TEC/Peltier CPU Chilled Water Cooling

[4Ryan6]

Ruwed If I completely remove my 5ghz overclock I would only be able to use one pelteir assembly and even with just one running, it would go lower than I want it to with no CPU overclock.

You keep talking about CPU TDP exactly what CPU are you running?

Edit: Belay That Question I looked at your Overclock.net thread and you are running an Intel 6 core 5820K TDP 140w and that is where you are basing your TDP numbers comparative to my 3770Ks 77w is that correct?

I would like to quote something you said in the OCN thread in the very beginning first post of your thread.

My largest worry at this point is that the TEC's are not going to be large enough and I may either need a fourth or to step up to the next size TEC. I currently have three 12715's on the way, running at 12v they should be running a Qmax of around 120w, so all together that should give me 360-ish watts, kinda. It's the efficiency loss to --> copper block --> water that has me worried..... Oh well, nothing to do but try!

You mention TDP equality with your setup but how does that compare with raw temperature output of the CPU that has to be accounted for?

In your thinking you should be able to get a closer comparison to the 3770K vs 5820K wattage load by disabling 2 of your cores it is the same 22nm manufacturing process and I do not run Hyper Threading either, because in some applications it hinders performance instead of helping.

Hyper threading does effect overall CPU temperature that has to be dissipated so does running system memory speeds past your CPUs 2133mhz max memory controller design specifications, I don't know what memory speed you are actually running?

If your cooling is indeed comparable with 2 of your cores disabled you should (CPU luck of the draw of course), be able to overclock the 4 cores to 5ghz, but even if you cannot reach 5ghz go as high as you can for a thermal reservoir comparison?

Why don't you test that for a comparison?

If you run the tests with 2 of your cores disabled it will give a better overall thermal capability to see exactly where your cooling is falling performance wise, it will also help in future understanding of what it takes to cool higher TDP CPUs, and I am definitely interested in that shared data?

 

[Ruwed]

Ruwed If I completely remove my 5ghz overclock I would only be able to use one pelteir assembly and even with just one running, it would go lower than I want it to with no CPU overclock.

You keep talking about CPU TDP exactly what CPU are you running?

Edit: Belay That Question I looked at your Overclock.net thread and you are running an Intel 6 core 5820K TDP 140w and that is where you are basing your TDP numbers comparative to my 3770Ks 77w is that correct?

I would like to quote something you said in the OCN thread in the very beginning first post of your thread.

My largest worry at this point is that the TEC's are not going to be large enough and I may either need a fourth or to step up to the next size TEC. I currently have three 12715's on the way, running at 12v they should be running a Qmax of around 120w, so all together that should give me 360-ish watts, kinda. It's the efficiency loss to --> copper block --> water that has me worried..... Oh well, nothing to do but try!

You mention TDP equality with your setup but how does that compare with raw temperature output of the CPU that has to be accounted for?

In your thinking you should be able to get a closer comparison to the 3770K vs 5820K wattage load by disabling 2 of your cores it is the same 22nm manufacturing process and I do not run Hyper Threading either, because in some applications it hinders performance instead of helping.

Hyper threading does effect overall CPU temperature that has to be dissipated so does running system memory speeds past your CPUs 2133mhz max memory controller design specifications, I don't know what memory speed you are actually running?

If your cooling is indeed comparable with 2 of your cores disabled you should (CPU luck of the draw of course), be able to overclock the 4 cores to 5ghz, but even if you cannot reach 5ghz go as high as you can for a thermal reservoir comparison?

Why don't you test that for a comparison?

If you run the tests with 2 of your cores disabled it will give a better overall thermal capability to see exactly where your cooling is falling performance wise, it will also help in future understanding of what it takes to cool higher TDP CPUs, and I am definitely interested in that shared data?

I don't have a problem testing that!

I'll disable HT and turn off core's 4 and 5 and leave 0-3 running tonight when I get home, that's actually a really good idea for equalizing the power between the chips.

I'm not sure where the TDP compares to the raw temp output, but this chip is a beastly heat producer. Just as a quick example. With a chilled water temp of 10c°, at an idle at desktop with a 4.4Ghz OC, my CPU temps are still 16-18c°. At an idle with the stock configuration my idle temps are only 11-12c° with a 10° res temp. That's a whole pile of additional heat.

I'm more than happy to do some testing with the setup, I'm really curious to see just how much added cooling power it will take to reach anywhere close to the same performance that you currently have. I don't think that my cooler is quite as good as your's, I just have to many inefficiencies left that drag my temps down, plus the larger TEC's I am running are much closer to their Umax so are less efficient there too, more heat for the same cooling output.

As to my actual CPU setup with my standard OC:
All cores enabled - 4.4Ghz
Hyper Threading - Enabled
Cache OC - 4.2Ghz (with a increased voltage also)
Memory OC's to 2666, the System Agent (or Ring) Voltage was bumped quite a bit also, def not stock.

So there is a quite a bit of extra heat added into my OC aside from core voltage.

I can reach 4.4 at those above settings with 1.25 core voltage. I wasn't able to stabilize 4.5 even at 1.32, but I may be able to with some cores disabled and all the extra stuff off. I'll test the cooling at 4.4 for a straight across comparison to my previous attempts first.

I'm more than happy to share all my results! I can't be the only person running one of these who is interested in below ambient cooling but doesn't want to go through and insulate everything......

 

[4Ryan6]

It will change when you disable 2 of the cores and also drop your memory speed down to 2133mhz which is the highest you can run without overclocking the memory controller on your CPU.

Overclocking the memory controller is no big deal at 4.5ghz or below that point, but when you are approaching 5ghz the additional heat from overclocking the memory controller can cause stability problems, keeping you from actually stabilizing at high multipliers like 50x, and beyond.

If you are willing to test, that will be valuable info!

 

[Ruwed]

Yup, more than happy to test that! I'm curious to see where we end up with this as well.

 

[Ruwed]

Alright.

Some more testing accomplished.

With 2 cores disabled and HT off, the frequency at 4.4Ghz, the chip never really got over 30c°. Here is where the inefficiencies of my system are felt...

I ran the 100% load test for 20 minutes (Real Bench H.264 Video Encoding) and watched the temp rise from 9c° to 12c°. (this is with all the TEC's powered on)
I hooked up a wattage meter to the outlet, the system at an idle drew 139 watts, with 100% load on the CPU it was drawing 200 watts.

I then went back in and enabled the 2 cores and HT, left the frequency at 4.4Ghz, everything else was still stock.
I ran the 100% load test for 10 minutes - The CPU temp reached 38c° and the res temp was at 12c°, up from 10c°.
Wall power was 145 watts at idle and at 100% CPU load was drawing 274 watts.

I'd be curious to see what the actual wattage from the wall on your system is at idle and full load Ryan. Would you happen to a kill-a-watt tester?
http://www.walmart.com/ip/25886580?wmlspartner=wlpa&selectedSellerId=0&adid=22222222227018143581&wl0=&wl1=g&wl2=c&wl3=40880693912&wl4=&wl5=pla&wl6=78810730832&veh=sem

The difference from TDP is so large that it's hardly even fathomable.... I mean really??? Even at a full OC I'm only 9 watts over the TDP, and if you take into account the efficiency loss on a PSU, I'd bet the actual CPU draw is even lower than that!

I need download an actual CPU tester and see if there's a difference... Aidia64 here I come!

aaaand 30 seconds later the verdict is....
Aida64 draws even less than the H.264 encoder in the benchmark in RealBench. 256 watts.

I'm going to go reset the computer to all stock levels again and see what the draw at stock settings are.
And the verdict:
At idle 130 watts. At 100% load using H.264: 190 watts.

I've got to be missing something in my testing... is it possible that I'm getting the super low wattage due to the CPU being cooled by 9c° water?

So, here are a couple of conclusions that I'm drawing.
1. My system is pretty damn inefficient.
2. I can't even hold res temp against 60 watts. No, I'm not sure where the final temp stabilizes at, I stopped the test early, but it was still rising 20 minutes later and at 12c° with all the coolers running. That alone is enough to tell me that the cooler is insufficient for what I intended.
3. The TEC's I have must be running hotter due to the lower Umax on them.
4. I'd really be curious to know what the same meter reads on a OC'd i5 3700k system. The TDP's are obviously worst case scenarios. I can't conceive of any situation where this CPU would be able to draw double the power at 100% load with stock settings. Maybe prime95?

 

[4Ryan6]

I can give you my APC readings but that's also powering the monitor.
Idle @ 330w
Run Intel Burn Test @ 457w
Run Firestrike @ 627w

All were the highest number reached each test.

That's my main machine the peltiers are on their own P/S.

 

[Ruwed]

Yeah, I threw the kill-a-watt just on the tower but I don't see the monitors making much of a difference, they should be running about the same all the time, so you can probably discount them.

Something has to be reading wrong on my meter. There is no way that this CPU is only drawing 60 watts at 100% load at stock. That's less than the 3700's TDP!

I'm not sure what to think at this point. Either I built the worst cooler ever, or something is reading wrong.

Would you be willing to run RealBench? I'm not willing to run intel burn test on my CPU as the instruction set it uses can damage the haswell-e CPU's. edit: I ran it for a bit, it actually uses Less power the the H.264 encoder in RealBench.

Or even Aida64, it's a much smaller download.

Well, one thing that this cooler has definitely shown me. I have a dud of a chip over 4.4Ghz.

I can reach 4.4 @ 1.25v, which isn't bad at all. But I've been trying to get 4.6 stable, I'm all the way up to 1.35 now and still can't stabilize the OC, RealBench keeps shutting down with instability detected. Not real sure I'm willing to dump even more voltage into the chip for a measly 200Mhz, I'm running a bit cooler than my daily OC at this higher voltage, but that's because the system voltage is on auto due to no memory OC.

So now to decide... spend the $25 and risk getting a chip that can't achieve 4.4.... or just say screw it and stick with what I've got.

 

[4Ryan6]

So now to decide... spend the $25 and risk getting a chip that can't achieve 4.4.... or just say screw it and stick with what I've got.

At this point in the technology timeline there's almost zero reasoning to invest any money in any CPU until Skylake arrives on the scene and we see how it performs, good or bad.

 

[4Ryan6]

You could have a bad Kill a Watt meter, test it against a known appliance draw like a toaster or something clearly labeled as to it's power draw?

 

[4Ryan6]

Correction: For those interested in pump longevity I had to edit my post regarding the Swiftech MCP655 that failed, after digging deeper into how long it had been running, it was way past the 2 1/2 years I had estimated, it has actually been in use for over 4 years.

So kudos to Swiftechs version of the Laing D5 design for making a good quality pump.

 

[Ruwed]

I've used the meter against heat guns and it's always seemed pretty close, but that's at 1200w and 1800w tests, I've never tested the low end of the meter.

I'm trying someone elses suggestion today. Basically running 5 TEC's at half power. Don't have any high hopes though, I'm not sure that a 10° heat sink temp difference will be enough to really help much.

And I still have the same issue of the OC on the haswell chip. I know there are chips out there that reach 4.8 at low voltage, but those are the super stars. The average is much much closer to 4.4 at 1.3. So I'm above average for the 4.4 and below average for anything over. In either case, a better cooler won't allow me to overclock without going into potentially dangerous voltage levels for the chip.

 

[4Ryan6]

I'm trying someone elses suggestion today. Basically running 5 TEC's at half power. Don't have any high hopes though, I'm not sure that a 10° heat sink temp difference will be enough to really help much.

How did the half power test go?

 

[4Ryan6]

But using math, apples to apples, you are running 178.5 (based on the specs here: http://www.shop.customthermoelectric.com/12711-9M31-24CW-Thermoelectric-Peltier-Module-12711-9M31-24CW.htm ) watts and I'm running 171 watts per peltier.

Just for clarification purposes I originally purchased my peltiers from FrozenCPU however since FCPU has closed it's doors and is no longer in business, the identical specs of the peltiers I am running are these.

http://www.shop.customthermoelectri...F1A3C.m1plqscsfapp05?keywords=12711-5P31-26CW

 

[Ruwed]

The half power test actually went pretty good (The cooler is currently still running that way as it's so much cooler in my office). Sorry I didn't update yesterday.

My absolute low end is only about 12c°, but I ran a 15 minute stress test at 100% when I very first fired up the computer, so the Res had not had a chance to drop below it's startup temp yet. The Res was reading 21c° when I started.

After about 15 minutes the Res was reading 22-23 (bouncing) and my CPU temp had started at 48 and was now at 52c°.

Not bad for a cooling system that's not adding any noticeable heat to the room!

The final setup ended up not being perfect, but workable.
2 large TEC's both at 6v,
1 large TEC that I ran in parallel with smaller TEC, the large TEC ends up at about 4.3v and the smaller TEC ends up running about 7.5v.
Plus one more small TEC on the 5v line.

It definitely says something about running TEC's at half power. If you can afford to purchase the extra TEC's, even something like an EVO can keep them cool and they add almost no heat output to the room.

I wish I had a better CPU for OC'ing. As I mentioned earlier, I just can't see the justification in keeping the TEC cooler running. It's awesome, works great, definitely keeps my CPU cooler than water did, but honestly, if I can't increase my OC, why would I pay the bill to keep this thing running constantly?

It's my CPU that's holding me back now.

I'm not too worried about skylake, unless they OC to 6-7Ghz, there won't be any need for me to swap from a 5820 to skylake. And even then I would only switch once this CPU has a hard time performing daily tasks. I'm a 5-7 year upgrade kinda guy. I'll upgrade the RAM and GPU's throughout that time, but the main system should last me about 5 years minimum if there isn't a large technology shift between now and then.

Thanks for the info on the Peltiers! I'm considering building a WC TEC box next. It's more of a long term project. If I trade in my CPU. I've learned quite a bit from what I've done this time around and I think I could use it to make an even better cooler. But the money isn't there right now for a project of that magnitude, I may start working on pieces slowly over time. Buy one here, one there and try to come up with a design that will fit into my office, or even possibly put it in the garage and feed some tubing into the office from there, not sure I want the increased temps in the summer though. It's been over 100° here the last month or so.
I dunno, just brainstorming on that one.

For now I'm going to leave the box running for a bit. If I can find the 3rd chinese block, it's possible that I'll add in the 6th TEC just to see what happens. Probably split two smaller TEC's to 6v each and move the larger TEC over to 5v. I think I could get a pretty substantial increase in cooling power and reduce the heat even more if I do that.

If I was running an i5 I think this box would do exactly as I expected and allow me to reach 5Ghz. There is a possibility that I would have to swap the larger TEC's out for something that isn't running right at the Umax to improve efficiency, but even with these TEC's the cooler works.

There is absolutely no doubt in my mind that building a water chiller with TEC's is a completely viable cooling solution, especially if you are trying to push your hardware without going sub-ambient.

Granted, if I would have listened, I could have built this cooler for quite a bit cheaper. But then we wouldn't have had the opportunity to do so many comparisons and testing

 

[4Ryan6]

Thanks for the info on the Peltiers! I'm considering building a WC TEC box next. It's more of a long term project. If I trade in my CPU. I've learned quite a bit from what I've done this time around and I think I could use it to make an even better cooler. But the money isn't there right now for a project of that magnitude, I may start working on pieces slowly over time. Buy one here, one there and try to come up with a design that will fit into my office, or even possibly put it in the garage and feed some tubing into the office from there, not sure I want the increased temps in the summer though. It's been over 100° here the last month or so.
I dunno, just brainstorming on that one.

For now I'm going to leave the box running for a bit. If I can find the 3rd chinese block, it's possible that I'll add in the 6th TEC just to see what happens. Probably split two smaller TEC's to 6v each and move the larger TEC over to 5v. I think I could get a pretty substantial increase in cooling power and reduce the heat even more if I do that.

If I was running an i5 I think this box would do exactly as I expected and allow me to reach 5Ghz. There is a possibility that I would have to swap the larger TEC's out for something that isn't running right at the Umax to improve efficiency, but even with these TEC's the cooler works.

There is absolutely no doubt in my mind that building a water chiller with TEC's is a completely viable cooling solution, especially if you are trying to push your hardware without going sub-ambient.

Granted, if I would have listened, I could have built this cooler for quite a bit cheaper. But then we wouldn't have had the opportunity to do so many comparisons and testing

I appreciate your efforts as in the information you've shared along the way regarding your discoveries is valuable.

Some of that information is backing up what I had already learned the hard way, and some was covering ideas I had pondered but not endeavored to undertake. (Like voltage regulating the peltiers)

The entire purpose of this cooling is to reach higher overclock levels of the CPU that traditional water cooling cannot reach, and be able to run those overclock levels under load 100% stable.

Unfortunately if your CPU just cannot reach those higher levels, why pay the additional electrical cost of the active cooling in the first place as it is a waste of money, that traditional cooling can do much cheaper.

My initial discoveries that led to the creation of this thread started with the initial expense of buying one peltier as the rest of what was needed to complete the first TEC assembly I already had on hand, from past upgrading and traditional cooling experimentation.

From that point it became an obsession of trial and error to discover where it shined and where it fell short and what worked best until finally discovering and excellent combination and assembling a dependable chilled water cooling setup.

Ruwed, You are a very smart individual, I hope you will take what you have learned from your endeavors and move forward with them, once the unknown becomes known, what then?

Ryan

 

[Ruwed]

I appreciate your efforts as in the information you've shared along the way regarding your discoveries is valuable.

Some of that information is backing up what I had already learned the hard way, and some was covering ideas I had pondered but not endeavored to undertake. (Like voltage regulating the peltiers)

The entire purpose of this cooling is to reach higher overclock levels of the CPU that traditional water cooling cannot reach, and be able to run those overclock levels under load 100% stable.

Unfortunately if your CPU just cannot reach those higher levels, why pay the additional electrical cost of the active cooling in the first place as it is a waste of money, that traditional cooling can do much cheaper.

My initial discoveries that led to the creation of this thread started with the initial expense of buying one peltier as the rest of what was needed to complete the first TEC assembly I already had on hand, from past upgrading and traditional cooling experimentation.

From that point it became an obsession of trial and error to discover where it shined and where it fell short and what worked best until finally discovering and excellent combination and assembling a dependable chilled water cooling setup.

Ruwed, You are a very smart individual, I hope you will take what you have learned from your endeavors and move forward with them, once the unknown becomes known, what then?

Ryan

Thanks man!

I'm not sure where I'll go from here, I'm still playing with the TEC's, can't help myself, it's a fun hobby. I may end up trying to get a chip that OC's better than this one does. If I can, well then I'm planning on building another TEC cooler. Maybe I'll go water cooled this time with a couple of large copper blocks custom designed and venturi cooling for the hot side? Heh, crazier things have been done I'm sure.

I think that in order to get this CPU type cooled down to a reasonable level I need a few things.
1. A CPU that can actually OC well at safe voltage levels.......
2. More than likely 4 large TEC's, like 350+ Qmax that I can run at half voltage to increase the COP and keep the dT down.
3. Probably radiators, because you've shown that air coolers work great and I just reinforced that, now it's time to test water :lol: I was looking at the magic 360's, their only $50. I could get three of those and still be into less than I am for all these coolers I'll never manage to get rid of now! (Plus I now have so many fans floating around that I'm not sure what to do with them all....)


I appreciate you posting up your initial idea and successful project, it's what inspired me to test mine. I've seen that PWM control is possible, I was just actually reading about it tonight. It may be of interest to you to try controlling your setup. You could use solid state relays and PWM controller that feeds from your MB, then you would be able to turn on all three TEC's and generate less overall heat under less loads. Have them cranked up at certain temps and turned down when not needed. Would be interesting to see how effective the PWM control is on a setup where we already know that the cooler is cooling more than the CPU is adding! I would sub to that thread and follow avidly for sure!

I'll keep an eye on this thread for sure and pop in from time to time, and I'll even try not to be argumentative with morons who doubt....

 

[4Ryan6]

I ordered my GTX Titan Heatkiller water block from Aquatuning USA, because the water block was cheaper than any US place I had shopped and the shipping was also cheaper.

This was my first experience ordering from Aquatuning USA, and I want to convey how pleased I am with their service!

It seems they really want your US business, from what I gathered Aquatuning USA is like a middle man? possibly? so to speak, because my water block was shipped to me from Germany, and that is Aquatunings main distributorship.

I love a company that stays on top of informing you of whats going on, I received an email from them that the water block I had chosen was out of stock, I chose the next step up in price which was about $2.50US and they had that block and completed and shipped the order.

The block I chose was actually $23.00 cheaper than Performance-PCs the cheapest I found with local US E-tailers, and the shipping to my hands was exactly what they claimed it would be!

So if you are buying in the US and are reluctant to choose Aquatuning USA, forget that, Go for It!

They will be getting more of my business!

That said: The Watercool Heatkiller water block is an engineering work of art, the attention to detail is amazing, it is such a high quality product and it's performance literally runs circles around the Swiftech Komodo that comes on the EVGA Hydro Copper water blocks.

Since the cooling on the backside of the card that covered the memory chips was so inadequate with a back plate, I just used the Watercool RamSinks on the 12 memory chips back there.

it kinda seems that looks have become more important to some of these graphics card companies on the back side of some of these cards than actual raw cooling performance, there is really no comparison to the cooling that was on the card originally till now, when it comes to the memory temperature.

All the memory chips shooting the RamSinks with a laser thermometer were running at 22c except the 2 closest to the PCI-E slot which were 24c.

At least now I know what's going on, temperature wise, on the back side of the card.

So all is well and running great!

With the Heatkiller water block, benchmark load temperatures so far have never exceeded 28c and it idles at 20c with a 26c ambient, and that is because the radiator cooling the GPU is 2 1/2 feet away from my room AC.

And that radiator is the Mighty MO, the Watercool MO-RA3.

Edit: Under a full gaming load for about 2 hours of play, it did reach 30c for about 24 seconds according to GPU-Z.

 

[tech-wreck]

are phobya blocks any good? aquatuning sell a full copper cpu waterblock for €35 delivered. atm that is the only part i may have to replace - i have a feeling the mounting holes in the xps cpu plate will be very proprietary.

edit: unless i can make it fit with power tools and one of these http://www.aquatuning.de/water-cooling/cpu-water-blocks/cpu-mountinig/10718/aquacomputer-umruestsatz-halterung-sockel-1366/1156/1155/1150/775-fuer-cuplex-kryos-delrin?c=493

 

[4Ryan6]

are phobya blocks any good? aquatuning sell a full copper cpu waterblock for €35 delivered. atm that is the only part i may have to replace - i have a feeling the mounting holes in the xps cpu plate will be very proprietary.

edit: unless i can make it fit with power tools and one of these http://www.aquatuning.de/water-cooling/cpu-water-blocks/cpu-mountinig/10718/aquacomputer-umruestsatz-halterung-sockel-1366/1156/1155/1150/775-fuer-cuplex-kryos-delrin?c=493

Phobya blocks are not the highest cooling performing of the CPU water blocks, but they're still good water blocks.

The link you posted did not seem specific from everything on the page so it was not clear what you were referring to?

 

[tech-wreck]

the xps cpu waterblock is fixed to a rectangular piece of aluminium which was mounted on the self-contained cooling unit, not directly to the processor. i may be able to 'reshape' the ally mounting bracket and use a kit like that to fix it to the motherboard. if not I'll have to murder the piggy bank and go shopping

 

[4Ryan6]

Are you attempting to affix that entire unit you pictured to your CPU of today, is that what you are contemplating?

 

[tech-wreck]

erm, sort of? apart from new tubing and maybe a CPU block, i will be reusing the parts in the picture.
i have a 2500k in a z68, but I'm reluctant to mess with them... that pc has been working flawlessly for over a year, and the case is another unique custom job that i don't want to re-paint.

my (current) plan is to acquire some old hardware, like 775 old and modded to take a xeon, and test the cooling setup as dell designed it. then rebuild the system with an insulated reservoir instead of a radiator. and possibly toying with a couple of other ideas along the way... i read somewhere the C0 stepping X5460s are good overclockers, and a quarter of the price of a top end C2Q.
if that is successful, and by that i mean not a resounding fail, i will be looking for a Z87 or Z97 and a processor capable of some serious clocks.

 

[CompuTronix]

Hi Guys,

I just wanted to pop into this thread for a moment to let everyone know that I finally had the pleasure of meeting Ryan in person this past week, and seeing his rig.

We conducted a very detailed covers-off inspection of the entire chilled water system, followed by running it through the paces. This included an assortment of utilities such as Intel Burn Test and Prime95, as well as gaming. During testing, I closely monitored his software temperatures and hardware instrumentation. The thermal balance, stability and performance of the chiller system, CPU / GPU loops and computer system was flawless.

As the author of the Intel Temperature Guide here at Tom's, and having worked many years on liquid helium chilled superconductive magnets as an MRI engineer, I'm uniquely qualified to comment on Ryan's achievements.

I'd like to say for the record, that as a perfectionist in my own right, I was thoroughly impressed by his approach to below ambient cooling, documentation, depth of knowledge, and especially the quality of workmanship and attention to detail.

Excellent work, Ry! Thank you for the invitation. I enjoyed the technical aspects of our visit, as well as the "Southern hospitality" shown by you and your wife. Hopefully on my next visit we can spend more time gaming on that 5.0GHz rig, rather than testing it!

Best regards,

CompuTronix

 

[4Ryan6]

@ CompuTronix

Thank You Sir, For Those Kind Words!

It was a Pleasure and Honor, to have you visit our home!

Ry

 

[4Ryan6]

@ Computronix

Since your visit is still fresh in your mind I would like to ask you a question to confirm a test I ran for you, if you will.

To be specific the coolant temperature was at 12.3c when I energized the 3rd TEC assembly, the 3rd TEC assembly does not add cold to the reservoir instantly it takes a couple of minutes before it begins adding cold to the reservoir you saw on the temperature reader when the coolant temperature began dropping.

I opened Intel Burn Test at that point and began running it, and my question to you is:

A: Did the coolant temperature increase while IBT was running?
B: Did the coolant temperature stay the same while IBT was running?
C: Did the coolant temperature continue to decrease while IBT was running?

Do not worry this is the only question I will ask you. Maybe?

It is just an eye witness confirmation for someone else benefit.

Thank You! Ry