Author Topic: Thermo-electric modules  (Read 1525 times)

0 Members and 1 Guest are viewing this topic.

Offline raptor1956Topic starter

  • Frequent Contributor
  • **
  • Posts: 869
  • Country: us
Thermo-electric modules
« on: September 22, 2016, 12:28:39 am »
I just picked up some (10) peltier modules from Amazon (see link) as I have a project that needs some cooling and I just received them.  I did a quick test to see if they worked and which side was hot and which was cold so I fired up my Tekpower power supply as it can do a bit more than 5A at a bit over 30V.  I set it to 5.2A, the max setting and 12V and tested all ten pieces I'd received.  The first one I tried drew no current so I tested with the leads reversed -- no current.  OK, not a good start.  The next one powered up just fine and I was able to tell, very quickly, which side was hot and which side was cold and marked it with an H and a C.  I tested the rest of them and all told 3 of them were DOA while the other 7 appear to be functioning as they should.

The meager specifications on the Amazon page indicates the following:

Size: 40mm x 40mm x 3.8mm
Voltage : 12V
Umax (V) : 15V
Imax (A) : 4.3-4.6A
QMax (W) : 72W
Power Cord : 300mm
Net weight : 22g
Color : White

https://www.amazon.com/gp/product/B00K67XZF4/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1

I found that the temp on the hot side got hot real quick and I didn't want to run them without a heatsink on the hot side so I grabed one of the heat sinks I picked up for the project -- they're 50mm x 50mm x 20mm and are therefore a bit larger than the 40mm x 40mm modules.  I put one heatsink on the cold side and another on the hot side then powered back up again with the idea of running longer and getting a feel for how hot and cold it would get with the heat sinks.  After just a couple minutes the hot side heatsink was quite noticably warm bordering on hot and the TC contact probe indicated 65C (150F) at a voltage and current of 12.5V and 2.4A with the current going down from a high of about 2.6A at first power.

OK, so I have a few questions about these modules and about the best method for using them.  First, increasing the voltage didn't seem to cause an obvious increase in current so I have to wonder if going above, say, 12V gains any cooling effect or if it only results in more heat on the hot side.  Second, would it be wise to use thermal paste between the module and the heatsink and if so is there a particular type of thermal paste that's recommended?  Lastly, while it's easy to calculate the power consumption from the voltage and current delivered is there a good formula that one can use to calculate the heating and cooling power and how that is effected by voltage?


Brian
« Last Edit: September 22, 2016, 09:03:43 pm by raptor1956 »
 

Offline raptor1956Topic starter

  • Frequent Contributor
  • **
  • Posts: 869
  • Country: us
Re: Thermo-electric mpdules
« Reply #1 on: September 22, 2016, 03:01:49 am »
One more question -- what are the practical limits for hot and cold temperature for these modules.  This is probably a two level answer as the limit might be useful limits or it could be actual maximums that you do not want to exceed or risk killing them.  My guess here is that the hot side temp is the limiting factor but I guess it's possible that there might be a minimum temp as well.

When you stack them to increase the temp differential knowing what the effective cooling/heating power is would be even more important to know.


Brian
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14172
  • Country: de
Re: Thermo-electric mpdules
« Reply #2 on: September 22, 2016, 02:51:28 pm »
There is an upper limit for the temperature of the module. Depending on the module type this could be somewhere around 100 C. Too high a temperature can destroy a module, so one has to be careful in using it without good cooling of the hot side.

The power can be approximated as:
 P = P_max (1  -  (I/I_opt)²)) - dT * L

I_opt is usually a little smaller than the maximum current.
L is the thermal conductance, it is about equal to dT_max / P_max.
Usually dT_max is about 50-60 K depending on the quality of the module.

One should definitely use thermal paste with heat sinks. Due to the thermal resistance the effective parameters of the TEC are reduced. So P_max and I_opt and thus also dT_max get a little smaller.
For best efficiency the current should be smaller than I_opt, if the temperature difference is smaller.

The voltage is mainly set by the resistance of the module, a slight extra contribution comes from the thermal EMF, but not really much.
 
The following users thanked this post: raptor1956

Offline raptor1956Topic starter

  • Frequent Contributor
  • **
  • Posts: 869
  • Country: us
Re: Thermo-electric mpdules
« Reply #3 on: September 22, 2016, 07:44:56 pm »
There is an upper limit for the temperature of the module. Depending on the module type this could be somewhere around 100 C. Too high a temperature can destroy a module, so one has to be careful in using it without good cooling of the hot side.

The power can be approximated as:
 P = P_max (1  -  (I/I_opt)²)) - dT * L

I_opt is usually a little smaller than the maximum current.
L is the thermal conductance, it is about equal to dT_max / P_max.
Usually dT_max is about 50-60 K depending on the quality of the module.

One should definitely use thermal paste with heat sinks. Due to the thermal resistance the effective parameters of the TEC are reduced. So P_max and I_opt and thus also dT_max get a little smaller.
For best efficiency the current should be smaller than I_opt, if the temperature difference is smaller.

The voltage is mainly set by the resistance of the module, a slight extra contribution comes from the thermal EMF, but not really much.

Great info Klein, really appreciate it...

I've done some more testing and there appears to be a good deal of sample variation with the current at 12V.  I also had another module fail making it 4 out of 10 that do not work.  I had the voltage at 13.5V, below the max value, and after a couple minutes it went open circuit even though I had the hot side cooled to keep the max temp below about 65.

So, it's not looking like the batch I received is very good and the problem I see is that these mostly Chinese made peltier modules are pot luck with respect to quality.  Looking at Digikey the prices for very similar looking modules is upwards of 10X the price and frankly beyond the practical price point I was looking at given a need for 6-12 of these modules per system -- the price for these modules alone could exceed $500USD and that's not far off of where I anticipate the unit would sell for when finished. 

The other observation I have is that the amount of power resulting in heating is WAY more than the amount resulting in cooling and that even modest cooling is going to demand a pretty massive heatsink and fan system to take the heat away and larger heatsinks and more powerful fans add more cost to the picture.  I just wish these modules were more symmetrical with respect the heating and cooling output.


Brian
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf