Thanks guys,

Ordering some 1/4" fittings for the next round of testing. Adding legit pump and res into the loop ........ :)

Nice temps on water!

Well I finally got my 1/2" x 1/4" fittings

Tomight I introduced a res and 655 B pump that would be the same as a D5 multi speed pump on setting 4

Using the Y & S fans at an ambient of 20.5C

1.408v 4889MHZ (65) (67) (65) (55) = 63C

being it is a half of a degree celsius and the total package only dropped by .75C I have to say this would be with in the margin of error, so no benefit here ......... but wait there's more, for some reason that I have not figured out yet it has raised my idol temps 2C and is still giving me the same load temps ....

This is actually a good thing as I am having no condensation what so ever, So the res and pump combo in effect has had a positive impact on the overall validity of using this ina 24/7 rig

http://www.overclockaholics.com/foru...1&d=1360382129



http://www.overclockaholics.com/foru...1&d=1360382129

And although not applicable to the testing the pump res has made a hair shy of 5GHz possible even with this POS chip, temps arn't bad considering the volts. On indoor ambient water it would be a melt down .........

http://www.overclockaholics.com/foru...1&d=1360382694

The next and final step of this preliminary testing will be to introduce a Fesser Quad Exchanger into the loop. I am thinkiing it will increase my idol temps yet again, this does not bother me as I hope it will contain the wicked curve I get under load. Hoping for a few more degree's .......... :clapping:

With a decent chip who knows may be able to get 5.0 Ghz 24/7 stable :)

Are you water cooling the hot side of the TECs or the cold side?

The cold side and the hot side is disapated into the aluminum coil and the fans pull ambient throught the coil

wouldnt it be better to cool the hot side ?

Thats how TEC cooling usually works, because Radiators add temps to subambient cooling however he is blowing past the TEC wattage capabilites so i warming the TECs in exchange for adding maximum cooling capability.

Its why the boreas only sold 2 units lol.

Here it is, the good bad and the ugly ....

Adding a rad to this system turned out to be an epic fail. Adding the ambient to the loop did raise the idol temps to a comfortable level that would in fact never condense.
However the Boreas unit simply does not have to capacity to cool the added heat that is added to the loop

Ambient = 20C

1.408v 4889MHZ (75) (79) (74) (67) = 73.75C

http://www.overclockaholics.com/foru...1&d=1360504583

http://www.overclockaholics.com/foru...1&d=1360504583

While these results were for told by Rick I could have saved a lot of work by listening ........... :laughing: But I had to find out for myself and had some fun doing it and was able to bring pretty strong results being the test conditions were controlled as close as I possible could to obtain viable results

Here is a table showing all the results from the test in one place

http://www.overclockaholics.com/foru...1&d=1360504583

While this is not the typical here at OCA as we are usually talking about chillers, Phase, dice or Ln2 temps this unit seems to be pretty much a joke in those terms. However I am not trying to compare this unit in those terms, but in terms of a water cooled 24/7 rigs

While it is undisputable this unit does have limited capacity, It is also undisputable that it is crushing ambient water to the tune of 15C, in those terms this unit is nothing less than impressive as all of us have played with water and fought over a few degree's sometime or other. So while I was Hoping for a 20C to 25C advantage over ambient water I find it hard to be disappointed with the results

To sum it the final spec going into my upcoming build will be small res w/D5pump. I will be using a higher quality high flow block to maximize flow in hopes to grab a degree or two. I will be surrendering a few degrees as I am opting to use the Yates Silent fans as the trade off is worth it for a 24/7 rig IMO

Thanks for reading ....... :thumbsup:

So, what are your plans on testing this on the cards Brian? I really would like to see how the Boreas can hold up there.:)

Going to give one of these a spin on the next project, but will likely have to move from my mid-towers to full in order to fit it.

http://www.newegg.com/Product/Produc...82E16835709001

I have decided to use a dual loop. I just do not think this unit is up to the job if I add GPU's into the loop, I am going to use this to maximize CPU clocks as I think I will reap the best results, The thinking is with 2 over cards at 230 watts a piece at stock clocks I will be looking 460 watts and if I go tri 690 watts ... that's stock ......... :ohcrap:

Overclocked it could get ugly .......... :blink:

My thinking is I would have to water couple this unit and even then it could struggle. That plus would probably have to buy higher capacity TEC's. I am speculating but I am thinking possible only a 5C advantage over ambient, that being said I have the room for 3x 480 rads to cool them so one to argue I will have ambient water coming back to them.

So looking at the TEC to cool the CPU and typical ambient water cooling for the GPU's ...Mind you that the typical water cooling will be over the top

I would have like to see as well, but thinking it is a waste of time and money as I would have to buy blocks that would not be used for the build just to find out where I stand and with the CPU testing results I am convinced this unit would be crushed by a tri SLI set up. Single card set up I do believe would be a worthwhile endeavor, I think this unit would bring some great results under those circumstances.

Probably a wise move. I long ago relegated the Boreas to bench duties as it just shines there...good performance and easy to make quick changes cross platform.

Most likely I will stick with high-end air coolers like the one linked above for my gamers...much less to deal with and can still run decent overclocks if desired. Latest video cards are beastly on their own with stock cooling. :good:

From the test results I am excited to get started on the build, it will defenatly be way overkill for simple 24/7 gaming, but that is what this upcoming build is all about, Overkill, excess and cool looking are the goals

basically building it for the fun of it all ........ :laughing:

Benching has become an LN2 only sport , at least if you want to compete at a high level

I would like to see you build a furniture grade such as this one... http://www.tomshardware.com/forum/27...om-wooden-case I know you have the skills...I would like one myself. :clapping:

Funny you should say that, I compenplated doing a built in. Ultimantly I decided against it as I thought upgrading would a major PITA ... and TBH do not have the milling tools to do a top flight job. Although not sure if Bren still does cabinetry? he possees such tools and skill set required to build furniture grade from the ground up.

The best I could do would to be to by a mice piece and mod it from there

But a simple granit toped desk with a cut out and a glass insert is all you would need for a great start.

Yep... Air don't cut the mustard with these Tecs.

Keep on modding Witch. The more you play, the more you'll understand. ;)

Remove min 2 watts of heat for every unit chilled and you'll have some awesome temps.

I can tell you that the cooler the hot side loop gets the better. Don't be afraid of trying Ice in a bucket on that loop to get them Peltiers well into negative temps on the chiller side.

Ever need anything with this TEC stuff for suggestions bro, please don't be afraid to ask.

EDIT:

In fact, I have a 130w (chilling power) TEC. All I need is an addy, and I can send it out. I'd like you to experiment with it on the side. Use different voltages and what not.

Not sure if starting a little smaller would help any with a TEC that can actually cool a Cpu by it's self. up to you, I have a PM box here too ;)

Thanks Shrimp,

But as I have been saying this will be a 24/7 rig that is contained with in a case's confines, Although I wanted better temps as stated 15C less than ambient water is nothing to scoff at on a 24/7 rig

Buckets of ice, seperate chillers, cold boxes phase, dice, LN2 or simply throwing a 480 out the window all would be viable options for better temps

These methods while effective are not what I am looking for. It all has to fit in a case that you can pick up and carry away. This is for a custom build gamer. at 15C under it's counterparts in the catagory should provide epic results

Pretty much rolling with what i have at this point, I was thinking about your idea of using two of the blocks for cooling them with water but while I am convince this is an epic idea I am alsoi convivced this would condense as I am on the verge of condesing using air. This would be an usatisfactory for the parameters of the build

When I want to bench seriously I would be straping the pot on the bencher anyway

Any one can throw a pot on something. That's just sooo easy. Dry Ice right at the walmart even.....

Cpu Block - Rad - Res - Pump - Chiller blocks - back to Cpu block.

Tec hot side

TEC block Rad - Rad -Rad (push pull fans) -res pump - back to tec block.

The hot side TEC blocks will get hot. To optimize heat dissipation, put fans right on those hot ass water blocks.

Expect needing negative Cpu liquid temps in order to perform 40c loads on 130w roughly. -15 to -25 ought to do it. Condensation depends on hose thickness. The thicker the better. Contain the chilled liquid will work best.

Some minor sealing around the Cpu heat sink couldn't hurt for just in case. I don't think you'll get frost on and through the board in my honest opinion. If you move enough air flow around the water block and hoses, you might not see condensation at all.

I hear ya bro.

BTW,

Nice ride you are sporting these days .............. :thumbsup:

Couldn't beat the 229 price tag at MicroCenter. A little bit of a drive but worth it IMO.

I should try TEC cooling it? (seriously) I have a peilter that'll handle this at stock. This 3770K actually runs pretty cool at stock on regular water. I was thinking of de-lidding it, but I have troubles enough with the chip past 4.8ghz and I'd hate to run 1.5+ volts just for 5ghz and benchmarking. I know I don't have a single TEC that could handle it alone past probably 1.250v give or take. At least not a 40mm TEC.

Another neat thing I've wanted to try is cold plate heat sinks. TECs need a lot of surface area, especially to be cooled. They are constant heat pumps. http://www.aavid.com/product-group/liquidcoldplates

What are your temps and core for 4.8 ?

1.380v min and 1.410 seems to keep stability better. Temps are in the mid to late 60s core #1 likes to spike the hardest to 72c - settle in temps are in the 60s mid to high short tested 2 hours OCCT and Prime95.

Anything more to the above I can run the chiller up to 1.5v for stability at 5.1ghz. haven't seen 5.2ghz without a blue screen.

Most settings you gave me for the Gene applied and work pretty good for the 5ghz. I did benchamrk Pimod, nothing much more than that.

With HT shut off, 5ghz is in the 60's and with only 2 cores I can get it in the 50's Spikes in the late 60s.

But 5ghz daily stable won't happen with this one unfortunately.

BLCK clock made a max of 107 postable 106 in windows on 35x multi.

additional commentary:

The GTX-480 on the same loop, it clocks 940/2050 pretty much max on the first PCB for the card. It humps along pretty good. If I removed this from the loop, the Cpu temps "should" drop dramatically. I just don't have enough stuff to make 2 loops. Space really isn't a problem....

Nice results ..... :thumbsup:

whats your ambient ......... ?

Temps seem very low :)

Ambient is 19c - 22c - Dual 120.2 rads - modded ThermalTake waterblock from BigWater series - Using s775 mounting plates (the board had holes for it! Tubes are 1/2" OD 3/8" ID Nylon clear liquid is a mix of ammonia and alcohol and Thermal take green and CoolerMaster neon reactive Blue and some windshield washer fluid.

So at stock keeping the i7 at around 45c (loaded) give or take unchilled is no problem. I still got to get F@H going for some long hauling. I did let it cruise at 4.2ghz on 1.288v over night. I didn't complete any packets, but it seems to fold the same as my AMD just has more threads to do it with. Surely no doubt it's faster, but the work units are different so I won't know any numbers until I get that itch again.

Cute

Mmonia and alcohole, I have read about but never seen anyone do it, better than distiilled and nuke ?

I think you will like the Intel's , there defenatly the big blocks of the CPU world. Being I don't know what platform I will be building on it is hard to say what clocks I will end up at. But 5GHz 24/7 would be cool, but definatly overkill for most tasks. Who knows if work does not pick up soon I may just be putting my current SB-E platform in it with some new GPU's and possibly a better board.

I would like to see what Haswell brings to the table and if Intel goes the way of TIM or soldering the IHS. hopefully the latter

I'd hope for solder myself... better heat transfer.

Told ya I like my diamond paste. Just takes a little bit to settle in ;)

from my testing TX2 seems to work best for ambient ...... :thumbsup:

But did not test the stuff you are talking about

Antec Formula 6 and they did make a formula 7. I found the 6 to be almost as good as the 7 at a cheaper price.

Most Thermal pastes work well if just the right amount is used. But I'm sure you know the ropes plenty well. TX2 is great stuff. Probably works better than what I'm using by a deg or so. I'm not that picky. I'll just go adjust my thermostat on the wall :D

Well it is back to the drwing board, the unit did not fit in the case, Bone head move on my part. need a few weeks and will be working on water to water..........

Thanks for the idea Shrimp, we will see if half the capacity with better disapation can have a positive result.

Just have to figure out how they are staged

diamond paste is a bitch to work with, and since it needs to really be super heated to apply it well (I stuck it in front of an electric heater for 15 minutes on time and it spread nicely) I would guess it is not very effective (CPUs just dont get that hot).

Not to mention that two applications and you can no longer read the IHS

Super heated?

Well AS5 also needs a burn in too. Was a very popular thermal paste.

The Diamond paste needs to be used in the proper amount. Not too much not too little type of thing.

To each his own - I've had good luck with house hold items too!

As5 burin I never found to be true. Always ran better wet than dry.

As for superheated I was referring to application for diamond paste.

I would love for that stuff to be great, I got enough of it for free and the idea behind it is awesome. It just sucks though, I really cant say enough bad stuff about it for the average user. It really sucks and everyone should stay away from using it. thats my opinion, I have tried it for a while over the last few years (various IC iterations) Its is finger breaking to get it out of the tube, there is no spreading it, it erases your CPU and like every other thermal pastes it makes no differnce in temperatures.

but whatever is easiest to spread. if you change yuor heatinsk a whole lot? Use ketchup or toothpaste. (LOL JK on the last bit, just saiyng AS5 or other easy spread is fine)

I use glied oc extreme and tried the free sample oc icd 24 diamond paste was just a few degrees cooler on idle and load but that stuff stains cooper it seems and fades the letters on the ihs and smells when you ran it for a while when removing the cpu block

I like TX2 for ambient and the Gield OCE for cold

That is the program I am on

TX2 did 3C better than Ceramic on the SS

Now that could either be it is better or I had a better mount when I did the TX2 as opposed to the Ceramic

sorta pissing in the wind with TIM, think every one has their favorite and I believe everyone has measured differences, but as above the mount itself is the deal breaker IMO not nessesarily the TIM ...

I know the white cap OCZ Freeze did a good job and they changed to a blue cap and it blew ........ :Dizzy:

Wow you got a sample tube and based an opinion on all diamond paste for it? Or you got a specific tube of Antec F6 or F7 and had a bad experience?

I had no better luck with AS5 than I have with using different types of never seize. Wouldn't recommend it to any one. Or any paste similar to it.

Not sure.... When I delided the 940BE, I saw a massive temp drop putting the core right on the waterblock. It worked better than the solder dropping temps some 10c or so. Can't remember that one, been a while. But I do know once you delid something, the only time the IHS plate goes back on is for sale ;)

@ Witchdoctor.... TIM on sub zero temps is good if it doesn't freeze. Personally, I could care less if the pot freezes to the board, I've used water under my LN/DICE pot on a few occasions when tim was running low.

Fun stuff, yes arguing TIM and performance is a moot point among extreme cooling individuals......