Question About Current Draw From Peltier Element

Status
Not open for further replies.

alexchamp29

New Member
I have a TEC1 - 12706. Specifications detail Imax at 6 amps. I would like to run five of these elements in parallel but my bench top supply can't supply the current needed ( I'd like about 5 amps each). I do however have a 600W power supply unit designed for PCs. The PSU has a single 12V rail with max current of 46 amps.

I know that some loads such as motors/ lamps only draw the current they require in accordance with Ohm's law. Other components such as diodes will continue to draw current until the destroy themselves and require current limiting mechanisms.

My question is, which does a peltier element fall under? I suppose I could connect the element to the PSU and see if it ignites in a plume of smoke but nearly 50 amps isn't something I want to take guesses with. Any help would be appreciated.
 
My question is, which does a peltier element fall under?
It looks pretty much like a resistor, albeit one whose resistance varies as a function of the temperature difference between the hot and cold sides. The TEC1-12706 data sheet shows the relationship between voltage, current and delta-T in several graphs on the second page.

I suppose I could connect the element to the PSU and see if it ignites in a plume of smoke but nearly 50 amps isn't something I want to take guesses with.
Doesn't look to me like you're in much danger of smoke or fire; the data sheet shows the 12706 drawing about 4.5 amps at 12 volts to produce a delta-T of 40 degrees C.
 
It depends on the temp difference on the heatpump and the max current for the heat pump.
Do you have the pumps datasheet.
The other thing is that some computer power supplies are not that keen on having all the current draw just on the one rail, as they are usually regulated on 2 or more rails, depends on the supply itself.
 
It is a single rail supply.
So I guess what I'm asking is do I need a current limiter to power a 6 amp TEC with a 46 max amp supply?
Does it draw current like a lamp or like a diode?
 
Ok thanks so much for the info.
Just so I understand a little better, is it because it is a closed circuit component and does not require a minimum forward voltage to complete the circuit?
 
Just so I understand a little better, is it because it is a closed circuit component and does not require a minimum forward voltage to complete the circuit?
I don't know what you mean by "closed circuit component" or how you think that distinguishes it from anything else electrical.

The important point is that resistors, lamps, and your TEC element are all basically resistive, which means there is a more or less linear relationship between the voltage across them and the current through them. That linear relationship might be dependent on temperature (the resistance of a cold incandescent tungsten light bulb is only about 7% of the resistance at its normal operating temperature of roughly 2800 degrees K) or other factors, but it is still linear with respect to instantaneous voltage. That is, there is no threshold below which the device doesn't conduct, and above which it does.
 
I just meant that for component like diodes, current doesn't flow unless that threshold is met, like forward voltage.
I see what you're saying though it makes sense.
Any resistive element must follow Ohm's law right?
 
I just meant that for component like diodes, current doesn't flow unless that threshold is met, like forward voltage.
Strictly speaking there isn't a threshold per se in diodes, transistors, LEDs or anything else based on PN junctions. Rather, the relationship between voltage and current is logarithmic instead of linear. In silicon diodes and in the B-E junction of BJT, the voltage across the junction will be proportional to the log of the current within a few percent, often over more than a 10,000:1 range of currents. For a diode-connected BJT (that is, with base shorted to collector), that range may exceed 1,000,000:1.

Plotted on a linear scale for current, this relation looks an awful like there's a sort of "threshold" voltage below which not much current flows, and above which current increases rapidly; but in actuality the relation is very close to logarithmic.

I see what you're saying though it makes sense. Any resistive element must follow Ohm's law right?
Correct, keeping in mind that the resistance may change, either a little or a lot, from the effects of self-heating due to the power dissipated in the element.

EDIT: I found some test data in my files for Vbe vs. Ie for a couple of diode-connected transistors, plotted two ways.

First, a linear plot:



Then plotted with current on a logarithmic scale:



The top plot shows how we get the idea that there's a "threshold" of voltage (albeit kinda soft) for current to flow through a PN junction; but the bottom plot shows the real story.
 
Last edited:
Do not connect your peltiers withour adequate heatsinks on both sides otherwise they got so hot they destroy themselves.

Mike.
 
Ok thanks so much for the info. I think I have a better understanding of ohm's law now and feel more comfortable about using muy power supply.
 
It is a single rail supply.
So I guess what I'm asking is do I need a current limiter to power a 6 amp TEC with a 46 max amp supply?
Does it draw current like a lamp or like a diode?

The supplies that I had used were either bi-polar or unipolar depending on the application.

Can you destroy a Peltier, operating it within the allowed voltages - yes,
That will depend on max temp, delta T, and I. "I" is likely to be exceeded within the operating voltage.

Exceeding Delta T is the real meltdown problem as well as the absolute temperature.

So, the device moves heat from one side to the other depending on the polarity. The device is really slow doing it,

It's great to use the device with a known heatsink, like ground water.

Condensation isn't a nice thing either.

So, operating at room temperature with a ground water sink was one way we operated the devices, This application used a controller that used a thermister. So, this was surface to water cooled surface, This actually was a commercial microscope stage with a vacuum hold-down.

We made smaller versions 1" x 1" with a vacuum hold down.

We used a large one with the sink being a fan, again to control a little higher than room temperature. 25 C would have been ideal, but we could not spare a controller, so a power supply was used open loop. This was surface to an air cooled surface.

The last application was quite odd. We needed to cool an enclosure. That was air to air.

I think there may have been some laser diode systems with thermoelectric cooler in them

Some applications were used 24/7.

They are not very efficient.
 
The problem with operating peltiers without a heatsink is the overall power input. You are adding 60W of power to the system, so the cool side may be 50C lower than the hot side but the hot side could be at 400C. To cool the cold side you need to cool the hot side by the amount you want to cool the cold side PLUS the power you are adding via the power supply. The hot side heatsink/fan may have to dissipate 110W to cool 50W.

Mike.
 
So my project is a simple cloud chamber. Cascading 3 TECs along with a 130W heatsink, I was able to reach a temperature of ~-30C. That's all good and well but I wanted to go lower to get a large saturation area in the chamber.
Trying to power three 6 amp TECs plus another that I was using as a heater, I realized that my cheap bench-top power supply couldn't provide the needed current and I still wanted to add another cooler to the cascade.
That is when I found a 600W PSU. A single, regulated supply of 12V at 46 Imax should be more than enough (4 TECs = 24 amps + the heat element at around .5 - 1 amp)
I just wasn't sure if the TECs would require current limiting resistor like an LED would or if they would only draw as much as they needed (12V/~2 Ohms@20C= 6 amps).
Everyone's information was really helpful as I am a chemistry student and feel a bit out of my element when it comes to electronics.

P.S. To give you a crude idea of my setup:

+=======+ >TEC1
------------------------------------------------- >Copper Plate
+=======+ +=======+ +=======+ >TEC2,3,4
| | | | | | | | | | | | | | | | | | >(3)130W Heatsink/Fan
 
Last edited:
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…