Need help with peltier cooled laptop case

[aarond]

Bought a pair of these coolers:

http://cgi.ebay.com/ws/eBayISAPI.dll...=STRK:MEWNX:IT

I was hoping to power it with this:

http://cgi.ebay.com/ws/eBayISAPI.dll...=STRK:MEWNX:IT


Just got the coolers in the mail. Hooked up to a 12v source one side gets warm immediately and the othe side gets cool for a few seconds and then the whole thing heats up. I'm sure one side gets hotter than the other but dang... I realize you are supposed to heatsink the hot side. So then I got a big hunk of 1/8 thick aluminum and held it up to the hot side. Powered up the "cooler" and the cold side just got hot at a much slower rate. Do these things need some type of forced cooling to get a large difference btw the two sides?

I was hoping to replace one side of this:

http://www.harborfreight.com/18-inch...ase-39196.html

with a piece of 11" x16" x 1/8" thick aluminum (the sides of the case are actually plastic) and use that to passively cool the hot side of the cooler. My plan was to insulate the plate on the inside of the case to keep the hot plate from just transferring heat back into it. Seemed like a simple idea but I'm not sure my plate will do what I need it to do passively.

Any input?

My goal is/was to build a sand (think the fine talc like stuff found in parts of the Middle East) resistant case that I could store my laptop in to use in a remote setup. I'm not sure how much heat the power supply will put off but it is an efficient switching supply that I am hoping is pretty efficient. My laptop supply is rated to over 60 watts but it barely gets warm to the touch (not sure what he typical current draw is though). I'd like to have it in the case as well.

Seemed like a good idea on paper but I can't help but think I'm building more of an oven than a cooler, lol.

Help......

[Brewski]

Are you trying to cool the laptop without a source for cool air? By putting the coolers in a case without an inlet for air your going to limit the systems ability to keep the laptop cool because the heat will still be trapped in the case heating the air the laptop is trying to use for cooling. I think even if the hot side of the cooler was outside of the case with the cool side in you'd still have the problem of the laptop heating the air faster than the cooler could cool the air.

I've had really good luck with a Zalman NC2000 and my wife uses a Zalman NC1500 I'd recommend both as good coolers but they still need to be able to draw in fresh air to really keep a laptop cool.

Take it easy
Jay

[aarond]

Basically I am wanting to build a more or less airtight case. I was hoping the coolers would be enough to passivley cool whatever internal heat is created (laptop screen won't even be on). The cooler are rated for 137 watts (x2). That is a lot of heat but I'm not sure how efficient they are so who knows how effective it will be. They show these things freezing water on youtube so I know they CAN have a dramatic cooling effect.

[Brewski]

I'd still think that in the same enclosure the hot side would neutralize the effects of the cold side in terms of cooling since that heat doesn't have anywhere to radiate to. Plus you'd still have some heat generated by the laptop even though you weren't using the screen. The video card arguably the hottest part of a laptop while gaming would still be creating heat if you were sending your video out to an external screen. It might be feasible but I'm not seeing how.

Take it easy
Jay

[killersoundz]

Basically I am wanting to build a more or less airtight case. I was hoping the coolers would be enough to passivley cool whatever internal heat is created (laptop screen won't even be on). The cooler are rated for 137 watts (x2). That is a lot of heat but I'm not sure how efficient they are so who knows how effective it will be. They show these things freezing water on youtube so I know they CAN have a dramatic cooling effect.

This kind of reminds me of the video of the girl with her solution for global warming to "put air conditioners out side to cool the air down". You can't create cold air without heat, air conditioners make more heat than cold, plus the heat generated by the source of electricity. !!

In this case, if you put these in an enclosed case, just like an air conditioner they are not 100% efficient, in fact less than 50%, meaning they make more heat than cold (as you've observed), so in a case it will work as a heater. Even if it was 50% efficient it would only cancel each other out in an enclosed case. I'm skeptical of just how efficient they are at cooling, probably like 5%, but anyways you would need to put the "hot" side on the outside of the case, with a heatsink...like you would a window air conditioner.

[kmibb]

I've seen peltier plates/thermoelectric coolers for desktop computers with overclocked CPUs. They need water cooling on the warm side. At the least, a nice copper heatsink with heatpipes and a fan. These things produce serious heat, and I've only seen it successfully installed with water cooled setups. That aluminum block is definitely not sufficient.

I think what you're using it for isn't a good application. They are generally used to cool something, not just to dissipate heat. For overclocked CPUs that generate heat, these are used to cool the CPU. Most CPUs don't need to be cooled, they just need a heatsink to dissipate the heat.

[aarond]

Maybe I concentrated on explaining the trees so much that I forgot to point out the forest.

The 11 x 16 aluminum panel that would replace one side of the case IS THE HEATSINK.

The side portion of the panel inside the case would be insulated to keep the cooler air in and away from the hotter heatsink. Heatsink on top part of case w/ coolers mounted below on the "top of the inside".

I am familiar with thermodynamic laws...

http://www.youtube.com/results?searc...=0&oq=peltier+

[kmibb]

I realize the aluminum is a heatsink, but it's aluminum, which is not great for dissipating heat. Also, it's passive, and you need active cooling to support the peltier plates (or a heatsink with infinite surface area :P.)

[killersoundz]

I realize the aluminum is a heatsink, but it's aluminum, which is not great for dissipating heat. Also, it's passive, and you need active cooling to support the peltier plates (or a heatsink with infinite surface area :P.)

Actually aluminum is second best only to copper for thermal conductivity...

[donradick]

What you are trying to do is probably possible. I don't know specifics, but I've seen several discussion threads on this site:
http://www.silentpcreview.com/

[neildavis]

I made a "personal air conditioner" with some peltier devices many years ago, and found out that designing with these devices can be tricky. There are two main issues. First, as others have pointed out, you need a large heatsink and possibly a fan to cool the "hot" side. The devices are designed to be run at a set "deltaT" operating point, which is the heat differential between the plates. But if the hot plate is too hot the cold side won't be cold--just X degrees cooler than the hot side.

Second, you need to making sure you are providing the right amount of current (see the attachments). The devices have an efficiency curve such that they are very inefficient at high and low currents. If the device isn't running at a point where it is efficient, you will not have much heat pumping and both sides will get hot. You need to look at the specs for the device and figure out the best operating point for the deltaT that you need. You will need to limit the current, either by using a resistor or a regulator. You can't just hook them up to a voltage source.

So you will need to juggle the amount of heatsinking versus the current versus the efficiency to arrive at the deltaT you need. There are some online calculators and lots of info on these devices--the link is just the first one I found.

You can see the tradeoffs involved by comparing the two attached charts. You will get a lot of heat pumped at around 8A for the device in the example, but the second graph shows that the efficiency (COP) drops off rapidly for currents higher than 2-3A. With reduced efficiency you will need to put in a lot of power (heat), so you will need a big heatsink.

http://www.ferrotec.com/index.php?id...ail&mod_id=118

[kmibb]

Actually aluminum is second best only to copper for thermal conductivity...

But copper is almost twice as good...

[aarond]

Good input Mr Davis...

I have my head wrapped around about 90% of that info. I really wish i could find some performance graphs for my specific TEC.

Would you happen to have any input based on the specs I did find?

Couples: 127
Umax (v): 15.2
Imax (A) 9
DeltaTmax: 62
Qmax: 89.2
R(ohms): 1.42


There is also a table of specs found in the eBay auction I linked:
http://cgi.ebay.com/ws/eBayISAPI.dll...=STRK:MEWNX:IT

I was going to post those as well but I wasn't certian how to format it.

Anyone have any input regarding my app based on these specs?

[1100xxben]

A peltier is simply a heat pump. It will transfer heat from the "cold" side to the "hot" side. So, to work effectively in your scenario, you need something attached to the cold side of the peltier to absorb the heat on the inside of the case, and you need something on the outside of the case to transfer the heat from the peltier to ambient air. If you're trying to pump 137 Watts (x2), and assume 50% efficiency (that might be an optimistic number), then you need to dissipate >400 Watts from that 11x16" plate to the air, and that's not going to happen if you want the inside to maintain a reasonable temperature. You will need some sort of active cooling on the outside, and possibly something with a larger surface area (fins). And, as neildavis pointed out, running them at a full 12V isn't necessarily the best thing to do.

[aarond]

A peltier is simply a heat pump. It will transfer heat from the "cold" side to the "hot" side. So, to work effectively in your scenario, you need something attached to the cold side of the peltier to absorb the heat on the inside of the case, and you need something on the outside of the case to transfer the heat from the peltier to ambient air. If you're trying to pump 137 Watts (x2), and assume 50% efficiency (that might be an optimistic number), then you need to dissipate >400 Watts from that 11x16" plate to the air, and that's not going to happen if you want the inside to maintain a reasonable temperature. You will need some sort of active cooling on the outside, and possibly something with a larger surface area (fins).

So if I could get the cold side of the cooler closer to the output of the laptop fan output (hottest part of the case) that would be ideal? Not sure it is practical but I can think about it if it is worth something.

Also, what I'm picking up is that a fan blowing on the outside of the heatsink is a good idea/necessity... easy enough to do. This isn't going to be a mobile setup. Keyboard, monitor, and mouse will all be outside of the case.

... And, as neildavis pointed out, running them at a full 12V isn't necessarily the best thing to do.

You are the second person I have noted in my searches that stated VOLTAGE limiting was a concern. neildavis specifically mentioned CURRENT limiting and that's what the graphs he posted represent. Am I missing something

[neildavis]

You are the second person I have noted in my searches that stated VOLTAGE limiting was a concern. neildavis specifically mentioned CURRENT limiting and that's what the graphs he posted represent. Am I missing something

Well, the device itself has an impedance, so you can control it either way. Look at the I/V curve for one of the Ferrotec devices (I attached the curve for model 9500/127/060 B). This device draws approximately 4.5A when using 12 volts. But the efficiency is fairly low at 4.5A--it would be better to run it at 3A so you will have less wasted heat that you need to get rid of. So for 3A you would need about 8 or 9V.

Unfortunately, these peltier devices don't have a simple "sweet spot". You need to figure out how much of a temperature differential you want and how much heat you can dissipate with your heatsink and fan, and then set the operating point by using the manufacturer's curves.

If you don't have the data sheet, I would just find something similar from a vendor that has the data posted. Your thermal resistance calculations (heatsink to air) are going to be a rough estimate anyway, so you just need to find a datasheet that is somewhat close to figure out a good operating point.

My experience with peltier devices was frustrating--I kept having to use bigger and bigger heatsinks to finally make my project work. The peltier device itself requires a fair amount of power, plus you need to pump out the heat from the components you are trying to cool.

[Crayborn]

Skip on the POS Harbor Freight case and get a Pelican. You will not be happier.

Chuck