Login |  Register |  FAQ
   
Post new topic Reply to topic  [ 58 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next

New water block run request

 Post subject: Re: New water block run request
PostPosted: Tue Dec 01, 2015 4:28 pm 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 150
Here is it's current size.

Attachment:
size.JPG
size.JPG [ 36.42 KiB | Viewed 7101 times ]


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Wed Dec 02, 2015 8:49 pm 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 150
Just so everyone knows im completely redoing the block from scratch, though it will look pretty much the same.

Soon ish i'll do a bunch of tests with these variables.

Water block base thickness
Inlet / out let locations
Fin thickness


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Wed Dec 02, 2015 9:34 pm 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 150
Ruff new base

Attachment:
new base.JPG
new base.JPG [ 135.55 KiB | Viewed 7092 times ]



So this is 4mm longer than v2

The lip is their cos im assuming i'll be running a much thinner base. That ridge is their to retain rigidity. if the base is 1mm thinner then that ridge will be 1mm high.

I'm very convenient that reducing the base thickness by 1mm will result in significant gains at 800 watts. (i guess 2c-4c)


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Thu Dec 03, 2015 11:36 am 
Offline

Joined: Thu Jul 03, 2014 12:51 pm
Posts: 22
That looks bad to the bone. I'm looking forward to this project. I have to release the funds from my bank as they put a hold on purchasing the TEC from China pending verification. Should be here in 3 to 4 weeks.
I have the PS here and an Arduino board. I plan on affixing sensors to both the cold and hot plates as well as in the water loop. I have sensors with LCD displays in my cart at Performance PCs but am going to look at possibly making a controller with a temperature display and PWM TEC control. Should be fun.


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Thu Dec 03, 2015 4:28 pm 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 150
Thick8 wrote:
That looks bad to the bone. I'm looking forward to this project. I have to release the funds from my bank as they put a hold on purchasing the TEC from China pending verification. Should be here in 3 to 4 weeks.
I have the PS here and an Arduino board. I plan on affixing sensors to both the cold and hot plates as well as in the water loop. I have sensors with LCD displays in my cart at Performance PCs but am going to look at possibly making a controller with a temperature display and PWM TEC control. Should be fun.



Just so everyone knows now this has turned into a project and will take a while before i actually start making it. But this is all fun stuff.

Now i need some help from some one on the specs of the TEC. Since we're going to play with base thickness correctly modeling a TEC is going to be very important. It would be widely wrong to assign a 800 watt heat load to a 62mm x 62mm square block because this is not the truth. Within the TEC there are 200 or whatever modules within it. There are gaps between them / their conductor. This means there's little horizontal heat spreading going on and the fore the WB base needs to do this. So to accurately calculate the base thickness i'll create a real TEC with gaps and then we'll beable to find the correct base thickness.

So someone needs to get his info from the manufacture.

Conductor dimensions.
Gaps between them.
Indent from the out side of the tec (cos the conductors don't start at 62mm)

Attachment:
peltiertheory_1244821990.jpg
peltiertheory_1244821990.jpg [ 32.54 KiB | Viewed 7025 times ]


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Thu Dec 03, 2015 6:43 pm 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 150
Right this is the SIMPLIFIED new block. It's simplified to make testing of the 3 variables easier

Attachment:
new base.JPG
new base.JPG [ 44.44 KiB | Viewed 7025 times ]


Attachment:
new cover b.JPG
new cover b.JPG [ 26.59 KiB | Viewed 7025 times ]


Attachment:
new cover t.JPG
new cover t.JPG [ 19.9 KiB | Viewed 7025 times ]


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Fri Dec 04, 2015 7:29 am 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 150
So i pulled apart one of my 50mm 243 tec's and it has a 0.7mm gap So this gives me something to go on.

This is all important cos like i've previously mentioned the hot and cold copper blocks need to spread across this gap. This will be crucially important to the waterblocks base thickness.

Also previously when pretending that the TEC was one large continuous 60mm X60mm TEC this did not treat it like it really is which is 486 independent cooling modules that are unrelated to each other.

This individuality is important cos each one is only capable of moving it's % of the total Qmax So for a 600 Qmax TEC each one can only move 1.234 Qmax

This is all important because if you concentrate all your cooling to say the centre and neglect the out side of the TEC you can only ever move the 1.234 from the centre ones no matter the cooling applied and you may have now lost 1/2 of the cooling potential of the out side ones. (this all assumes there is no heat spreaders involved.)

So to get accurate hot and cold side thickness we need to cut the TEC up into lots of little ones. This is something i've never heard of anyone discuss let alone model up.

Although a 243 TEC actually has 486 modules they are electrically connected and this connector acts as an internal heat spreader so i'll be modelling then up in pairs as rectangles including the connector. This should be more than accurate enough for what we're trying to achieve.

You'll notice that in the corners on one side of the TEC sone modules will be missing, this is because this is how a TEC actually is because this is where the wires come out.

Image

Lastly the TEC's ceramic covers will provide some heat spreading though poorly. Im not going to factor this in, so mine will be worse case, In saying that i don't know for a fact that the gap for the TEC's we'll be using, they maybe greater.


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Sat Dec 05, 2015 2:04 am 
Offline

Joined: Tue Nov 24, 2015 5:02 pm
Posts: 18
sorry I got sidetracked for a few days with a TEC chiller experiment.

Just read all this. Good stuff re new block design, I'm always keen for optimising stuff and you certainly have the simulation skills to try to get us the best block possible. Thermonamic is very reluctant to give out much info other than how much money they want us to pay them :P


However, in this thread on OCN, foxrena held up a CTE 400w TEC to the light to see all the couples, then he held up the 669w TEC to the light for comparison. the first one is CTE 400w, second one is Thermo 669w

CTE 400w http://cdn.overclock.net/8/86/900x900px ... 31.47.jpeg

Thermo 669w http://cdn.overclock.net/b/b0/900x900px ... 33.38.jpeg

that's probably the best we can get, if you can import that into cad and template the couples you will be close to the actual size, the CTE is clearer because its ceramic is slightly thinner I believe, but it appears that thermonamics couple size might be slightly larger, they appear to have smaller gaps, hard to tell. Thermo 669w is on the right, slightly thicker than CTE 400w http://cdn.overclock.net/3/32/900x900px ... 28.08.jpeg

They are both supposed to be 199 couples. Both 62x62mm. thermo is 3.8mm thick give or take.

Now, I had asked thermo for a quote for the TEC with Aluminium Nitride Ceramic which has far better thermal conductivity than the standard Aluminium Oxide Ceramic, like 4-10 times better I think. I also asked for the 200*c CuSn solder. The price premium for the Aluminium Nitride is crazy though, USD$450+shipping, so unfortunately I don't think I can stretch to that.

However it would be fantastic to know a result through simulations, ie model it up with 2 different thermal conductivities of the ceramic and see if it would change things. It's too expensive to buy though so it's a waste of time simulating that I guess.

I would love it if you could allow for a middle cold plate to accommodate stacking TECs. The bottom cold plate should have the same design regardless of 1 or 2 TEC's, but the middle cold plate probably needs to be 3mm or so, it doesn't have to spread heat, just has to trasnfer it directly from 62mm TEC to 62mm TEC, it has to be about 3mm thick though so I can drill a thermistor into it.

Also, do you see any problems stacking 2 62mm TEC's? it deviates from the norm of reducing size of TEC for every stack, but I think same size would work better because of the way TEC's work as you are pointing out here, ie designed to transfer heat directly not transversely, so having the same size TEC's stacked should perform better, and then leave it up to the bottom cold plate to spread the heat from IHS out to 62mm cold plate. Although having a smaller TEC on the bottom would be better from that perspective as the heat wouldn;t have to travel so far in that stage. Anyways, just ideas, but I'm thinking a thermonamic 511w 62mm TEC at 12v and a thermonamic 669w 62mm TEC at 12v stacked will actually provide very good performance and delta and still keep power consumption from TEC's at about ~360w. The 511w is a 241 couple TEC, so its pretty efficient for a first stage.

511w http://www.thermonamic.com/TEHC1-24127-English.pdf
669w http://www.thermonamic.com/TEHC1-19940-English.PDF

This is the thread where foxrena tested the 669w TEC http://www.overclock.net/t/1425145/ther ... t_21945475 it didn't perform as well as claimed in the data sheet, but he estimates it performs about 20% better than Customthermoelectrics 400w TEC, so that's what led me to thinking about stacking, to achieve the delta I want and to make it worthwhile purchasing your block and also to do your block justice, its a good piece of kit and I think it deserves to get some real performance out of it, also it will legitimise TEC's a bit in the eyes of the doubters if we can get close to a 40*c dT at 100% load on a heavily overclocked CPU.


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Sun Dec 06, 2015 1:49 pm 
Offline

Joined: Tue Nov 24, 2015 5:02 pm
Posts: 18
Ok so I've just installed my 4670K ES CPU and OC'd to 4.8ghz 1.275v, I'm in desperate need of this 62mm TEC block Ultra hahahaha. No I can wait, but my current kryotherm 111 couple 13.7v 23.1a TEC is just not capable of handling the heat load, I end up with close to 0*c delta and my temps would be better on water cooling alone. So, no rush, but yes I am 100% buying one of your blocks :D


Top
 Profile  
 
 Post subject: Re: New water block run request
PostPosted: Tue Dec 08, 2015 8:24 am 
Offline

Joined: Thu Jul 03, 2014 12:51 pm
Posts: 22
I'm a buyer too!
That's a very good picture of the TEC that you posted Liam. I could make a diagram of the shadows cast through it with a CAD program if you think that would help give you an idea of the coupler layout. Let me know.
Would the raised outer edges effect the heat transfer at the edges of the TEC?
John


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 58 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next


Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  

cron

Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
CoDFaction Style by Daniel St. Jules of Gamexe.net