Swiftech MCW7000-T for Corei7 announced

[Ultrasonic2]

duckbillclinton, I suggest you read a bit more on TEC's before making claims you are unable to back up. A 360w Qmax TEC requires nearly 600w of power to run at 360 Qmax. Assuming the voltage is lower than Umax then you no longer get the full 360w Qmax. As muffy stated these results just don't make any sense.

Muffy is me by the way . i got the nick name from eating a muffin everyday for 5 years


BY THE WAY ChillerZero wasn't using a 62mm TEC he was using a 50 mm TEC ... hes said he was using a 226 watts of QMAX TEC. A 62mm TEC wouldn't have fitted his block anyway

yes you will go sub-zero.

i go subzero with only 226w of TEC on Core i7 -10°C idle @ 4180Mhz HT on, load +28C.
Now i ordered a 300W TEC for it, and i will testy it on this one.

I assume we're talking about this post cos. This is theoretically imposable even in a best case scenario .. You have a CPU generating 230watts of heat being cold by a 226 watt TEC and your loads are 28c .. that it theoretically impossible.

So you must be forgetting to tell us some thing OR an i7 at that Vcore and frequency doesn't generate 230 watts of heat ?

It' all very fishy to me

[duckbillclinton]

duckbillclinton, I suggest you read a bit more on TEC's before making claims you are unable to back up. A 360w Qmax TEC requires nearly 600w of power to run at 360 Qmax. Assuming the voltage is lower than Umax then you no longer get the full 360w Qmax. As muffy stated these results just don't make any sense.

Hey, bud, I did say on my post a 360W TEC will need 600W PSU. What I tried to say was we don't need 700W PSU, lol, since a 600W can overload a bit (+- 5 to 15% depending on spec). I do undertand lots TEC rated 24V is indeed 24.1V, and a 360W Qmax TEC will need 360/ 0.6 = 600W power. I have been working in far east for 3.5 years now, and yes, before I posted here, I made 2 phone calls to 2 TEC factories here and talked to their engineers (the sales people routed my calls since they can't handle my questions).

Please don't get me wrong, dude. You are the one that is cool enough to know one of the key factor for TEC cooling is power supply, but you need to really read through my posts, lol.

However, one thing that puzzles me on this forum is that even ChillerZero is so hardcore on extreme cooling (with his 20+ pages work log covering modded SwiftTec TEC cooling), lots still don't believe him, even there's no reasons he should scam on the screenies. In fact, i think both of us do know ChillerZero's biggest problem on his setup is him having a gimpy 320W PSU.

By the way, I do have a question for you though, does Qmax change base on temperature? How come you mentioned 437W Qmax (50c), and 360W Qmax (27C)? If the temperature does play a factor, isn't that suppose to go the other way around?

Peace,
DBC

[duckbillclinton]

Im still surprised that no one credible has tried a 62mm TEC on a i7 yet ?

In theory ChillerZero's results are impossible. But thats based on the CPU load calculators being correct. Which they may not be. So for me im still waiting for someone else to give it a try. I've got a 62mm TEC and water block pitty i dont have a i7 id be happy to sell it too

lol, yes, the calculator you used is off. I had the same doubts as you at first, and that's why i went through all the trouble to dig through Intel's tech papers and ATX 2.2+ documents. With ChillerZero's gimpy 320W PSU, the most Qmax his TEC could do is 192W. So, it is actually more shocking to know a 192W Qmax tech could actually manage the heat produced from 4.0 ghz OC with full load.

DBC

[plasmatique]

Swiftech MCW7000-T the waterblock phantome

[Ultrasonic2]

.

By the way, I do have a question for you though, does Qmax change base on temperature? How come you mentioned 437W Qmax (50c), and 360W Qmax (27C)? If the temperature does play a factor, isn't that suppose to go the other way around?

Peace,
DBC

Qmax dTmax and all the rest increase as the hot side temp increases. However they dont increase faster or at the same rate as the hot side is increased. This means your always better off with a colder hot side temp

[Ultrasonic2]

lol, yes, the calculator you used is off. I had the same doubts as you at first, and that's why i went through all the trouble to dig through Intel's tech papers and ATX 2.2+ documents. With ChillerZero's gimpy 320W PSU, the most Qmax his TEC could do is 192W. So, it is actually more shocking to know a 192W Qmax tech could actually manage the heat produced from 4.0 ghz OC with full load.

DBC

This is why people have trouble believing him. We're not saying he's lying it just seems a bit strange and would like someone else to verify his findings

[duckbillclinton]

Qmax dTmax and all the rest increase as the hot side temp increases. However they dont increase faster or at the same rate as the hot side is increased. This means your always better off with a colder hot side temp

isn't this suppose to go the other way around, higher ambient temp (hot side) cause the Qmax and dTmax to decrease (more inefficient given the TEC is hotter)

DBC

[Ultrasonic2]

isn't this suppose to go the other way around, higher ambient temp (hot side) cause the Qmax and dTmax to decrease (more inefficient given the TEC is hotter)

DBC

Simple answer is NO

Complicated answer is NO


What make you say this ?

[Ultrasonic2]

There's no doubt that a TEC will use more electricity to cool a load to 20c if it's hot side is 60 over 20 but that doesn't make our statements wrong.

[duckbillclinton]

There's no doubt that a TEC will use more electricity to cool a load to 20c if it's hot side is 60 over 20 but that doesn't make our statements wrong.

Na, dude, what I meant is besides the simple and complicate answer of NOs, I need some reference for reading, as it is an unknown territory for me. I am not saying anyone is wrong, but if i have doubts, i would seek answers. Could you give me more details on the 2 big NOs thing? Thanks

DBC

[Ultrasonic2]

Na, dude, what I meant is besides the simple and complicate answer of NOs, I need some reference for reading, as it is an unknown territory for me. I am not saying anyone is wrong, but if i have doubts, i would seek answers. Could you give me more details on the 2 big NOs thing? Thanks

DBC

This is a great site

http://www.tetech.com/FAQ-Technical-Information.html#2

Here is a link to a TEC that's rated at 25c and 50c
http://www.tetech.com/temodules/grap...99-1.4-0.8.pdf

And as you can see we're 100% correct

[duckbillclinton]

This is a great site

http://www.tetech.com/FAQ-Technical-Information.html#2

Here is a link to a TEC that's rated at 25c and 50c
http://www.tetech.com/temodules/grap...99-1.4-0.8.pdf

And as you can see we're 100% correct

good stuff, thanks! Now i need to print all these out and find me a pen and some papers for calculation, reading is always good.

By the way, I didn't really question the legitamacy on anyone's comments but merely was trying to discuss, but I guess the bad part about internet discussion is, words don't carry a tone with them, so it does sound awkward sometimes by just reading, lol.

[Ultrasonic2]

good stuff, thanks! Now i need to print all these out and find me a pen and some papers for calculation, reading is always good.

By the way, I didn't really question the legitamacy on anyone's comments but merely was trying to discuss, but I guess the bad part about internet discussion is, words don't carry a tone with them, so it does sound awkward sometimes by just reading, lol.

You my find this of some serious use
http://www.overclock.net/peltiers-te...bout-tecs.html

oh and here's a calculator so you dont have to do the number . it's not perfect but it'll give you an idea
http://watercooling.co.nz/node/60

It certainly seems to me that you shouldn't have been commenting on ChillerZero system since you clearly dont know enough to make that judgement call ?

[duckbillclinton]

You my find this of some serious use
http://www.overclock.net/peltiers-te...bout-tecs.html

oh and here's a calculator so you dont have to do the number . it's not perfect but it'll give you an idea
http://watercooling.co.nz/node/60

It certainly seems to me that you shouldn't have been commenting on ChillerZero system since you clearly dont know enough to make that judgement call ?

Well, to be honest, the part I don't know is temperature affecting Qc, and it has really not that much to do with ChillerZero's setup. Because the assumption i was taking is the hot side of TEC being properly cooled, and it is a fact seen from ChillerZero's MCW-6500 setup. Also, we did see his CPU temp was at -10c when idle, and the increased heat from full load changed the temp to 30c. Though I can not be 100% sure on his rig, but one thing I can be sure is i7's TDP since I digged through all the spec documents and did calculations on my own. The error margin I used was pretty high, and no matter how hard i tried to inflated the TDP number, an i7 shall never exceed 360W TDP running at 5 Ghz with 1.6V Vcore (for 360W TDP, the i7 will need to draw at least 360W electricity from ATX power supply, and it is unlikely). Those 2 numbers are already insanely high. So Gabe's proposal is legit given that if his MCW7000-T could really cool the hot side fast enough.

Also, regarding ChillerZero, we all saw is he's pretty hardcore now running to phase change cooling. So one thing we can assume is, it is unlikely he lied on the 226W (192W actual on 320W PSU) TEC cooling part. He used a not so powerful 50mm TEC on his MCW-6500 was just taking a blind shot, but somehow it worked, lol. However though, using a 50mm TEC by him is a pretty damn good choice. You know, one of the TEC factory I called in mainland here from China confirmed they have 50mm TECs that with 199 couples can do 24.1V with 28A, 675W, and that's a whooping 405W Qc converted. In conjunction with a 1000W PSU (i could only find 600W or 1000W, no in between), and if we could cool the TEC's hot side at 50s while full load (edit: it should be 30s, not 50s, giving if i7 pump out 200W heat, and TEC being 100% efficient at 405W), we might be to able to get sub zero. IMHO using 62mm TEC is not optimum, as the extra space not covering the CPU die went to waste (even with large copper plate). Smaller TEC and cold plate are always more effective, but the hot side will be harder to cool. On the other hand, the pity part though, is the factory i called won't sell it to me unless i order them in min quantity of 100, lol, bastards.

After all, like I said before, this whole original thread is Gabe announcing MCW-7000T block, and we all want to see if that's a dud, right? For me, i am a firm believer it will work

[Ultrasonic2]

Well, to be honest, the part I don't know is temperature affecting Qc, and it has really not that much to do with ChillerZero's setup. Because the assumption i was taking is the hot side of TEC being properly cooled, and it is a fact seen from ChillerZero's MCW-6500 setup. Also, we did see his CPU temp was at -10c when idle, and the increased heat from full load changed the temp to 30c. Though I can not be 100% sure on his rig, but one thing I can be sure is i7's TDP since I digged through all the spec documents and did calculations on my own. The error margin I used was pretty high, and no matter how hard i tried to inflated the TDP number, an i7 shall never exceed 360W TDP running at 5 Ghz with 1.6V Vcore (for 360W TDP, the i7 will need to draw at least 360W electricity from ATX power supply, and it is unlikely). Those 2 numbers are already insanely high. So Gabe's proposal is legit given that if his MCW7000-T could really cool the hot side fast enough.

Also, regarding ChillerZero, we all saw is he's pretty hardcore now running to phase change cooling. So one thing we can assume is, it is unlikely he lied on the 226W (192W actual on 320W PSU) TEC cooling part. He used a not so powerful 50mm TEC on his MCW-6500 was just taking a blind shot, but somehow it worked, lol. However though, using a 50mm TEC by him is a pretty damn good choice. You know, one of the TEC factory I called in mainland here from China confirmed they have 50mm TECs that with 199 couples can do 24.1V with 28A, 675W, and that's a whooping 405W Qc converted. In conjunction with a 1000W PSU (i could only find 600W or 1000W, no in between), and if we could cool the TEC's hot side at 50s while full load, we might be to able to get sub zero. IMHO using 62mm TEC is not optimum, as the extra space not covering the CPU die went to waste (even with large copper plate). Smaller TEC and cold plate are always more effective, but the hot side will be harder to cool. On the other hand, the pity part though, is the factory i called won't sell it to me unless i order them in min quantity of 100, lol, bastards.

After all, like I said before, this whole original thread is Gabe announcing MCW-7000T block, and we all want to see if that's a dud, right? For me, i am a firm believer it will work

i personally have not debated whether or not gabe's block with the 360 will work .. i have no dobt it WILL work. im debating ChillerZero results.

However moving to the 360watts Qc TEC and lets say a load of 359watts with a hot side temp of 27c whats the ruff temp of the cold side ?

[duckbillclinton]

i personally have not debated where or not gabe's block with the 360 will work .. i have no dobt it WILL work. im debating ChillerZero results.

However moving to the 360watts Qc TEC and lets say a load of 359watts with a hot side temp of 27c whats the ruff temp of the cold side ?

That will be up to the TEC's spec, and also it has something related to the stuff you sent me to read early. For now, TEC1 in general has a dT of 60-70C if running in full efficiency. So, I will say, if it's 100% efficient on the TEC, the cold side should be still close to 27C (edited here, i misread your post, my previous post was wrong on calc as well, with i7 pumping 200W TDP, and TEC being 405 Qc, with hot side being 30, not 50, we might get sub zero). The actual number will all depends. You can certainly point out i am wrong on the answer, but on the other hand, it seems to me that after so many rounds of our discussion today, you are still trying to discredit ChillerZero, but before you do that, at least check his worklog from the other site, then you will know he also tried another TEC with higher Qc with 2 large HSF with heatpipes, later he gave up and went to phase change cooling.

In addition, if you really want to prove CZ is wrong, setup your own rig with worklog like him, then we will see. I myself however, is planning to do some experiments on my own by year end, and that's why I called the TEC manufacturer. I also called a DC PSU manufacturer as well, and confirmed I can order a 24V 600W or 1000W PSU through one of its distributor. Also I found work shops that can do custom milling and laser cuts for my custom made water block. But this will all depend on my personal life, at this stage, I am not promising anything as PC cooling is just one of my hobby.

By the way, dude, this whole thing is not about CZ rig being legit or not, the problem is we had too much speculation on i7's TDP and no one really bother to find the real answer, and on some threads, some even questioned if Gabe's proposal is doable.

[Ultrasonic2]

That will be up to the TEC's spec, and also it has something related to the stuff you sent me to read early. For now, TEC1 in general has a dT of 60-70C if running in full efficiency. So, I will say, if it's 100% efficient on the TEC, the cold side should be 27-60 = -33C. The actual number will all depends. You can certainly point out i am wrong on the answer, but on the other hand, it seems to me that after so many rounds of our discussion today, you are still trying to discredit ChillerZero, but before you do that, at least check his worklog from the other site, then you will know he also tried another TEC with higher Qc with 2 large HSF with heatpipes, later he gave up and went to phase change cooling.

In addition, if you really want to prove CZ is wrong, setup your own rig with worklog like him, then we will see. I myself however, is planning to do some experiments on my own by year end, and that's why I called the TEC manufacturer. I also called a DC PSU manufacturer as well, and confirmed I can order a 24V 600W or 1000W PSU through one of its distributor. Also I found work shops that can do custom milling and laser cuts for my custom made water block. But this will all depend on my personal life, at this stage, I am not promising anything as PC cooling is just one of my hobby.

By the way, dude, this whole thing is not about CZ rig being legit or not, the problem is we had too much speculation on i7's TDP and no one really bother to find the real answer, and on some threads, some even questioned if Gabe's proposal is doable.

Oh i've read as much of his work log i can via google translation. I dont need to discredit him .. as i originally posted when he brought it up ether the TDP is wrong or he's for getting to tell us all something.

Your not helping him cos at every place you can you spout forth something thats not ture and your discrediting your self quite nicely which is not helping him.

Ie the answer to my question is the hot and cold side of the TEC is virtually the same temp it doesn't matter the size of the original TEC . All you have is 360 watts of Qc. Qc is the amount of heat that can be moved from the cold side of the TEC to the hot side to a delta of 0


SO if you load it with 359 then the temp of the cold and hot side will be virtually the same of 27c

So the answer to my question is not -33 its 27c ish. The Qmax of the TEC or the efficiency it's running at is irrelevant.

On the topic of efficiency you can't have something that's 100 efficient

Stop claim to know something when you clearly know nothing. you'd learn alot more if you just asked question instead if making up rubbish and posting it.

UMM as for me TECing stuff .. i have no problem with that.. i have a quad TEC water to water chiller under my desk and a Direct die 62mm TEC block

Both of which i made on MY CNC Milling machine .. i have plenty of detailed work logs . The Quad TEC chiller work log is at 26,346 view so far and my 62mm TEC block work log is up to 315 relevant posts so far.

You are right this whole debate we have been having has had nothing to do with Gabe or ChillerZero or their rigs .. it's about you posting rubbish claiming that it's right. It would be a dis-service to the TEC community as a whole to let it slid. Cos people like me will have to mop it up later.

So im trying to discredit you and i think i've done a pretty good job so far.

How can you possibly have a valid opinion about Gabe or ChillerZero systems when you CLEARLY DON'T KNOW ANYTHING ABOUT TEC ? Seem crazy to me ?

I ACTUALLY WANT TO HELP YOU which is why i've sent you those links

[duckbillclinton]

Oh i've read as much of his work log i can via google translation. I dont need to discredit him .. as i originally posted when he brought it up ether the TDP is wrong or he's for getting to tell us all something.

Your not helping him cos at every place you can you spout forth something thats not ture and your discrediting your self quite nicely which is not helping him.

Ie the answer to my question is the hot and cold side of the TEC is virtually the same temp it doesn't matter the size of the original TEC . All you have is 360 watts of Qc. Qc is the amount of heat that can be moved from the cold side of the TEC to the hot side to a delta of 0


SO if you load it with 359 then the temp of the cold and hot side will be virtually the same of 27c

So the answer to my question is not -33 its 27c ish. The Qmax of the TEC or the efficiency it's running at is irrelevant.

On the topic of efficiency you can't have something that's 100 efficient

Stop claim to know something when you clearly know nothing. you'd learn alot more if you just asked question instead if making up rubbish and posting it.

UMM as for me TECing stuff .. i have no problem with that.. i have a quad TEC water to water chiller under my desk and a Direct die 62mm TEC block

Both of which i made on MY CNC Milling machine .. i have plenty of detailed work logs . The Quad TEC chiller work log is at 26,346 view so far and my 62mm TEC block work log is up to 315 relevant posts so far.

You are right this whole debate we have been having has had nothing to do with Gabe or ChillerZero or their rigs .. it's about you posting rubbish claiming that it's right. It would be a dis-service to the TEC community as a whole to let it slid. Cos people like me will have to mop it up later.

So im trying to discredit you and i think i've done a pretty good job so far.

I ACTUALLY WANT TO HELP YOU which is why i've sent you those links

dude, no need to cry out, read my post on the edit time (5:37am US EST timezone versus your 5:54am EST). I corrected my mistake before your post. It's still near 27c, since (360 - 359)/ 360 * DT = aprox 0. I didn't cheat to catch my own error.

I did expect a flaming post from you crying out loud as soon as I realize I misread your post and made an obvious mistake. I am here to debate but not to make war.

If you do look into your post time versus my original post's edit time, and acknowledge I did get the answer right, then I would like to apologize for confusion caused and make peace. After all, you are a very knowledgebale person to be a friend. Also, FYI, i edited this post as well, because the words i used was a bit arrogant.

If you disagree, then this post should be the end of our discussion.

[Chillerzero]

the problem was in fact that the two TRUE's with the copper plate didn't get the heat enough out from the TEC.
Or maybe try with a less power TEC, maybe 170w or so.

it was tested on a P4 955 extreme with has almost the same heat as a i7, + the P4 955 EX was overclocked to give more heat and see what the TEC was handling.

but the TRUE getting mega hot, they didn't get the heat away from the TEC.

[zipdogso]

the problem was in fact that the two TRUE's with the copper plate didn't get the heat enough out from the TEC.
Or maybe try with a less power TEC, maybe 170w or so.

it was tested on a P4 955 extreme with has almost the same heat as a i7, + the P4 955 EX was overclocked to give more heat and see what the TEC was handling.

but the TRUE getting mega hot, they didn't get the heat away from the TEC.

And it's taken you 3 months to come up with that......hmmmm
This is pretty much a dead thread Duckbillclinton hasn't been back....after stirring up a hornets nest....
and Gabes new i7 cooler is no nearer appearing on the horizon...

Check dates before replying !!!

[Vinas]

You all got me excited to go back to TEC for no reason.

[zipdogso]

????

Oh sorry..... chillerzero bought back up the i7 thread and you thought it was on....

No sorry about that...nothing haappening with the i7 yet.

[craigbru]

Do you check this thread Gabe? It would be awesome if you could drop in here ant let us know what the status of this block is. I think there are quite a few excited at the possibilities, yet wary of the results. An update either way would be appreciated. If for some reason this block isn't going to be delivered, that information would be nice as well. I'm sure the community would understand if there were some technical or performance limitations. Don't keep us in the dark man!

[gabe]

Do you check this thread Gabe? It would be awesome if you could drop in here ant let us know what the status of this block is. I think there are quite a few excited at the possibilities, yet wary of the results. An update either way would be appreciated. If for some reason this block isn't going to be delivered, that information would be nice as well. I'm sure the community would understand if there were some technical or performance limitations. Don't keep us in the dark man!

Sorry guys, I spoke privately to a number of you but never officially updated this thread. I am sorry to report that I could not meet my goals with any of the TEC modules that I tried. The TDP of these quad cores is still too high for current TEC modules, and the results are not worth the energy expenditure. The solution would work fine for non-overcloked CPU's, but as soon as you increase the voltage and frequency, the performance curves intersect in favor of straight liquid cooling.

I am not giving up yet, as there might be upcoming technologies that could allow us to reach our goals, but for the time being, this project is shelved.

[craigbru]

Alright. Thanks for the direct response Gabe.