TEC Basics - Do they need to be driven?

[rhd]

I hate to start a thread for something that has surely been covered in length, but it's remarkably difficult to search for three letter words like "TEC"

This is a basic question -

I have a couple cheapo 12706 TECs sitting around here (Like These), and I've never known whether they need to be "driven" or not? They're rated 12V. Do I just supply them with 12V? Or should I be wiring them to a current regulated driver (maybe a 317 based design) etc.

 

[Trevor]

I hate to start a thread for something that has surely been covered in length, but it's remarkably difficult to search for three letter words like "TEC"

This is a basic question -

I have a couple cheapo 12706 TECs sitting around here (Like These), and I've never known whether they need to be "driven" or not? They're rated 12V. Do I just supply them with 12V? Or should I be wiring them to a current regulated driver (maybe a 317 based design) etc.

When I read the title I was confused. I thought you were asking if they had to be driven as in if you could just put them in a build not connected to anything and have them cool something.

...I've no experience with these either, so I'm interested in what others have to say on the issue. From what I gather they might be run on a duty cycle in conjunction with a thermistor to maintain a temperature usually? :can:

-Trevor

 

[Jufran88]

TECs are heats pumps which have a cold side and a hot side. The hot side transers heat to the cold side thus cooling the object its applied to.

I don't think TECs are driven, but fed directly.

However, with its other applications such as a low cost thermopile, you would need design an amplifier circuit to read the low voltage it gives off when heat is applied.

I don't know if this helps, but I tried.

You could try searching thermoelectric/peltier modules.

 

[Tech_Junkie]

I asked this question a while ago, but cant find the thread. IIRC yes you should use a driver. I cant remember the reasons why, but that was the answer. They should not be driven direct current.

 

[Jufran88]

I came across this information while googling:
Thermoelectric cooling, coolers, modules, heat sinks, exchangers, Peltier coolers, devices - TE Technology

Refer to #11 for powering a TEC

Hopefully this helps :beer:

 

[HIMNL9]

Yes, you need a current regulator for TECs for operate them properly ..... they are, basicaly, current devices, like diodes, so once you hit their voltage limit, a small increasing in voltage causes a big increase in current (the main difference is that, being them pure junctions, and not PN junctions, instead of a "threshold step", the working parameter is a curve.

Anyway, when you decide to use a TEC, always check for the maximum current, and adapt the driver to your needs basing on this value as maximum value (and NEVER run a TEC at full current without a heatsink on the hot side ..... the cold side can give you a maximum of 65 / 70 degrees of DIFFERENCE from the hot side, but nothing prevent the hot side to go hot enough for melt the junctions solder, if not heatsinked )

 

[Cyparagon]

It'll work fine off of 12VDC. All of the tiny fridges I've seen use a TEC directly hooked to 12V. 12V will give you near the full cooling power, but if you don't need that much power, it just makes cooling the hot side more difficult. If you want more control, you can drive/modulate it.

 

[HIMNL9]

Yes, fridges are directly powered from 12V, but mainly cause those TECs are made on purpose for have their maximum current at 13.8V (max charged voltage for a 12V car battery) ..... and also, they have usually 2 safety thermal switches, in serie with the TEC, both on the hot side (one is a bimetal self-rearming switch set usually at 80C, and one is a one-time thermal fuse set usually at 120C)

Never hook them directly to 12V or similar without any thermal safety switch.

 

[rhd]

So, I ignored your advice (actually, it hadn't been posted yet), and I wired my TECs (cheap $3 ebay 12706s) directly to 3x 18650s.

I sort of knew the basics. On the cold side, I used thermal adhesive to attach an aixiz heatsink/mount (no diode, this was just a TEC test) I didn't have a heatsink for the hot side, so I thermally adhered it to my metal kitchen sink... figured "this is a giant piece of metal, it should be able to dissipate a lot of heat"

Surprise surprise, this wasn't solid logic. Too much for these cheap TECs and/or my kitchen sink to handle. After getting really cold for about 30 seconds, the TECs failed, and just started getting warm on both sides.

But here's what blew me away - about two minutes later (after disconnecting them of course), the TECs would come back to life, and let me torture them all over again.

So apparently even these cheap ebay TECs must have some built in thermal protection. I'm not sure then why they actually started getting HOT (both sides) after failing. Presumably thermal protection would just cut the circuit all together, so why the all-around heating effect after thermal failure? Regardless, I'm pretty impressed. It's a neat feeling to just hold this tiny plate between your fingers, hook up the power, and feel one finger start to burn, while the other freezes.

I'm looking forward to learning how to actually implement on of these. This is ridiculous and absurd, but I think I'm going to try putting these in a portable similar to my "Blue Ninja" build. It would be nice to find some inexpensive TECs that are smaller than the 40x40 mm eBay cheapos.

 

[HIMNL9]

Well, the main reason for which the TEC started to get hot both sides, is that a TEC is a heat PUMP, not a heat adsorber, and can only grant you a DIFFERENCE of temperature between the two sides, not a specific temperature ..... also, it need to dissipate from the hot side BOTH the heat pumped away from the cold side AND the heat generated in the process (yes, it generate heat, pumping it away, as any machine)

So, basically, if you have, say just as example, a 40W TEC capable, that work with 12V at 3A, it mean that it can pump away from the cold side 40W of heat, and doing this, need to dissipate from the hot side 40 + 36 = 76W of heat (the one pumped away, PLUS the 12V*3A=36W generated in the working process)

Also, a TEC usually can grant you a maximum of 65 / 70 degrees of DIFFERENCE from the hot to the cold side, at full power working range, so it's obvious that if you keep the hot side, say, at 50 degrees, you can reach 15 or 20 degrees under zero, but if your hot side is not heatsinked enough and go over 70 degrees, also the cold side start to heat ..... if your hot side go at 150 degrees (easily reached at full current, without proper heatsinking), your cold side go to 80 degrees .....

The lucky part is that, usually, cheap TECs sold on ebay and similars have big junctions, that are difficult to damage, if not seriously abused.

 

[rhd]

Interesting -

So really, there wasn't any thermal protection kicking in. Rather, the heat generated by the whole process just got to the point of overwhelming the entire unit. It was still running, but the heat couldn't go anywhere, so it engulfed the whole unit. That's why when I disconnected, and let the unit cool down, it would work again (briefly, until the same thing happened)

Is that a reasonable understanding of what was going on?

 

[chipdouglas]

yes you can direct drive it. you will get pretty much full cooling/heating. with prober heat sinking. also you can use a driver for keep a cool temp around lasers that won't cause condensation... i hear dr.lavas flexmod works well with tec's.

michael.

 

[HIMNL9]

Interesting -

So really, there wasn't any thermal protection kicking in. Rather, the heat generated by the whole process just got to the point of overwhelming the entire unit. It was still running, but the heat couldn't go anywhere, so it engulfed the whole unit. That's why when I disconnected, and let the unit cool down, it would work again (briefly, until the same thing happened)

Is that a reasonable understanding of what was going on?

Yes, basically is that what is happened.

Luckily. you disconnected it before the temp reach the damage point .

 

[Trevor]

So, I ignored your advice (actually, it hadn't been posted yet), and I wired my TECs (cheap $3 ebay 12706s) directly to 3x 18650s.

I sort of knew the basics. On the cold side, I used thermal adhesive to attach an aixiz heatsink/mount (no diode, this was just a TEC test) I didn't have a heatsink for the hot side, so I thermally adhered it to my metal kitchen sink... figured "this is a giant piece of metal, it should be able to dissipate a lot of heat"

Surprise surprise, this wasn't solid logic. Too much for these cheap TECs and/or my kitchen sink to handle. After getting really cold for about 30 seconds, the TECs failed, and just started getting warm on both sides.

This is the best three paragraphs I've read in a while.

-Trevor

 

[RogueAgent]

Hello everyone!

I am brandy new here... I was searching Ebay and came across a listing for Thermoelectric Cooler Peltiers and a little before that I was searching YouTube on how to build a laser... put 2 and 2 together and this has gotten interesting!

I found this site: Thermoelectric cooling, coolers, modules, heat sinks, exchangers, Peltier coolers, devices - TE Technology

with small circular TEC's but they seem pricey and I don't plan on killing any components on my first build.

I had a CPU cooler that had a TEC unit on it... but it was also cooled by a fan... how can we cool the hot side of these beasts? Oh, something else that might interest you.... check out "Mineral Oil Cooling." Mineral oil is non conductive and would not short out any electronics.... hell I wouldn't want it on my batteries all the time... but im sure I can section that off a bit...

I am very interested in what is said here and will continue to watch for progress

 

[HIMNL9]

^ It's a common mistake to confuse "cooling" and "thermal exchange" .....

TECs are not "cooling devices", they are "heat exchangers" ..... same as mineral oil is not a "cooling fluid", it's a "thermal exchanger fluid" ..... the mistake is, usually, think that the TEC, or the mineral oil (or anything other used for exchange heat) "produces" cold ..... WRONG ..... they just carry more (or less) efficently the heat from one place to the other .....

But always, at the end, you need to DISSIPATE the heat, and this can be done only with irradiation or thermal exchange with something that is not part of the cooling circuit (air, usually, cause it's the more common thing )