Thermoluminescence device - general planning.

Hello everybody,

Let's start with a quick introduction: I am a chemist currently doing a PhD in biophysics. So we have this old broken Thermoluminescence device that came from some arctic biology ship expedition thing (in the 70s from the looks of it). However it is not working anymore. Buying a new one is not an option as there are only one or two companies making them and these don't have the specs needed for the science we would be doing on it. Basically all TL-Devices used in the Field where built by one guy that now has pretty much retired... (Also its more of a side project to my PhD so buying a 20000£ thing is not gonna happen)
So what I want to do is build a new one using some of the old parts we have, some new stuff and naturally an arduino.
As up to now the only thing I can do with an arduino is make some led's flash, and I am also not very good at electronics yet. However I can solder stuff and with some help I think I can put together all the stuff thats needed.
I won't bother you too much with the science of it, but just a short description of what the TL device will have to do in the end. With thermoluminescence you take a sample and heat it while watching for the light it emits in dependence of the temperature. In biology you do that with leaves or cyanobacteria for example and it gives you information about Photosynthesis and energylevels and populations and so on.
So the device has to be able to run various linear temperature gradients in a very defined way from about -70 to + 80 °C. I will use a peltier to do this (however the peltier doesn't need to be able to cool all the way down to -70 but it needs to withstand -200 C without breaking as I will use liquid nitrogen to cool the device down initially and flash the sample.) Flashing the sample will either be done by some high power LED's or a xenon flash. Detection will be done by a (cooled) photomultiplier. Temperature detection I want to do for now with a K-type thermocouple. (I know it doesn't go down to -200 but I only need to read the temperature as it warms up to about -70 so it should be ok for now, I don't want to use another type of thermocouple as there are no amplifieres readily available for the other thermocouples (or rather expensive) to my knowledge and I don't want to get into making that myself just now). Naturally the device has to be controlled by a computer ie when to flash, heat, hold record and so on. So that is the general project.

So there is a lot of stuff I don't know how to do just yet but I want to begin the project by connecting a thermocouple and a peltier padded with some copper blocks to an arduino and program it first to hold a certain temperature on one side and then do some gradients as well. Which I think will be a task that should be within the reach of my abilities. (also i should be able to use this with another TL device we have at the moment to see if that part is working before building the rest.)

So for that here is what I have come up with so far:


The only thing missing at the moment (as far as i understand) is a driver for the peltier as it draws to much power to be connected directly to the arduino.
So as It will need to do both heating and cooling, I thought of a H bridge for a DC motor might do the trick but most I found are only for up to 3 amps (also the motorshield doesn't deliver enough current)
Would something like this do the trick?

or do I need something like this?

or something completly diffrent?

So if someone could help me out on this that would be great... I think if someone where to point me in the right direction I can come up with how to connect it up using the datasheets and some logic but help is always appreciated. In any case I would draw up a scheme and let someone look over it if anybody would care to do so.
Coding wise I found a tutorial here and a description there and would try and see if I can put the pieces together. I am sure I'll run into some problems at some point but I'll see what I can figure out on my one until then.

A one more thing. I chose this powersupply (see list) because it delivers 15V 6.6 A which is close to the optimum requirements of the peltier. However It might be smart to by a bigger one to also drive the uno (or not ?) and other stuff that gets added on later like high power LEDs or xenon flash or even the PotoMultiplier (although i don't know what that draws atm and might want to have that on a diffrent circuit for changeability and so on) . so if i were to buy something like this
and put a potentiometer on it or a resistor of 1.3 ohms (if I understand ohms law correctly) what would happen. Am I now just dissipating some power as heat through the resistor or could I damage stuff somehow? (just pwm it with a mosfet to the peltier doesnt seem smart to me as the peltier would be seiously overloaded by the 24V and not work properly)
I think these are all very basic questions for people with more electrical engineering skills so sorry for not beeing smarter.
Well I hope you can making something out of this and I am happy to answer all questions you might have...

Cheers and thx for your help already.


Welcome Sven,

Very interesting project. Think you are well prepared, I would consider is using a separate power for the "high voltage" stuff and the Arduino and use optocouplers to keep them separated. One wrong wire can blow an Arduino and although they are not too expensive, it would cost you extra time.

Furthermore adding an LCD screen to show the temperature & optional parameters, optionally fancy red warning LED is system gets too hot/cold .

Sorry I cannot help you with the H bridge that can handle 7 Amps.

Hi rob,

thanks for the tip with the optocouplers! that might be a very smart idea. I didn't even know that those things existed.
As for flashy LED's and LCD's and all that other stuff that emits light, I know they are cool and easy to integrate with arduino but it's very unlikely that I will use any of this. As the experiments and equipment is extremely light sensitive I cant use anything that emits to much light. i will even have the arduino in a sealed casing to get rid of the on Board LED light.


Hi Sven,

very interesting your project and as Rob said you've prepared it pretty well. I am doing a similar project to control a Peltier cell to cool down an imager (digital still camera) used for astronomy photyography, so I am running in very similar problems as your.

Based on my project I'd like to make few comments on your post:

  1. Do you really need an Arduino UNO? Could you possibly consider to use an Arduino MICRO? The advantage is that you may supply Arduino with the USB of the controlling computer keeping separate its supply from the Peltier supply.

  2. An inexpensive Power supply that you may consider is the one inside desktops computers! They have a +/-12V @ 8A, +5V @10A and you can easily unmount it from an old desktop... (yeah that is the power supply of my project).

  3. H-Bridge is perfect for the bidirectional control that you need. The pololu board looks to fullfill your needs. You may implemented it using discrete NMOS, but if you're not very skilled in electronics I wouldn't try to make an H-bridge to save 40$ ;o).

  4. To develop the Arduino code I would suggest you to purchage the O'Relly books "Arduino Cookbook", there are plenty of practical examples inside that I think it will be very handy for you.

  5. Have you already consider how to develop the PC code to control the Arduino? Do you need to control other instruments with the same PC (e.g. via GPIB or so)? Since I am developping the same kind of application I can possibly share some C# code with you.

I hope you may find those comments useful....

1 Like

astronomy photyography,

can you share a nice one?

Hi Rob,

Unfortunately I am not passionate enough to freeze my butt all night long in the mountains during the winter to shoot pictures, I am just providing electronic support to them when they need.
Nevertheless, if you’re interested, the association publish some shoots here: (sorry for the italian website you have just to click of the pics to zoom them).


looks good, thanks!