Peltier in the loop would help cool it?

Im just wondering if the Peltier could cool the water faster than the cpu could heat it. End result being not needing an insulated reservoir. Ideal for micro atx builds.

Like I said before tho, Im just brainstorming. A setup like yours is perfect. Not trying to reinvent the wheel here haha

 

[PCNoob27]

I seen those vram heatsinks on your titanx for sale in packs of 10. Did you buy the other 2 separately? I first built my pc 3 years ago and just recently started overclocking.

My build specs:

3770k @ 4.7 ghz 1.285 vcore (4.8 w/o ht same vcore)
Z77x-ud3h
8Gb g skill ram
1Tb wd hdd
Gtx 690
TX850 psu

Considering selling the cpu, motherboard, ram, gtx 690 and upgrading to x99 or just rocking my setup until broadwell-e comes out.

Either way im upgrading to the titan x and probably add an ssd. Not too get off topic, I waited years before I overclocked and the chip is handling heat and v core very well. Could it be that it had sufficient time to 'break-in' so to speak? Im using just a zalmann cnps 9900 and oem case fans. Temps are high 70s under prime95 small ffts full load. Even ran cinebenchr15 while running prime95, which I think just cancelled prime95 momentarily because temps dropped almost 20c until cinebench finished.

Tested temps with and without HT. There is about a 10c difference. Even managed 5 Ghz HT off with 1.375 vcore. Temps high 80s prime95 for 20 minutes. Crashed cinebench almost immediately haha. How much would temps drop if I added an h100 aio?

 

[4Ryan6]

Perhaps I need to add a video to the sticky on the teardown of a GPU in order to evaluate the stock cooling solution.

That's a good idea, Please do so.

I would have gone into much more detail with my graphic card discoveries, (picture wise), but my drive to get to a resolution to the situation after wasting the money on a full coverage Heatkiller water block, was greater than an in depth expose', Sorry!

Any additional info to help others is appreciated!

Ry

 

[4Ryan6]

Thanks for the reply Ryan. I was working near Augusta, Ga all last fall/winter, let alone the summer, so I know all about the heat . Sorry about the confusion in my first couple questions. To clarify, similar to your initial setup, except without the insulated reservoir. Just 2 waterblocks total. One on the cpu and one on a Peltier/tec. Obviously one would need a small reservoir with a pump, but would a single Peltier be sufficient in cooling the cpu in a smaller setup like this?

So, to get a better understanding, the loop goes like this.

Small reservoir/pump flowing water into a waterblock (modified like yours) attached to a Peltier/tec, then continue flowing to the cpu waterblock, and finally back to the reservoir/pump.

Maybe adding a second Peltier in the loop would help cool it?

Im just wondering if the Peltier could cool the water faster than the cpu could heat it. End result being not needing an insulated reservoir. Ideal for micro atx builds.

Like I said before tho, Im just brainstorming. A setup like yours is perfect. Not trying to reinvent the wheel here haha

First of all, the answer to your question totally depends on the power handling capabilities of the peltier used?

Without an insulated reservoir you have no cold buffer at all, so your cold produced by the peltier would have to be way more that the heat load generated by the CPU, or else the CPUs heat load exceeds the cold output of the peltier and probably will cause it to stall out in it's cold production. (and that is bad!)

I seen those vram heatsinks on your titanx for sale in packs of 10. Did you buy the other 2 separately? I first built my pc 3 years ago and just recently started overclocking.

My build specs:

3770k @ 4.7 ghz 1.285 vcore (4.8 w/o ht same vcore)
Z77x-ud3h
8Gb g skill ram
1Tb wd hdd
Gtx 690
TX850 psu

Considering selling the cpu, motherboard, ram, gtx 690 and upgrading to x99 or just rocking my setup until broadwell-e comes out.

Either way im upgrading to the titan x and probably add an ssd. Not too get off topic, I waited years before I overclocked and the chip is handling heat and v core very well. Could it be that it had sufficient time to 'break-in' so to speak? Im using just a zalmann cnps 9900 and oem case fans. Temps are high 70s under prime95 small ffts full load. Even ran cinebenchr15 while running prime95, which I think just cancelled prime95 momentarily because temps dropped almost 20c until cinebench finished.

Tested temps with and without HT. There is about a 10c difference. Even managed 5 Ghz HT off with 1.375 vcore. Temps high 80s prime95 for 20 minutes. Crashed cinebench almost immediately haha. How much would temps drop if I added an h100 aio?

The VR heatsinks did come in a pack of 10? I only used 8 of the 10?

I run 5ghz on my 3770K with HT off at 1.375v, but with the chilled water cooling it is rock solid.

Regarding the H100 question, I do not know, I have never used an AIO cooler, nor would I.

The greatest failure running radiator water cooling for CPU overclocking reasons is not having enough radiator cooling field to handle the CPUs overclocked heat load!

That is why I would never use an AIO CPU cooler! (Because none of them can truly handle the overclocked CPUs DeltaT, because they do not have enough radiator cooling field, they are designed to fit in most even small cases so they do not possess the needed cooling capabilities.)

There is a big difference between cooling satisfaction and cooling regret, and Oh Well, I was hoping for better results but I'll settle for what it is!

Obviously I am not one of the latter as I will not be satisfied with regret, else I would have just Oh Well, and kept running Heatkiller waterblock, but I knew it could be better so I continued until I got it cooled like I wanted.

It is an expensive attitude that has cost me plenty of money over the years, and a lot of lessons learned the hard way!

But, No Regrets! (It is just money, you cannot take it with you when you leave this earth!)

The CPU is factory designed to protect itself, (Whether it is an Intel or AMD CPU), some of it's protection you can monitor and control, and some of it's protection features you cannot monitor and control.

The underlined above shows up when you have unexplained performance drop off when running a benchmark and discover your performance was better without the overclock, or certain levels of overclocking begin showing a performance drop.

That happens when the CPU internally protects itself, simply because it is getting too hot, too fast.

From my own discoveries!

Depending on how fast the heat is rising, the CPU responds to control that temperature, that's why chilled water cooling and running below ambient works so well, you are actually running below the CPUs design parameters.

So under those circumstances some of the auto protective features of the CPU are never activated.

These are my discoveries from running below ambient cooling for years and you won't find verification in CompuTronix Intel Temperature Guide, because it is 100% accurate IMO, but it is totally based on CPU design ambient cooling.

And now it is time for my 2nd cup of coffee!

 

[4Ryan6]

There are no cracks round the reservoir I've had it out and inspected and had it on yet no leak on out of the case once it there is a single drop of fluid every so often that appears on the paper towel that I placed inside. But I just can't seem to pin point it.

Obviously you do have a leak, keep looking for it, you have got to find it and stop it!

 

[eL3ctro]

I found the leak cracked in the same place as other models squirting out of a hole although took it back I picked the Corsair GTX H100i GTX hopefully a good choice seems to not keep CPU as cool runs a little bit hotter 2 degrees under-load prime95 but about 8 degrees lower on idle. 22 degrees idle - 79 - 80 load

 

[4Ryan6]

I found the leak cracked in the same place as other models squirting out of a hole although took it back I picked the Corsair GTX H100i GTX hopefully a good choice seems to not keep CPU as cool runs a little bit hotter 2 degrees under-load prime95 but about 8 degrees lower on idle. 22 degrees idle - 79 - 80 load

Hopefully it won't leak!

 

[PCNoob27]

Thanks for the input Ryan. You answered some questions I didn't even ask! Sometimes when I would run cinebench to test different frequencies, with different voltages, I would get mixed results.

I realized my questions I had asked yesterday were answered, in a roundabout way, after further reading in pages 3-4 in this thread. The same questions are being asked over and over it seems. I am guilty of it as well haha sorry. I have found something interesting that maybe you could answer.

It seems that most of the Peltier setups that mount directly to the cpu have excellent cooling capabilities while at idle. When it is subjected to a load, the temps rise back above ambient. Sometimes 20c above ambient. From what I've learned about your trial and error with the Peltier, regarding hot and cold sides of the Peltier effecting each other, it seems that when the load is applied, the heat from the cpu is causing the cold side to get too hot. Add in the heatsink and fan cooling the hot side even more, the efficiency of the Peltier is significantly reduced.

Could also be that when at idle, the Peltier has sufficient watt dissipation and when under load it is too much for that size of Peltier to effectively handle? The tests I have seen vary from 150-245w Peltiers.

About the vram heatsinks.. I seen packs of 10 for sale and the titan x had 12 GBs of vram. Maybe I missed something? What are the smaller heatsinks on the right side of the gpu in your pictures for?

Thanks for answering my questions. It helps alot. What is your next project?

 

[4Ryan6]

Thanks for the input Ryan. You answered some questions I didn't even ask! Sometimes when I would run cinebench to test different frequencies, with different voltages, I would get mixed results.

You're Welcome, but you'll probably be telling me to shut up before it is over with!

Since I have been defending this cooling solution since the beginning, wading through all the initial doubts of whether this would work or that would work, I do find myself going over old discoveries and sometimes rambling.

However there are good points to share that can save someone a lot of time and money discovering what I already have discovered the hard way, so I Ramble On. (Ahh "Ramble On" Led Zeppelin, reminds me I haven't listened to that song in a while, must rectify that right now!)

[video="https://www.youtube.com/watch?v=a3HemKGDavw"][/video]

Wow did that ever bring back memories!

I saw them live when they did one of their US tours, never forget it!

Now back to the thread.

I realized my questions I had asked yesterday were answered, in a roundabout way, after further reading in pages 3-4 in this thread. The same questions are being asked over and over it seems. I am guilty of it as well haha sorry. I have found something interesting that maybe you could answer.

It seems that most of the Peltier setups that mount directly to the cpu have excellent cooling capabilities while at idle. When it is subjected to a load, the temps rise back above ambient. Sometimes 20c above ambient. From what I've learned about your trial and error with the Peltier, regarding hot and cold sides of the Peltier effecting each other, it seems that when the load is applied, the heat from the cpu is causing the cold side to get too hot. Add in the heatsink and fan cooling the hot side even more, the efficiency of the Peltier is significantly reduced.

Could also be that when at idle, the Peltier has sufficient watt dissipation and when under load it is too much for that size of Peltier to effectively handle? The tests I have seen vary from 150-245w Peltiers.

You read and came to your own conclusion!

That is absolutely freaking fantastic!

I'll just add that what you discovered regarding the load temperature is exactly why I do not direct mount to the CPU as that mounting allows the CPU to affect the peltiers output.

Mounting the peltier to a flow modified water block with a heat sink handling the hot side allows no direct effect of the CPUs load heat, as the TEC assembly is dumping it's generated cold into the insulated reservoir and the cold stored in the reservoir is being pumped straight to the CPU.

In that respect the reservoir becomes a cold storage buffer, and the CPUs load generated heat is countered by the peltiers balanced cold output.

About the vram heatsinks.. I seen packs of 10 for sale and the titan x had 12 GBs of vram. Maybe I missed something? What are the smaller heatsinks on the right side of the gpu in your pictures for?

OK, the Swiftech Copper BGA Memory Ramsinks come in a package of 8, I bought 2 packs, the Enzotech MOS-C1 Copper MOSFET Sinks, (IE: Graphics Voltage Regulators), come in a package of 10.

I used copper on the front side of the card, (the power side of the card), and seeing as how they are upside down I used Thermal Adhesive to secure them, I used low profile Watercool Aluminum Ramsinks pack of 12 on the 12 memory chips on the backside of the card, and since they are flat and on top just used the sticky thermal that came with the kit, they come in either a pack of 8 or 12.

The 3gb of memory on the back side of my Titan has never even been used by any application I have ever run, I am sure it has a reason for being there, past card hype and sales, but as yet, I have not discovered why it is even back there?

The backside memory was factory passively cooled underneath a back plate, so I've left it passively cooled, but much better than the factory did!

Thanks for answering my questions. It helps a lot. What is your next project?

My next project is to shrink down this cooling to a more manageable solution, BUT (and that's a big but!), I am waiting to see what Skylake brings, as I am not actually rolling in money and these experiments cost money, so I either get to upgrade to Skylake or continue experimentation.

That Sir, depends on what Skylake brings to the table, is it worth upgrading to, or not? (which will be hopefully around Christmas. )

Broadwell wasn't worth messing with, so we'll see?

Which ever takes precedence over the saved up money, shrinking it all down is on the future drawing board.

 

[PCNoob27]

Led Zeppelin... was my favorite band back in the 'experimental' days those were the days. I, myself, have issues with rambling. Must be something in the water.

Anyways, skylake 66-6700k are rumored to be released this August. I don't think we'll see much improvement from where we sit now (current overclock). Rumor is that IPC is suppose to be up to 29% faster than 4790k., but the leaked benchmarks show most improvement to be in igpu dependent tests. This should be taken with a grain of salt, as nothing is ever as it seems, until proven otherwise. A rumor is just a rumor..

Of course, with each new chipset, there is new bells and whistles to be had.

 

[4Ryan6]

@PCNoob27

Now don't get offended, my following comments are from my own past experiences, OK.

Those are not bad temps you're running with your 3770K OC, considering the Dinosaur Zalman CNPS 9900 cooler you're running.

I had one of those earlier Zalman models that the fan wasn't sandwiched inside, and the problem with that style of Zalman cooler is you cannot add another cooling fan, and if you have to replace the fan it will be a Zalman fan, because nothing else will fit it, but the Zalman fan.

I lapped the base of mine and that was a bad day, it was like trying to move the base with an elephant sitting on top!

Worst lapping project I ever attempted, and it got revenge too, cut the crap out of my hands, both of them, If you're not careful installing it, it will filet you!

The only air cooler I had that was worse installing was the original Tuniq Tower, that thing cut me regularly!

Do you have plans of replacing that Zalman?