For a project I would like to drive 1-2 peltier modules. For that I need to provide 0-4A of current (8.6V max) per module. I would like to set the current using an arduino. I thought, this should be something easily available, but it is not. At least I could not find any available solution.
I had a look at dedicated ICs for controlling peltier modules, but this seems a bit overkill. I don't need extremely precise temperature control. I would just like to operate the peltier module with a certain current value, which I can set with an arduino.
There are probably a lot of different routes to go, but since I'm not an electronics developer, it is not easy to know which solutions are adequate for the task. I'm willing to invest some time and develop my own circuit if needed, although I would prefer to rely on proven solutions.
I don't know which and how many components you have. But I have looked at what we have at school, and what we learned the previous year. All the components are chosen in such a way that they fit together best.
The diode doesn't have a name, because you can take any. As long as it can pass at least 30V. This component is placed in such a way that it does nothing until you switch the relay. It will cancel out all the self induction voltages from the relay.
The LED just indicates if the relay is switched on or off.
The big blue cube in the middle of the paper is the relay. It will switch the peltier element on and off.
The whole circuit just needs +15V and the Arduino signal at the +5V input. I have included all the calculations. But because I am Belgian, all the symbols and names are the way we learned them. I am in my last year of high school, so I think it isn't that difficult. And of course it can be done much better. But this is the way I learned it and would do it.
If you have any questions, please ask them. I would be happy to answer.
I would suggest adding a temperature sensor on the element, as a back-up. So the Arduino can shut down the element before it fries itself. This sensor can just go to one of the analog pins and be read by the Arduino.
If you want to choose between cold and warm, just copy the whole circuit, but take the same element and rotate it 180°.
This is my answer from another post, so it is possible that it isn't completely what you're looking for.
Thanks for your reply. No, I don‘t think it is the circuit i need. First, a relay is not an option because of the noise. And then if I understand your circuit, you can only turn on/off the voltage but you can not drive the peltier module with a specific current.
Voltage and current are equivalent on an almost constant (Peltier element) resistor (U=R*i). Effective current or voltage can be produced by PWM, or effective heat can be produced by a two state regulator.
Or search for constant current regulator circuits..
well, I don‘t use the peltier for heating. The COP of the element is highly dependant on the current and therefore in my understanding PWM ist not ideal since efficiency will be significantly lower.
True. If efficiency is very important to you, then use a constant current controller, which will be an analog circuit.
A circuit like the following is popular, and with proper choice of the R1/C1/R2/C2 low pass filter components, either a PWM signal or an analog voltage source can be used to provide the current control voltage.
Before getting more complicated, test your Peltier device with a rheostat in series to vary the current and measure the current with an ammeter. No reason to get complicated in the beginning of the project.
I have done various measurements with my lab power supply. The current varies with the temperature of the hot side of the peltier. I don't know yet how I will drive the peltier. Maybe, I will just set a certain voltage (current) and done. But I might as well look into doing some sort of PID controller...
I have looked at circuits with the LMV358, but wasted energy across R3 is quite high. I don't need 4 amps, but 3 for sure. This would already mean 3x3x0.1 = 0.9W across R3.
Thats why I had a closer look at the LMR33630 and LM5117. The latter has a current control mode, but if I understand the circuit you still need a shunt resistor where you will waste energy!
Here is the problem in a nut shell. The Peltier device is a bunch of series connected metallic diodes, as you already know from your research. Each diode requires a specific voltage level to begin conducting current. Each diode conduction voltage will be slightly different.
The sum of those voltages is the rated voltage for the Peltier device. What is the rated voltage for your specific device? That is the voltage you must supply in order for the heat movement to occur.
What happens to the voltage when you limit the current to the device?
I don't need the peltier for heating/cooling. I use it for humidity removal. This works fine, and I have tested that. But now I need something to drive my peltier other than from my lab power supply.
@ Paul
I don't know what you intend to tell.
The peltier has 8.6V, 4 Amps. If you limit the current, voltage will drop. So?
You do that by cooling water vapor, to condense it. Using a Peltier device is one the least efficient methods of dehumidifying available.
I'm reminded of one of the most misguided, poorly researched Indiegogo campaigns ever conceived -- the "Fontus" solar powered, self filling water bottle.
I have testet with currents from 1 to 4 amps. If you have a look at the data sheet, curves are given for the whole current range. I don‘t understand exactly what happens inside the Peltier, but I know for sure that it still cools
Hi,
Look at the specs for the peltier, check on the operating temperature difference between the two surfaces.
Don't get difference to high as you will degrade or destroy the device.
Need a decent heat sink to get the heat away,
I have a small calibration oven that uses a peltier, uses the circuit similar to @jremington posted in post #8, I keep monitoring the differential temperature.
Mine heats and cools.
That depends. Wasted energy in the mosfet could be much higher.
The shunt can be replaced with a hall sensor, like the ACS712-5A. And the op-amp is not needed if you use a logic level FET (connect filter directly to the gate). Feedback can be done in code.
Leo..
You could reduce R3 by a factor of 10 then use the 2nd op amp in the LM358 pkg to amplify the shunt voltage by a factor of 10 before feeding it to the first op amp.
