Identifying a thermistor?

[Things]

Hi all,

I've got a laser module here that has died, and inside it contains 2 TEC's for regulating both the diode and crystal temperatures. Each heatsink also has a small temperature sensor embedded in it, but I can't see any numbers or anything.

However, naturally your first guess would be a thermistor, and sure enough, it's resistance changes dramatically with temperature.

The resistance decreases with heat, so I think it's an NTC type. However what I'm not sure of, is how I can actually get an accurate temperature reading from it, without knowing what model it actually is. I've had it just sitting here on my bench for a while, trying to let it settle, and it seems to be reading around 9.9K (Although still slowly rising), and using an IR thermometer (Which isn't the best, as it's an aluminium heatsink), I get a reading of about 26C. Does this seem on par with a typical 10K thermistor?

Just trying to figure out how I could accurately get temp readings from these knowing basically nothing about them - Thermistors being simple devices I'd imagine there's quite a bit of variability between them.



Cheers,
Dan

[bingo600]

Aren't those 10K NTC's spec'ed to 10K @25C
I'd say that fits your measurement.

Is the NTC that brass thing ?

/Bingo

[KJDS]

I'd put some black tape over the heatsink before measuring it's temp with an IR thermometer.

[SeanB]

Just makr it with a black marker, permanent or dry wipe.

[muvideo]

Hello from what you measured it should be a 10k NTC, but it's not enough, you need also "beta" to
have the temperature-resistance relationship. In that applicaiton, probably it's a good quality
part, you could caracterize it measuring it's resistance to 2 or 3 known temperatures and then use
it for "precision" work. How  much precision depends on your ability.

The best way to model the NTC is the Steinhart–Hart equation, that needs at least 3 test temperatures
to obtains equation parameters (3rd order), but once obtained it will be a very accurate model of the NTC,
this means that you could convert the NTC resistance to temperature with less than 0.1°C error, depending
on your meter and setup accuracy you could do much better.

A lower precision but simpler method is to use the beta equation, this is a second order approximation.
You need 2 known temperatures, for example 0°C and 100°C easy to obtain from distilled water, measure the
resistance at these temperatures and work back the equation parameters,  in your case will be beta and
R_T=0 or R_T=100, but in general are beta and R_T=25. From your values
you could calculate R_T=25, and check how much is the predicted resistance at 25°C, if it's near
10kohm (+- tolerance) probably you have the right parameters.

It's easy to find the equations details and calculators online (check also murata, vishay and wikipedia), but
if you want more details feel free to ask 

Fabio.

[Things]

Thanks all, I hooked it up to an Arduino with the usual voltage divider setup and ran some random code I found online, and do get readings very similar to the ambient temp. The actual temperature reading itself isn't critical, as long as I can tell something isn't getting too hot, stability is the key.