Where can I learn the basics on selecting and using wide range thermocouples?

[cdev]

I want to get a better idea of the exact temperature of an aluminum block for reflow and also of the internal temps in a bench power supply I want to redo the fan on, and also probably, soon, a toaster oven so I can remodel it to be a temperature controlled reflow oven. All fairly common uses which I hope means I can use the sme brand and type thermocople for all of them..

Are any extra pieces of hardware needed between the thermocople and - say, with a Raspberry Pi or arduino

what is the best way to do this, with minimal external components? Or maybe should I just get a cheap PID controller?

 I have a GDM-8251 bench multimeter that needs a thermocouple and an adapter to use banana plugs.  (The distance between its plugs is around 2 cm/20 mm)

Can I satisfy these all with one thermocouple model (Type k?) and simply buy a few of them?

(and an 20 mm adapter?)

( so that whatever I use can be swapped around and reused elsewhere as needed)..

Sorry to ask such a dumb question.

[pieman103021]

I want to get a better idea of the exact temperature of an aluminum block for reflow and also of the internal temps in a bench power supply I want to redo the fan on, and also probably, soon, a toaster oven so I can remodel it to be a temperature controlled reflow oven. All fairly common uses which I hope means I can use the sme brand and type thermocople for all of them..

Are any extra pieces of hardware needed between the thermocople and - say, with a Raspberry Pi or arduino

what is the best way to do this, with minimal external components? Or maybe should I just get a cheap PID controller?

 I have a GDM-8251 bench multimeter that needs a thermocouple and an adapter to use banana plugs.  (The distance between its plugs is around 2 cm/20 mm)

Can I satisfy these all with one thermocouple model (Type k?) and simply buy a few of them?

(and an 20 mm adapter?)

( so that whatever I use can be swapped around and reused elsewhere as needed)..

Sorry to ask such a dumb question.

This is a good and very valid question. Most thermocouples output an analog signal, so you will need at the very least need an ADC, then implement a LUT to convert the output of the ADC to a temperature. You would also have to do something called cold junction compensation, which there is abundant information on through Google.
To make all of that easier, there are chips which do all of that and output to i2c or spi, which I highly recommend doing. One example of this would be the maxim MAX31855.

Do some more research, and decide what will work best for your applications.

[Vtile]

Thermocouples measure the temperature difference between the junctions of the thermocouple element, so if you need a true temperature with static reference point, then you need to make not one but two temperature measurements. That is what the cold junction compensation really is, measuring the "ambient" temperature and mathematically taking it account to get real temperature of the hot(or frozen  ) junction.

Type K should be as good as any other standard type for your application, by far the most used (easy to get) one as far as I know. The problem with any thermocouple is the non-linearity and relative unaccuracy. Again something you need to be aware of to get correct readings (might not be needed if aprox. is working).

[jmelson]

Thermocouples give VERY small voltages at modest temperatures.  You get a voltage differential of about 41 uV per degree C with type K.  So, for a reference junction at 20 C, and sensing something at 220 C, you get about 8.2 mV.  So, tiny voltage offsets will totally swamp the measurement.

You can get simple temperature controllers on eBay, and just ignore the control function and use it as a readout.  (Or, you can use it as intended, with a solid state relay, to control the heater.)

You have to be crazy, in my book, to design a thermocouple readout circuit, it is just a HARD problem due to the tiny voltages involved.

Jon

[Vtile]

Take a quick search about RTDs, for your application Pt1000 might be pretty good device. It is easy implement and as a standard part relatively cheap and do have much better linearity than thermocouples over the whole measurement range. Easiest is to find something like LM35 with proper operational range.

For the original question on the tittle: https://en.wikipedia.org/wiki/Thermoelectric_effect

[cdev]

Thermocouples give VERY small voltages at modest temperatures.  You get a voltage differential of about 41 uV per degree C with type K.  So, for a reference junction at 20 C, and sensing something at 220 C, you get about 8.2 mV.  So, tiny voltage offsets will totally swamp the measurement.

You can get simple temperature controllers on eBay, and just ignore the control function and use it as a readout.  (Or, you can use it as intended, with a solid state relay, to control the heater.)

You have to be crazy, in my book, to design a thermocouple readout circuit, it is just a HARD problem due to the tiny voltages involved.

Jon

I can see how that would be a problem. At  this point I am just trying to get an idea of the lay of the land, so to speak.

The several applications I want to be able to use whatever i end up with on are all "hot" applications, where there will likely be significant amounts of heat over ambient that I want to control.


I'm surprised that so many multimeters offer it, under the circumstances, actually, and that they have reasonable accuracy.