Login |  Register |  FAQ
   

Post new topic Reply to topic  [ 4 posts ] 

TEC1-12706 delivering only half the expected Qc

 Post subject: TEC1-12706 delivering only half the expected Qc
PostPosted: Thu Nov 19, 2015 4:06 am 
Offline

Joined: Thu Nov 19, 2015 2:07 am
Posts: 3
Hello,

I'm trying to cool a power laser unit to some 20°C. As I need to shift 40 watts of excess heat, I figured I could use two TEC1-12706 Peltier elements connected in series (electrically), mounted in parallel. These Peltiers have a Qmax of some 50 watts, a Delta Tmax of 66°C, an Imax of 6.4A and a Vmax of 14.4 volts.

The laser unit has a machined brass base, and is stacked on top of two '12706 Peltiers, which are placed on a large heatsink with a fan underneath:
Attachment:
2xTEC1_12706_problem.jpg
2xTEC1_12706_problem.jpg [ 111.75 KiB | Viewed 7610 times ]

At the moment, some 5 kilograms (10 pounds) of pressure is applied to the top of the laser module.

I also built a 4-channel thermometer for simultaneously monitoring several temperatures. In this picture, the top temperature is the actual laser temperature, the second one is the temperature of the brass base (the Peltier cool side), and the third one is the heatsink temperature (the Peltier hot side). The fourth temperature is the ambient temperature (18.7°C). The multimeter shows the actual Peltier current (4A). The bits of tape with the red dots on the left show the temperature sensor positions.

At 40 watts of Qc, I expected to get a Delta T of at least 20°C with two '12706 Peltiers running at 4A -- according to the datasheet, one Peltier should be capable of shifting 25 watts with these parameters. But unfortunately, I can't even get a Delta T of 10°C. Only with 20 watts of heat load can I achieve a Delta T of 20° -- at the cost of pumping 100 watts in the Peltier elements. This seems to me a very poor performance.

This is what I tried so far:
- Different types and brands of Peltier elements, both cheap and expensive. And yes, I found that there were some duds among the eBay types, but the good ones from eBay show the same performance as a set of $30 Laird Technologies 56760-500 Peltiers (identical size as TEC1-12706, with similar specifications). So the lack of efficiency is not due to bad Peltiers.
- Different types of heat compound material, both the cheap white stuff and the far more expensive Arctic Silver variety. It does make a bit of difference, but no more than perhaps 1 or 2 degrees.
- More pressure on the Peltiers. This gives no real improvement.
- Checked for 'heat leaks', but I can't find anything that is likely to cause a >20W heat flow back into the module. Also, it's not that the module picks up ambient heat -- the ambient temperature is the lowest of all and if anything should help cool the module down.

So I guess I must be doing something wrong, but I still have no real idea what -- both the datasheets and online Peltier calculators tell me that a set of two '12706's should be up to the job, but they obviously aren't, at least not on my workbench.

So does anyone here have an idea what I might be doing wrong? Or perhaps what else I could test? yes, I could of course switch to yet heavier Peltier elements, e.g. TEC1-12708, but that means that I need to get rid of yet more heat. It's almost like forgetting to release the parking brake, and wondering why on earth that car won't accelerate. So I really hope it's something simple I overlooked...

Thanks in advance, best regards,

Richard


Top
 Profile  
 
 Post subject: Re: TEC1-12706 delivering only half the expected Qc
PostPosted: Thu Nov 19, 2015 8:22 am 
Offline

Joined: Thu Nov 19, 2015 2:07 am
Posts: 3
Update: I may have found a cause -- the heatsink may not be perfectly flat. After applying substantially more CPU heat paste between the Peltiers and the heatsink, I get much better results: 18°C on the cool side, 38°C on the hot side, and still 40 watts of heat transferred. So I guess I should be more wary about the mechanical aspects of this (I'm an electronics engineer).

Anyway, comments, tips etcetera from more dyed-through-the-wool TEC experts are of course still welcome!

Best regards,

Richard


Top
 Profile  
 
 Post subject: Re: TEC1-12706 delivering only half the expected Qc
PostPosted: Sun Nov 22, 2015 12:59 pm 
Offline
Site Admin
User avatar

Joined: Sat Nov 06, 2010 11:28 am
Posts: 151
hello

it's a bit redundant now. But i believe the fundamental question is why cant i achieve the performance that it should theoretically achieve. The short answer is that they are rated in ideal conditions that can't be replicated in the real world. So results are much worse. However 99% of peoples 'problems' revolve around the quality of peoples cooling they apply to the hotside of their TEC's. While everyone claims their heatsinks are great in reality they are poor.

Better results will normally be as a result of improved cooling and an increase in Qmax with a reduction of input power


Top
 Profile  
 
 Post subject: Re: TEC1-12706 delivering only half the expected Qc -- SOLVE
PostPosted: Wed Nov 25, 2015 8:03 am 
Offline

Joined: Thu Nov 19, 2015 2:07 am
Posts: 3
OK, final update: I measured the heatsink surface carefully, and it turned out that it was far from flat -- there was almost 1 mm deviation over a distance of 100mm -- more than enough to play merry hob with the heat transfer from the Peltiers, even with loads of thermal compound.
In short, I machined the surface flat with less than 0,05mm deviation overall, and lo and behold: now I get no less than 27°C Delta T, with a Qc of 40 watts! This is almost exactly what I would expect!

So my problem is solved -- and I learned that one should never trust the eye when judging the flatness of a surface.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 4 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