Author Topic: Disipating 60W of the hot side of the peltier cooler  (Read 2533 times)

0 Members and 1 Guest are viewing this topic.

Offline SauliusTopic starter

  • Newbie
  • Posts: 9
  • Country: lt
  • Green green green green....
Disipating 60W of the hot side of the peltier cooler
« on: August 12, 2019, 02:47:41 pm »
Hi fellow enginerds,

I haven't done much heat dissipation calculations. So I would like some advice.

I will be cooling some devices with a thermoelectric cooler, and I need to dissipate at least 60W on the hot side of the peltier element.
I am trying to decide on the most reasonable solution.
Distentions of the peatier element are 55x55mm.
I was thinking of using a heat sink, but not sure what size should I use and what size fan should I use.

If anyone has any advice, it will be greatly appreciated.
Thanks :)
 

Offline ejeffrey

  • Super Contributor
  • ***
  • Posts: 4033
  • Country: us
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #1 on: August 12, 2019, 03:53:16 pm »
That is about the right size and power budget for a standard CPU heatsink and fan.  I would first look if you can just use one of those off the shelf.
 
The following users thanked this post: NiHaoMike

Offline ConKbot

  • Super Contributor
  • ***
  • Posts: 1400
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #2 on: August 12, 2019, 04:47:07 pm »
How close to ambient do you need to keep the hot side? I.e. If you're just temp stabilizing a laser or something, pretty much any desktop CPU cooler should suffice.
If your trying to stay as close to ambient as possible, something like a NH-D15 is about as good as you'll do in the consumer segment for a 1-piece solution. (I.e. not an All In One sealed water cooler)

Anandtech did proper thermal measurements with a 60W load (convenient for you )  https://www.anandtech.com/show/9415/top-tier-cpu-air-coolers-9way-roundup-review/12

6.5C above ambient at 60W, with thermal resistance coming in at 0.0944C/W for the 60-340W range.
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15797
  • Country: fr
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #3 on: August 12, 2019, 05:34:37 pm »
I would also consider a CPU cooler. Beware that the typical contact surface of a CPU cooler is smaller than 55x55mm, so that may leave a peripheral belt on the Peltier module excessively hot?

What are you going to do with this Peltier cooler?
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7043
  • Country: nl
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #4 on: August 12, 2019, 07:36:36 pm »
There's vapor chamber heat spreaders on Aliexpress nowadays, anyone ever tried them?
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8998
  • Country: us
  • Retired, now restoring antique test equipment
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #5 on: August 13, 2019, 03:39:34 pm »
Be careful with the thermal design.  The heat flowing out of the hot side of the Peltier device is the sum of the heat flow into the cold side (from your source) and the power required to move that heat through the Peltier junction.  That required power increases as the temperature difference across the junction increases, since you are pumping heat “uphill”.  An undersized heat sink will get hotter and increase the required power into the Peltier, making the sink hotter...
 
The following users thanked this post: SiliconWizard

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15797
  • Country: fr
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #6 on: August 13, 2019, 03:59:23 pm »
Another point (why I asked the intended use) is that going below ambient temperature will give severe condensation problems. You can't directly use a Peltier module to cool down something without taking appropriate measures against condensation.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9330
  • Country: fi
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #7 on: August 13, 2019, 04:28:34 pm »
People almost always underestimate the cooling required for the Peltier hot side.

Note that the max dT of a Peltier is specified at zero thermal power transfer. dT=60 degC means that if you maintain hot side at +60, your cold side will maintain at +0 only if it's completely isolated and not cooling anything.

If you actually want to cool something, practical dT is going to be somewhere between 20 and 30 degC. This tends to get you about half of the rated power - i.e., a "60W" Peltier will cool by 30W and heat by 90W. This would be hot side actively cooled at +30 degC, and cold side at about +0 degC. And removing 90W of heat without getting the device surface over 30 degC is the difficult part. A CPU cooler won't do it, it's designed to dissipate 90W while keeping CPU under about 60 degC.

You can cool something to room temperature just by using a heatsink at a room temperature, so if you are using a Peltier, you likely want to go below room temperature.

So I'm guesstimating you want to have your cool side at least below +10degC, possibly even colder. Let's say you want +5degC. This would mean the hot side cannot be allowed to heat to over about 35 degC!

This means, your heatsink design needs to have only about 10 degC temp rise (measured between the Peltier hot side and the ambient) - preferably even less! - to make any sense. This means ridiculously low thermal resistance numbers. This is a Peltier-specific problem - other things (semiconductors, CPUs, etc.) that require cooling are usually just fine with dT=50 degC between the ambient and the device. So to cool a 60W Peltier properly, you'd need a heatsinking solution typically used to cool "400-500W things".

Note that heatpipes and similar vapor phase change based cooling devices won't work - they all need certain dT to start working at all, the thermal resistance is not a linear function. Fine for things that can run at, say, 60 degC, like CPUs, but not OK for a Peltier hot side that you want to cool to almost room temp.

You are left with water cooling, or a well designed fan cooling with a massive heatsink with thick enough base for proper heat spreading (think about 15-20 mm for aluminium or 10mm for copper!), then a lot of fan-cooled fins on it. These behave fairly linearly, and oversized enough, can cool with very low dT.
« Last Edit: August 13, 2019, 04:30:49 pm by Siwastaja »
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8998
  • Country: us
  • Retired, now restoring antique test equipment
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #8 on: August 13, 2019, 04:42:19 pm »
I agree with Siwastaja.  At the back end, efficient transfer of heat from the sink to air requires a high heat-sink temperature, which decreases the capacity of the Peltier.
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7043
  • Country: nl
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #9 on: August 13, 2019, 05:26:00 pm »
Note that heatpipes and similar vapor phase change based cooling devices won't work - they all need certain dT to start working at all, the thermal resistance is not a linear function.

It's not linear, but the non-linearity isn't all that pronounced ... that CPU heatsink already has 0.1 degree per Watt at 6 degrees above ambient. Looking at the results from 60W and 100W, if there is any point where the heat pipes nearly don't work at all it has to be at negligible power and temperature difference. For all extent and purposes they do work, a far behind the comma offset in the temperature above ambient won't make the difference between working and not working.
« Last Edit: August 13, 2019, 05:35:26 pm by Marco »
 

Offline ledtester

  • Super Contributor
  • ***
  • Posts: 3279
  • Country: us
Re: Disipating 60W of the hot side of the peltier cooler
« Reply #10 on: August 13, 2019, 06:10:26 pm »
For calculating the performance of TEC you might be interesting in this video (starting at 2 mins. 40 secs.):



 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf