# peltier elements as heating and cooling

Hello guys,

I need to generate a temperature gradient from 90-80 °C to 4-25 °C. The gradient must be applied inside a device filled with very little amount of water. The length of the channel, where the water is going to flow, is 5-6 cm.

I was thinking to use two Peltier element, one as heating unit and the other as cooling (to be sure that the temperature is really lower than 25 °C.

Do you think that this is really feasible using 12V pertier elements?

Thank you very much for your feedback,

Francesco

Sure, although a resistor might be a simpler heater.

Which temperature is inside and which one is outside? What is the ambient temperature?

Most Peltiers can only generate 70C difference across the two sides. This is at ZERO heat flow. The moment you have some heat flowing into the 'cold' side, the differential is less.

I am hoping that you are not trying to cool water from 80C down to 25C in 5cm distance. Even with a small trickle of water, that is a lot of heat flow for such a small device. Do you know how to calculate the heat flow rate?

Are you trying to heat one stream of water and cool another? Then your 'hot' stream needs to be a lot larger than your 'cold' stream. A typical 12V*5A Peltier is pumping out 60 watts of heat at all times. The whole thing heats up very quickly. To get the 'cold' side down below ambient temperature, you need a heatsink which can extract 60W of heat at a 70C differential. A typical CPU-cooler size heatsink with a fan bolted to it is only just able to meet this specification. Water cooling can do better than this but in a closed-circuit the water-to-air heat exchanger becomes much bigger, just moving the problem elsewhere.

I was thinking to use two Peltier element, one as heating unit and the other as cooling (to be sure that the temperature is really lower than 25 °C.

If you reverse connections (with an H-Bridge or a DPDT relay) you'll reverse the hot & cold sides of the Peltier. If you use a relay, you'll need a 2nd relay to turn the thing on/off.

But, a resistor is a "natural" 100% efficient heater so it may be a better solution for heating.

The gradient must be applied inside a device filled with very little amount of water. The length of the channel, where the water is going to flow, is 5-6 cm.

You can calculate the amount of energy it takes to increase a given volume of water by a given number of degrees.* But, since you don't have perfect insulation, the amount of energy loss is unknown. (I don't know anything about the efficiency of Peltier devices.)

So, you might want to start by experimenting with a resistor-heater. Once you know the amount of energy you need, you'll have a better idea of your Peltier power requirements.

• Of course, the same amount of energy must be removed to cool by the same amount.

Dear all,

first of all sorry for the delay and thank you very much for your answers, I had to submit my thesis in these weeks so I was quite stressed and busy. I paused the project in these weeks for this reason.

For what everyone wrote, you suggest the use of a resistor instead of a peltier element. My choice on peltier was because I can find many peltier on the market, but not as many heaters. Therefore, [u]if you could suggest a suitable resistor[/u] I would really appreciate it.

@MorganS. The device is microfluidic, the channel has nanoliters of waters as total volume, while the two chambers at the extremities have less than 1 microliter. I worked before with microheating resistor elements and I can have a full gradient from 40 to 100 degrees in few cm. The microfabrication of such heaters is expensive for the first test, so I wanted to use cheaper solutions. Unfortunately, physics and electronics are not my strengths, I would appreciate your help for the feasibility of the project.

Thank you again very much for your kind and precious help

Unfortunately, physics and electronics are not my strengths

Have you got any idea how much power (Watts) you need?

I assume you know how to calculate power dissipation in a resistor?
Power (in Watts) = Current (in Amps) x Voltage = Voltage2/resistance = Current2 x Resistance.

[u]This style[/u] of resistor is good as a heater because you can easily mount it to something metal and get a lot of heat transfer to the metal (if that’s what you want). Or you can let it heat-up the air, but obviously you don’t want to submerse the electrical connections in water. This type of power resistor is available with ratings of 5W and up. Of course, you can use one with a higher power rating than you need.

Ok, if it's only nano liters then you might have a chance. The primary cooling problem is still getting heat away from the hot side of the Peltier.

Do you know how many watts of heating you need for the hot spots? Any resistor has a power rating that can be used. For example the standard small resistors used in most electronics are called 1/4 watt because they can heat air at 0.25W. You may need an unusual shape, but there's lots of options out there.