Peltier control?

Hi all,

New member here. I just bought a Peltier assembly that I saw a video on, and thought it would be cool to try one out. The assembly already has the module, heat sink, and the fan, all 12V. Peltiers are something completely new to me and I have no knowledge about their specifics. I was wondering, do these things need a control circuit? If so, what kind, and if not, why not? I'm kind of a beginner when it comes to electronics, but I'm willing to learn because these things look pretty cool

 

You need to get the specs for your specific peltier. It may well be capable of direct connection to a 12V supply. I have one that is. It has an effective resistance of ~1.2Ω, so it will draw 10A in that case. That's OK because it's rated for that. I run it using a 5V supply, so that it draws only 4A or so. It is under control of a thermostat controlling a MOSFET to switch that 4A current on and off.

The hot side of a TEC will get REALLY hot at the max rated current. It may draw 100W to move 10W from the cold side to the hot, and all 110W show up on the hot side. If your sink and fan can handle that, it will operate. But you can overheat a TEC very easily if you don't move the heat away.

Any TEC is far more efficient when run at well less than rated current. I think you can approach 2 watts consumed for each one moved, at low current.

 

Yes, you really need the specs. If nothing else, then count the number of elements. A nominal 12 volt Peltier will have 16 elements on a side and be square. A 16 x 16 array is 256 less two in the corners for electrical connections leaves 254 total or 127 pairs (N & P). The element height determines the current rating. Never operate a Peltier without heatsinks, especially on the hot side since enough heat can be generated in a few seconds to melt the solder internally and cause catastrophic failure. Avoid operating in free air unless you can keep the cold side above the dew point since condensation will occur on the cold side and the water will cause corrosion, as well as electrical and thermal short circuits. Generally in a refrigerator application the Peltier is operated in a hermetically sealed space. When installing a Peltier between two heatsinks be very careful not to crush it, especially on an edge. There is a real technique to doing this correctly; I have seen countless devices cracked by poor technique, and that is expensive. Each application is a little different, so I won't get into specifics here. My fingers are now calibrated by 25 years experience. These devices are quite voltage tolerant; a 127 couple device can be operated from zero to about 24 volts, but above 14 volts the heatsink has to become very efficient to keep the temperature rise down, and life will be reduced. Of course a negative voltage only means that you reverse the direction of heat flow. These are strictly DC devices, and AC ripple from the power source should be minimized. Naturally, there is a flip side. The device will generate a voltage if heat is applied to one side, and the other side is cooled. Generally the material properties are different for modules used for power generation; mostly due to the high temperatures required to get reasonable power output.

 

Thank you for your reply. The module I got has these specs:
Hot Side Temperature (°C) 27°C 50°C
∆ Tmax (°C-dry N2): 66 74
Qmax (watts): 71 78
Imax (amps): 7.4 7.4
Vmax (vdc): 14.7 16.4
AC Resistance (ohms): 1.6 --
Device ZT 0.74 --

Can I make a circuit that controls it to turn off at a certain temp like 15C and then turn back on when the temperature goes back up? Because I've been thinking of putting it in the side of an old 8 quart cooler I have and cool the inside and see if it'll work properly I'll probably have to use a lot of styrofoam to completely seal it, but luckily I have some of that lying around

 

Thanks a lot for your reply. I've been coming up with all sorts of ideas where I can use this and the one that really appeals to me is using it as a way to make my cooler work without ice. My cooler is pretty small so I don't think it'll be too much of a problem, especially if I put in another small fan on the cold side of the peltier? I've also checked and the module I have is mainly for cooling and not generation. You think I can run it from my car's cigarette lighter receptacle directly? That way I can make my cooler work wherever I go

 

On the cold side, you can probably get away with a large heat sink, maybe even just an aluminum plate. The challenge is greater to shed the heat from the hot side.

The design problem you have for your cooler is simple thermodynamics. The peltier can reach a ∆T at a current level. Let's suppose you target it to run at 4A. Assume you have enough heat exchange at the hot side to keep the peltier at 40°C. What will the cold side temperature? And how many watts can the peltier move under the conditions?

The answers may be that 1) the peltier doesn't get cold enough to freeze ice and 2) the watts the peltier can move is a small number compared to losses through the insulation. In other words I think there's a good chance you will be disappointed in the results. I could be wrong, but I suggest working through this project, on paper, before building.

 

If you are cooling, the hot side needs a seriously good heatsink. The reason for this is: the current flow creates a difference of temperature between the sides. So, if you want the cold side to get down to 0°C and your room is 25°C a temperature difference of 25 degrees is required. But if the current flow causes the whole thing to rise by 25 degrees then you are not getting anywhere!

Another way of looking at this is that the heatsink on the hot side has to remove the heat contained in the thing you are cooling PLUS the heat dissipated in the peltier by the current flow.

The heatsink will certainly need a fan. If you have an old computer lying about, rob the CPU heatsink and fan. These usually have a square pad about the right size for your peltier. Some good ones have a copper slug in the middle.

 

Give me a bit to write up a useful reply. Basically though, the module you have is way overkill for a small cooler. We use a 6A module at about 4 amps to cool a 40 qt box and get a delta T between 40 & 50 ºF. Unless you somehow reduce the voltage the module will be drawing about 6 amps and the heatsink will be running pretty hot. CPU heatsink/fan combos are too small for this module. You would need at least a 92mm fan (case cooler or power supply fans are about this size), and not a low speed one either.

BTW, the MAX specs on you module are tricky devils, but more about this anon.

 

Rather than do a full write-up myself, here is a very good primer on the technology:

http://www.tellurex.com/technology/peltier-faq.php

Look at items 25 through 30 carefully to understand the MAX ratings; if you have difficulty, then ask me about it. I've been working in the field since 1989.

 

Thanks for the input. I have a generously big enough heatsink that came as part of the the whole bundle along with a 12V CPU fan. I ran the thing directly from a 12V DC supply and the cold side starting freezing up in no time. I didn't run it for long though just incase running it directly from a DC source would kill it over time. Again, I'm no expert on this particular subject, so would running the module directly from a pure DC source kill it over time? I'm just asking this because unless I don't manually remove it, the thing will technically get power non-stop.

 

Wow! Never thought it would be feasible to cool something that big. Just wondering how you're getting the cold air to spread through the entire cooler? Is that all done using a small fan on the cold side? I bought another 10qt cooler, drilled a hole in the side and sealed the thing (heatsink + hot side fan on outside, module and cold side fan inside) and while the cold side gets very cold (-3C) the cooler temperature only went down by about 6C in an hour. Why would you think this was the problem? I only ran the test for 1 hr because I was worried that since I did not have a method of controlling the current (I directly connected it to a 12V supply) going in then the peltier would receive too much power as time wore on. Perhaps running the test longer would solve it? I'm thinking of using something like arctic silver just to help with the conduction part. But I can't think of any other reason why the air wouldn't get cold.

BTW, that link you sent me is excellent. I'll read the whole thing because it seems like a good read, and I'll get back to you if I have any questions. Thank you

 

I sorta went ahead and got the thing built before I read your message. Maybe I shouldn't have, because as of right now, i'm only getting a 6C temp differential, even though from where I'm standing the thing looks pretty sealed. The temp I got at the cold side is -3C, but the cooler itself is barely getting any colder, at least in the 1 hr time span I ran the test. Any ideas as to why this would occur?

 

Btw, would you recommend me to get another peltier? I know these things are pretty cheap, and I can get one better suited for the cpu heat sink + fan that I already have.

 

I haven't done the design math, but adding another peltier would likely be a good move. Two in series on a 12V supply, for instance, will lower the current and hot-side temperature while increasing the efficiency and cold-side surface area.