-5
\$\begingroup\$

I am trying to power a set of peltiers and am having some concerns, while I have a computer power supply I can dedicate to the product these are often high in the amps they can deliver, and the peltier specs say a max current that the device can handle, and sources indicate that the devices are in reality a short circuit of two seperate metals....I have tried a few lower amperage power supplies and they all seem to get dangerously hot no mater what my arrangement has been, and actually damaged one unit already by breifly connecting it to a cigarette outlet supply,though this was done as a test without any heat sinks attached it was only done breifly through a 2 amp car lighter adapter cord with a standard 10 amp fuse, so am affraid to risk this by hooking directly to a computer supply.

I am hoping that someone can advise me on a safe and preferably cheap solution to make the current from the laptop or other power supply safe for the peltiers to do their thing, but also have some concern as to the minimum currents being met. So better understanding on how exactly these work in reguards to if I were to have a 15v generator plant that only put out 1 amp would they still work ok as the plant should not overheat with the minimal resistance they have, I'd think and they could not draw more than it is producing, and it is resistance that is supposed to generate heat, in the transformers this heat may be from the inductuctance coils side more so than the resistance on the secondary id presume, though am no electrition just a newbie drawn experimenting with a something that seems cool.

I believe my modules are tec1-12705, but who really knows with chinese ebay products these days...

Alternatively is there a way to reduce the current from building up the heat in the transformers I have tried using, as a 15v 4 amp laptop supply has been tried as well as a 9v 1 amp but anything I use seems to get so hot and really dont need an electrical fire....thanks in advance for any advice to help with this product.

To add more specification I am trying to make a window air conditioning unit for a trailer window that has tilting slat pannels. I have heat sinks(with arctic freeze thermal paste) on the inside and outside with fans on both still (the heat sinks are from pentium 3/4 processors with the larger one on the hot side from old server units i got from a trade), neither heat sink seems to be making any remarkable change in temperature while running at any voltage I have tried, so am wondering how to best understand the heat pumping power of these things,as I understand there are people making refrigeration units out of them that can freeze things but cannot seem to get these to output much change at all. in fact I seem to only be generating extra heat from the power supply into the room.

I have heard of current limiting circuits though these are rather pricey and usually entail an adjustable interface which id just as soon not use, only wanting it to be fully on or off, but would like to be able to have the device be as efficient as possible, pwm suggestions make sense to a degree but presume this reduces the efficiency of the system by whatever factor of time it is in off status, id like to be able to have the device run at 15v but only feed it a minimum amount of amperage if this were possible to still yeild its effects.

I have an arduino I plan to use as a general temp. controller for the circuit once the power supply issues are worked out, so can someone please suggest the best method to power my peltiers for max performance/efficiency without destroying them?

\$\endgroup\$

closed as unclear what you're asking by Trevor_G, Olin Lathrop, Voltage Spike, brhans, winny Jul 14 '17 at 6:49

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ Driving peltiers requires more than just regulating the current. You also need to monitor the temperatures. This question is way too broad for a simple answer here unfortunately. \$\endgroup\$ – Trevor_G Jul 13 '17 at 19:27
  • \$\begingroup\$ i have that side covered with relays and temp sensors tied to an arduino, so is not relevent \$\endgroup\$ – user155835 Jul 13 '17 at 19:29
  • 1
    \$\begingroup\$ Please consider the use of capital letters where required to make your post more readable. You will find that your question is taken more seriously. If you don't bother then why would we? \$\endgroup\$ – Transistor Jul 13 '17 at 19:38
  • 1
    \$\begingroup\$ The very first letter exhibited sloppiness, but I kept going tentatively. Then more sloppiness by the end of the first line. By the middle of the second line the sloppiness and "Eh, I can throw anything I want at these dweebs" attitude was a clear pattern. That's where I quit reading, downvoted, and voted to close. If you don't care about your problem, there is no reason the volunteers here should either. You wouldn't hand in homework this disrepsectful, and your teachers are paid to read it. It certainly doesn't belong here. Screw this. \$\endgroup\$ – Olin Lathrop Jul 13 '17 at 19:40
  • 1
    \$\begingroup\$ BTW: Peltiers do get VERY hot... You need LARGE HEATSINKS and fans. \$\endgroup\$ – Trevor_G Jul 13 '17 at 20:10
2
\$\begingroup\$

With due respect to @Will, Peltiers will not "draw just about as much as you'll give it". Instead, for a given voltage they will pull approximately a constant current, proportional to the voltage. Assuming that what you have is indeed a tec1-12705 or equivalent, I'd advise driving it with 12 volts until you have more experience. Your current will be less than 5 amps. In principle (which means, if you read the data sheet and paid attention) 15 volts will provide about 5.3 amps, but only if you keep the hot side temperature at 25C. And I'll wager good money that you have not done so.

What I'm pretty sure you are doing (well, not doing, actually) is properly heat-sinking your hot side. You need to supply a decent (like 30 or 40 square inches, with fin depth of at least an inch) heat sink, and arranging for forced airflow will not hurt your cause, either. At 15 volts and 5 amps the Peltier must dissipate on the order of 75 watts, and a 2" x 2" flat surface will indeed get very hot at this power level. You will also need to learn about thermal interfaces when you connect to the heatsink, and you'll need either thermal compound or a Sil-pad. If you use thermal compound you must use a very, very thin layer, and you must provide symmetrical, uniform clamping pressure. Alternatively, a good thermally conductive epoxy can be used, but again you'll need to clamp it with even pressure while it cures. The idea is to keep a thin, uniform layer (usually 10 thousandths of an inch is a good goal) between the two surfaces.

Start by applying about 1 to 2 volts. You will notice that one side gets hot and the other cold. I notice that in comments you state that you have at least one temperature sensor connected to the hot side. Use this to monitor the temperature.

\$\endgroup\$
  • \$\begingroup\$ Yes,i do have the peltiers sandwiched between computer heat syncs, and am using arctic freeze thermal compound between both sides and the two syncs but cannot guarantee perfect pressure or thickness, just dabbed thin with my finger and kept the larger sync to the hot side.if it helps im trying to make a room air conditioner for a slat style window on an old trailer, wanted to make sure it worked first so have been trying various power supplies to test efficiency before attaching other electronics. \$\endgroup\$ – user155835 Jul 13 '17 at 20:52
  • \$\begingroup\$ @user155835 - it's a heat sink, not a heat sync. \$\endgroup\$ – brhans Jul 14 '17 at 1:06
  • \$\begingroup\$ @user155835 - If you've killed a module, your heat sinks are grossly inadequate. Although you've got heat sinks, have you included the fans? And when you overheated a unit, how cold did the cold side get? \$\endgroup\$ – WhatRoughBeast Jul 14 '17 at 2:23
0
\$\begingroup\$

The reason everything is getting so hot is because the Peltier device will draw just about as much as you'll give it. You're going to want to look at doing some pulse-width-modulation (PWM) of the device in order to reduce the duty cycle of your power supply (you're at 100% ON right now, and PWM will allow you to modulate it all the way between 0 and 100%). The cheapest way to do this is using a 555 timer to control a transistor or MOSFET or H-bridge between the supply and your Peltier device. If you're looking to do this quickly without designing a complicated circuit, then you could look around for a motor controller or something similar that uses an H-bridge and PWM.

\$\endgroup\$
  • \$\begingroup\$ pwm may be a good solution to reduce total load time on the transformers for the ones i have used but does not address the 45 amps 12v from the computer supply which would likely fry out the peltiers with the first pulse \$\endgroup\$ – user155835 Jul 13 '17 at 19:52
  • \$\begingroup\$ @user155835 just because your PSU CAN deliver 45A does not mean it will. The current is dictated by the load not the power supply. By my reckoning it will max out at 6.25A at 15V @ 25C \$\endgroup\$ – Trevor_G Jul 13 '17 at 20:04
  • \$\begingroup\$ @user155835 5A at 12V,, which is 60W of heat you need to dump.... \$\endgroup\$ – Trevor_G Jul 13 '17 at 20:14

Not the answer you're looking for? Browse other questions tagged or ask your own question.