Proper way to measure thermistor

[Rhybot]

Hello Phidget Forums,

I have a 2-wire 2252 Ω thermistor, and am trying to use Phidgets to measure the thermistor's temperature.

I have the TMP1200_0 Phidget right now. I wired it like a 2-wire RTD (per the TMP1200_0's User Guide), with:
1. EXC+ jumpered into RTD+
2. EXC- jumpered into RTD-
3. Thermistor+ going into RTD+ as well
4. Thermistor- going into RTD- as well

Via the ResistanceInput API, I set the RTD type to be RTD_WIRE_SETUP_2WIRE.

Is this the proper way of measuring a 2-wire thermistor? Is there a better Phidget I could be using, or a more standard wiring method?

Just checking, it's not super explicitly stated in the TMP1200_0's User Guide.

[KUROCHAMAME]

I have a 4-channel Phidgets 1048 already but I would like to change the device to 2 or more of the Phidgets TMP1200s to avoid noise coming from the moters.
Is it possible to connect two of more of the TMP1200 to the Vint Hub and use it with Artisan?
Do I need to buy some special probes for the TMP1200?
Or simple K type probe is OK on the TMP1200?

[jdecoux]

@Rhybot - Yes, that sounds like the proper way to measure a thermistor with Phidgets.

---------------------------------

@KUROCHAMAME - The TMP1200 is an RTD/resistance Phidget. It is not intended to measure thermocouples, but instead for use with RTD probes. You can find RTD probes linked in its product page under Connection and Compatibility.

It sounds like you could be looking for the TMP1100 Isolated Thermocouple Phidget, which will use K-type thermocouple probes. Yes, you can connect multiple TMP1100 (or any VINT Phidget) to the same VINT Hub. I cannot speak to how these are used with third-party software such as Artisan.

[luneart]

Probably way too late for you @Rhybot, but for other people like me who have similar issue:
"2-wire 2252 Ω thermistor" -> because 2252Ω (at 25°C) is a standard for NTC, I expect your thermistor is not a RTD, but a NTC (Negative Temperature Coefficient). The law between temperature and resistance is quite different!

RTD mesures the resistivity of a metal, resistance increases with temperature: Rt = R0[1+ α (t-t0)] => resistance is an affine function of temperature, eg ~ proportional.
NTC mesures the resistivity of a semiconductor, and as the name says, the resistance decreases with temperature. Temperature-resistance is accurately described by the Steinhart-Hart polynom:
1/T = a0 + a1·ln R + a3·(ln R)**3
2 points about that law:
* T is in KELVIN!! (substract 273,15° to the result for celsius degrees.)
* you can also find this law without the **3 term for either simplicity's sake (especially if low computing power available) or because manufacturer only provide 2 points (generally 25°C & 100°C, eg 298,15K & 373,15K), or with an additional **2 term.
* manufacturer will often provide a table of temperature/resistance points instead of these a0-a3 coefficients. Write the table in a spreadsheet, compute the equivalent table for 1/T and ln(R), and do a polynomial regression on the 2nd table.

The TMP1200's RTD API is only for... RTDs.
But you can use the resistance input API instead as @Rhybot said, and do the conversion in your program.