CoolIT Boreas

[Serpentarius]

actually CoolIT Boreas is a great idea ..... easy chiller installation .. hassel-free

but there's aren't any additional fins inside the wind tunnel (+rad) .. and it's cooling method is quite (un)efficent? the heat is not directly exitted ... there's lotsa gaps (turbulance?)
plus .... according to the specs ............... Heatsink: Anodized alloy, radial intake array (what alloy????)
not to mentioned the Chiller waterblock specs: Dual anodized alloy distribution, multi-channel (what alloy again??? assumed: copper + alu = the most common duo)

[NaeKuh]

I have to ask a FEW Questions now ..
Did martin make that chiller for you ?
Martin said the person he made it for was planning to uses it inline with a radiator? (i fort hard against this because it's fundamentally floor'd )
What kind of rads do you have ..
what TEC's do you have .
What load are cooling .
What happened when you used it last time

Martins design isn't that great but well made though mine is pretty bad too

Well, the rad eventually got pulled, because as you said, it did floor the performance...

Nol was testing the unit last, and said it had some problems holding a 300W load tester... Then the TEC's died... so i dont know if it was just overall bad TEC's or the 5 x 90W tec's martin installed just wasnt up to the power...

Last i heard Nol replaced the TEC's to 4 62mm TEC's, but he didnt get around to testing.. I should have the unit back soon tho and wanted to retest this.. But then again teyber is also finishing up on a chiller for me...

But if you can see a sensible way where its not gonna gobble 300W I would love to retry TEC's again on that chiller..

and yeah i removed that pic, because its a boras post, didnt want to flood it... also i was reading your graph more in detail and got somewhat confused....
Uncle Jimbo was half working on a thermostat to possibly cycle the tec's... but i dont know what happened to that either.. :P

[Spoiler]

yes and no. Chillers have there place in the future but the Boreas is rubbish it's fundamental problem is it's trying to achieve better than watercooling performance with effectively with one 120mm fan That's pretty much can't happen expect for when used with very light loads. But

The hot side of the boreas must be getting to hot to achieve descent cooling

120 watts of electricity is asking way to much from such a poor design and turning this up would just result in even worse hot side temps. OS with a highly overclocked CPU i think the Boreas is fundamentally floored by having to poor a hot side cooling

I am not quite sure I understand everything you have said, but shouldn't we wait for an in depth review before calling anything rubbish? Also, aren't there two 120mm fans exhuasting the heat?

I really don't know that much about TEC's, but like many other people, I am always on the lookout for easy but Extreme cooling. If this unit can prove to me that it can cool an i7 better than standalone water cooling...I'm interested.

[Ultrasonic2]

I am not quite sure I understand everything you have said, but shouldn't we wait for an in depth review before calling anything rubbish? Also, aren't there two 120mm fans exhuasting the heat?

I really don't know that much about TEC's, but like many other people, I am always on the lookout for easy but Extreme cooling. If this unit can prove to me that it can cool an i7 better than standalone water cooling...I'm interested.

Yes it does have 2 fans .. but two in series doesn't = twice as much flow

How much does the Bores cost anyway?

[aspire.comptech]

Yes it does have 2 fans .. but two in series doesn't = twice as much flow

How much does the Bores cost anyway?

Something like $550

Which is $150 more than the OCZ Phase unit.

[Ultrasonic2]

Well, the rad eventually got pulled, because as you said, it did floor the performance...

Nol was testing the unit last, and said it had some problems holding a 300W load tester... Then the TEC's died... so i dont know if it was just overall bad TEC's or the 5 x 90W tec's martin installed just wasnt up to the power...

Last i heard Nol replaced the TEC's to 4 62mm TEC's, but he didnt get around to testing.. I should have the unit back soon tho and wanted to retest this.. But then again teyber is also finishing up on a chiller for me...

But if you can see a sensible way where its not gonna gobble 300W I would love to retry TEC's again on that chiller..

and yeah i removed that pic, because its a boras post, didnt want to flood it... also i was reading your graph more in detail and got somewhat confused....
Uncle Jimbo was half working on a thermostat to possibly cycle the tec's... but i dont know what happened to that either.. :P

i wouldn't think 5x90 would be enough Qmax

What i would say is matrin drilled the holes in the chiller to take the TECs of X size so changing them might not be as good. but it probably doesn't matter to much. If you do get it back you must make sure someone hasn't stuffed it by bowing the bottom of it.

What was confusing about the graph ?

Oh i cheated on the Thermostat for the TEC's. im running a home made 600watt PWM now. it adjusts the current going to the tec's to keep my CPU at 25c irrelevant of load. Why 25? well i dont have to worry about condensation. it does load higher than 25c though. currently my fan speed it 38% and my TECs are at 44% (240ish watts)

By the way im not just some random guy i normally hang out at overclock.net but i've been stuck at work with not much to do so i thought id come over and see what you guys are upto

[NaeKuh]

ultra i know your not a random guy...

if you were id be a lot meaner to you..

i can tell you know your stuff, which is why im asking for help...

seems like you know how to get this unit back working which is my main hope..

So check the blocks for warping... okey, then what tec's you recomend me load on that unit? go with the original 4x62mm tec? or keep it on the 5x50mm tec?

[Ultrasonic2]

Well unfortunately the answer to your question is not simple (as you guessed) Really i need to know the answers to the questions i asked before. You see the answer is NOT as simple as saying more Qmax the better because that's NOT the case. This is because of the relation ship between Qmax Umax dTmax which i was trying to explain before. Well before i go any further i should qualify my statement that "More Qmax isn't always better" more Qmax is better as long as you disregard the electricity required to achieve X temp. Now in the real world the electricity consumption is very important. So we need to find the what i call the "Sweet Spot" this is the point where adding anymore or less Qmax would result in more electricity consumption or unacceptable temps. It is also the same point where adding more input volts relative to Umax will result in more electricity consumption or unacceptable temps

Now before i try my best to explain this i've already done that. you can read it on my site here
http://watercooling.co.nz/node/134

[HDCHOPPER]

thanks for the read ultra

[NaeKuh]

well another theory on the unit was to stack 10 tec's..

since the hot side is watercooled, the heat moved wouldnt be an issue, it would be an issue on the rad size tho, but now a days, with TFC monsta rads, i dont think the system could possibly ever be cold side limited...

this thread is very good, althought its on boras and we derailed it, a lot of good info on TEC's came out...

[Ultrasonic2]

Nah you pretty much never want to stack TECs. When you stack them on top of each other your multiplying the inefficiencies.

yeah the chat has been fun. Im off for a few days guys so have fun

[RealTelstar]

waiting for review

[NeedMoMegaHurtZ]

waiting for review

HWC (hardwarecanucks.com) had one a while ago...

EDIT: site seems down at the moment, but just dig around to find it when it's back up

Cheers

[Ultrasonic2]

pitty theres not True or decent W/C

[Uncle Jimbo]

Uncle Jimbo was half working on a thermostat to possibly cycle the tec's... but i dont know what happened to that either.. :P

I finished the PWM thermal TEC controller design a ways back - a couple of people have built one. I didn't get around to laying out a circuit board though.

A TEC can achieve more cooling than power in by picking an operating point that has CoP better than one. The graph below shows what range of Imax gets those results. I used a 10A Imax because it's easy to figure % Imax from the chart - results are similar (at a given temp) for most TECs.

For a 10C dT, you can get CoP of 3.4 at 12.5% of Imax - that means you get 340W of cooling for 100W in. Problem is you don't get much cooling at that low current so you need a lot of TECs.

Most commercial designs aim for CoP of 1 to 1.4 at 20C dT (30% to 50% Imax). That is more useful and still efficient, and you get significantly more cooling power.

[Uncle Jimbo]

ultra i know your not a random guy...

if you were id be a lot meaner to you..

i can tell you know your stuff, which is why im asking for help...

seems like you know how to get this unit back working which is my main hope..

So check the blocks for warping... okey, then what tec's you recomend me load on that unit? go with the original 4x62mm tec? or keep it on the 5x50mm tec?

There are other issues (such as getting the right torque on the TECs) - but if 4 62mm TECs will fit, that is less expensive than 5 50mm. Since the problem with the unit is hot side, we want to be at an efficient point - no more than 50% Imax. For the common '545' eBay TEC, that would be 8V or less. Actually, putting two in series on a 12V supply gives a nice operating point - 40% of Imax and a CoP of 1.4 at 20C dT. That will produce a total heat load of 120W per TEC, or 480W for the whole rig. Power in is about 50W per TEC, or 200W for 4 of them. You move 280W across 20C dT.

[eligray]

I can't believe those results!

[masska]

need to see more. that graph posted by NeedMoMegaHurtZ is hard to believe.

[masska]

not much of a review but here it is,
http://www.maximumpc.com/article/coolit_boreas

[RealTelstar]

I can't believe those results!

They are from hardware canucks, in this page:
http://www.hardwarecanucks.com/forum...review-16.html
(and next page)

I believe they are true. The Freezone Elite is in line with a decent watercooling setup, while the Boreas does really good. And at $500 (without case) it should!

[Anemone]

Would have helped if Canucks had just included a simple 240 or 360 cooling and one air, like a TRUE. It would have been a useful set of reference points. Their review still seems the most useful so far.

[RealTelstar]

Would have helped if Canucks had just included a simple 240 or 360 cooling and one air, like a TRUE. It would have been a useful set of reference points. Their review still seems the most useful so far.

True
Also they dont mention the ambient temp, although i dont think is high.

[demonkevy666]

this looks like 4 water blocks on an air coolers and 5 tec's one on the cpu and 4 on the air heat sink.

better blocks and an air heat sinks.

wonder what adding a radiator would do for a set like that.

[bldegle2]

the boreas is a bit expensive.....with minimum mods I got a $129 Original Freezone with the annoying 92mm stock fan removed and Dangergen 120mm xflo running a Phenom II @3.8..............

25*c idle, 45*c full load, 1.44v...the full load temps flat line, before adding the radiator the unit had problems containing the heat as it was designed for a single core.....

the secret is the pump, it must be low flow (3.5L/m) so the unit has a chance to supercool the water before it returns to the block....with quiet 120's on the unit and radiator it is much quieter than air for clocking...i don't understand why the radiator didn't help with the Boreas....

simple loop, warm cpu water hits radiator to get precooled, gets supercooled in the unit, and onward.....jeesh, if Freezone sold these as a kit with the radiator.....

i run the system without the MTEC controller, just an original controller with pentometer cranked to 100%...

properly tuned, these units work flawlessly, mine is going on its third year of operation on various setups....the radiator was added just recently, wish I had done it sooner.....when I first got the 940 PII the unit by itself could not contain the full load temps @higher vcores, radiator makes all the difference...

Freezone runs the pelts at about 5v.......

clock on...

laterzzzz..............

[SteveRo]

the boreas is a bit expensive.....with minimum mods I got a $129 Original Freezone with the annoying 92mm stock fan removed and Dangergen 120mm xflo running a Phenom II @3.8..............

25*c idle, 45*c full load, 1.44v...the full load temps flat line, before adding the radiator the unit had problems containing the heat as it was designed for a single core.....

the secret is the pump, it must be low flow (3.5L/m) so the unit has a chance to supercool the water before it returns to the block....with quiet 120's on the unit and radiator it is much quieter than air for clocking...i don't understand why the radiator didn't help....

simple loop, warm cpu water hits radiator to get precooled, gets supercooled in the unit, and onward.....jeesh, if Freezone sold these as a kit with the radiator.....

i run the system without the MTEC controller, just an original controller with pentometer cranked to 100%...

properly tuned, these units work flawlessly, mine is going on its third year of operation on various setups....the radiator was added just recently, wish I had done it sooner.....when I first got the 940 PII the unit by itself could not contain the full load temps @higher vcores, radiator makes all the difference...

Freezone runs the pelts at about 5v.......

clock on...

laterzzzz..............

what are you using for a pump - I am thinking about a boreas and have the Swiftech 355 with XSPC top - might be to much flow?
Did you try your setup with high flow and low flow?