Login |  Register |  FAQ
   

Post new topic Reply to topic  [ 9 posts ] 

unexpected performance of 12706 devices

 Post subject: unexpected performance of 12706 devices
PostPosted: Tue Apr 28, 2015 7:53 pm 
Offline

Joined: Thu Apr 16, 2015 11:18 pm
Posts: 4
I have a heat exchanger made up of 12 devices wired in parallel.
The structure is 2 heatsinks either side of the 12 devices. Each device is glued with silicon Plaster to each heatsink.
I am trying to cool the internals of a custom enclosure with a heat load of between 100 and 150W

When the unit is powered up with no heat load i see a temp of approx 0 deg C on the cold side and a temp of 22 deg hot side in an ambient of 20 deg C.

From this i feel my hot side is able to disipate the heat adequately.

We add the heat load and the internal rises to 28 degrees and the cool side rises to 25 degrees C

I am driving the 12 devices with about 5.5V at 13A and from the data sheets i would expect the heat pump capacity of the design to easily pump 150W of power.

Am i missing something ?

Dave kilshaw


Top
 Profile  
 
 Post subject: Re: unexpected performance of 12706 devices
PostPosted: Tue Apr 28, 2015 8:25 pm 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 151
do you have the data sheet for the tec's in question also to clarify your applying 5.5 volts and 1.08 amps per TEC ?


Top
 Profile  
 
 Post subject: Re: unexpected performance of 12706 devices
PostPosted: Tue Apr 28, 2015 8:48 pm 
Offline

Joined: Thu Apr 16, 2015 11:18 pm
Posts: 4
you are correct just over 1 A per TEC


Attachments:
TEC1-12706.pdf [238.04 KiB]
Downloaded 883 times
Top
 Profile  
 
 Post subject: Re: unexpected performance of 12706 devices
PostPosted: Tue Apr 28, 2015 9:37 pm 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 151
Ha yes Hebei .. that cost me countless money with their low cost TEC's that kept failing.


First thing to note based on their graphs later is they're based on a hot side of 50c not 300 kelvin which is the norm. Having this higher Th temp makes the Qc and Dt greater.

Based on their graph bottom left, with a Th of 50c the tec will move 57 watts to 0 delta for 6 amps

At 1.5 amps thats about 18 ish watts at a Th of 50c. But you in theory dont have a Th of 50c but less so it wont be moving 18 but less at 1.5 amps may be 17 or 16 watts at your Th temp.

But also you're saying you're applying less than 1.5 amps so at a guess it maybe moving say 12watts.

12 tec x 12watts moved = 144watts moved to a delta of 0 . Understanding that Qc is to a delta of 0 is extremely important that means if you had a heat load of 144watts then the hot and cold side of the TEC will both be the same temp making the TEC pointless.

Now in the above example we can in theory move 144watts to a delta of 0 but this is far to simplistic and will not be reality because both the heat sinks on the hot and cold sides are not perfect conductors of heat so both apply insulation to the TEC so while the heatsink may be at ambient the TEC's hot side will infact be hotter on the hot side and colder on the cold side. This is not good. This wastes a huge amount of cooling potential because a TEC has a VERY important Delta to watts moved relationship which is what that bottom left graph is showing

If you apply 6amps it can move 57 watts to a delta of nothing making the TEC pointless. Or if you go the other extreme it can move nothing to a delta of 70c this is also pointless in the real world where you want to cool something that is omitting heat energy.


So in short it's working for you exactly as it should.


Top
 Profile  
 
 Post subject: Re: unexpected performance of 12706 devices
PostPosted: Tue Apr 28, 2015 10:20 pm 
Offline

Joined: Thu Apr 16, 2015 11:18 pm
Posts: 4
i understand your points but if i increase the current through the TEC's i dont get any improvement !!

I wondered if this was due to the COP reducing as the current increased

Are you suggestying that some TEC are not very good ?

If so can you recommend a good source of 40 * 40 TEC's


Top
 Profile  
 
 Post subject: Re: unexpected performance of 12706 devices
PostPosted: Wed Apr 29, 2015 8:07 am 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 151
Hello Is there any chance to get a pic of this thing.

Where are you measuring the internal temps from are the air temps ?

When you increased the power by how much and what was the result

To be honest i suspect this will be poor for the same reason all other TEC projects that invole a load and that is poor hot and cold side thermal resistances.

www.customthermoelectric.com


Top
 Profile  
 
 Post subject: Re: unexpected performance of 12706 devices
PostPosted: Wed Apr 29, 2015 8:00 pm 
Offline

Joined: Thu Apr 16, 2015 11:18 pm
Posts: 4
Thanks for the reply

A pic is attached
The assembly is 300mm by 120mm

The cold side has one heatsink bonded with thermal Plaster to the cold side of the TEC and the hot side is also thermally bonded the same way to the hot side heatsink. Air if forced over the cold side by fans which cant be seen. In effect its a duct with the hetsink in it with the warm air flowing through. The hot side is venting to ambient with the fans helping keep the heatsink cool.

Ive have never used the thermal plaster (from ebay) before and wonder if its thermal properties are not as good as thermal silicon compound. We used the plaster as it sets like rubber and glues the heatsink to the TEC. Silicon grease doesnt set as a glue so another mounting method would need to be found.

Are there good / bad TEC's ive only bought them frm Ebay in the past and wondered if they were poor
performance compared with other reputable sources ?

The thermo couples were placed at the base of the fins of each heatsink one can be seen on the cold side

With regard to the power changes we seem to achieve optimum cooling at 5.15V per TEC if the voltage is encreased then the hot side heatsink just gets hotter but the cold side get no colder. This was why i was thinking that all that was happening was the extra power to drive the TEC's was heating the heatsink and the COP value was falling in the same manner.


Attachments:
heat exchanger.jpg
heat exchanger.jpg [ 606.06 KiB | Viewed 13941 times ]
Top
 Profile  
 
 Post subject: Re: unexpected performance of 12706 devices
PostPosted: Thu Apr 30, 2015 8:11 pm 
Offline
User avatar

Joined: Sat Nov 06, 2010 1:33 pm
Posts: 23
Location: Canada
If I'm understanding correctly, you have your TECs mounted between the two heatsinks using thermal glue. This is a sub-par mounting method. TECs perform best when mounted with fairly large compression forces. So mounting using screws or bolts is generally the recommended way to mount TECs. The compression force also tends to reduce the thickness of the TIM (thermal interface material) layer which leads to a lower thermal resistance between the various thermal junctions. This could explain the results you are getting.


Top
 Profile  
 
 Post subject: Re: unexpected performance of 12706 devices
PostPosted: Fri May 01, 2015 8:27 am 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 151
Hi are you able to provide more details

How do you know what the heat load is ? is this an electrical device your trying to cool ?

What cold side temp are you trying to achieve ?

Because it would be worth while working out what should work in theory or not



Im sure that replacing the TEC's and installing them in the recommended fashion will improve your performance. However you've already stated what the problem is. You said that when you increase the power to the TEC's all this results in is hotter heat sink temps. This points to the usual problem of the hot side cooling being to inefficient. When you're applying more power the TEC is achieving a greater delta between the two sides but this is not translating to lower coldside temps because you can't keep the hot side the same temp as before.


Pretty much you're idea is the same as has been tried countless times before and you'll need to bin your poor performing cooling design and go to something more efficient like water cooling


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 9 posts ] 


Who is online

Users browsing this forum: No registered users and 1 guest


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