Even though an electronic hobbyist since years, this will be my first instructable over here!

This project is not as a permanent solution for a digital thermometer, but a quick to build one just in case you cant find your dedicated one or if it stops working. It quickly displays a rough estimate, but precision values can also be obtained if you do a little math! Whats more, it uses just 2 components and costs around \$3 overall!! The only drawback is that you have to use a table to find your temperature, but as I said, this is not a permanent and dedicated thermometer.

## Step 1: Components Required

A Digital Multimeter - Just about anything will do, it just should have a 'Resistance' measurement which all of them have. A cheap one costs a little more than \$2 and the better quality you buy, the more precise the readings will be but here the cheapest will do.

A Thermistor - NTC 10K - Its an electronic component, basically a type of resistor whose resistance varies significantly with temperature, more so than in standard resistors. 10k indicates that the resistance of the thermistor would be 10k ohms at 25 degrees. You could find it online or get it from some place like radio-shack. This thing costs less than 25 cents in India.

Some wire (optional) - This is only if you need to measure the temperature of something more flexibly. You could use any wire which is not very thin, as they have a greater resistance and will affect the reading a little bit. Something which is as thick as the legs of the NTC or thicker is just fine. If you need just a feet or so of wire, you  could use thinner wires, but this wont be any problem here for now. Just make sure that you connect it such that the meter gets a good contact with the NTC.

## Step 2: Measuring

Connect the thermistor to the multimeter probes, either directly or by just inserting into the socket after removing the probe jacks. Set the Multimeter to the 20K range in ohms (20k=20,000 ohms).  The multimeter should start showing resistance readings by now, but wait for it to settle down. This variation is due to the cooling down of the heat from your hand while you inserted it. Make sure the thermistor has a good contact with the multimeter probes as a loose connection shows a higher value than what we need.

The table shown will tell you the temperature from the resistance reading on the multimeter. Find the range between which your resistance ends up and check the corresponding temperature, as simple as that!!!

You can measure room temperature, body temperature or anything u like, just make sure you don't touch the metal pins of the NTC or let it come in contact with any liquids which will ruin your readings.

If you need it to measure the temperature of something which is flat, say a metal tumbler, you need to get the NTC to have a better contact with the metal. For doing so, don't grind or remove the plastic from the NTC as the pins are pretty close to the outer regions internally, instead find a small metal piece, preferably a flat heatsink, put a small dab of thermal compound on one side of the NTC, place it on the heatsink, and then glue the edges so that it remains in place. Dont let the glue come in between the NTC and the heatsink.

## Step 3: Precision Temperature Calculation (optional)

This step is only for advanced users who need to find out the precision temperature from those reading. There are 2 ways and the original genuine way for calculation for NTC's is given below.

R = Ro exp( Beta/T - Beta/To)

where

R      - Thermistor resistance at T (K)            = Reading from Multimeter
T       - Thermistor temperature (K)                 = To be found, substract 273 from result.
Ro    -  Nominal resistance at To (K)              = 10000
To     - Temperature where Ro is measured = 25+273 = 298
Beta -  Thermistor material constant              = ~3960 for NTC 10K

However this method involves solving an exponential function which is not easy for most of us, so a less accurate second method for approximation by me has been given below.

T= T1 + (Rlow-R)/(Rlow-Rhigh)

where
T is the temperature reading we need
T1 is the min temperature reading corresponding to the range we find when looking at the table, for eg, for 9560 ohms, T1 is 26 degrees.
Rlow is the lower end of the resistance range where we find our resistance R in. It will be a value higher than R
R is the reading of our multimeter
Rhigh is the higher end of the resistance range where we find our resistance R in. It will be a value higher than R

example we get a reading 9.56 on the multimeter on the 20K range
so 9.56x1000= 9560 ohms
9560 comes between 26 and 27 degrees on the table, so T1 = 26 degrees
Rlow among that range is 9612 corresponding to 26 degrees which is higher than our 9560
R is 9560 itself
Rhigh is 9224 which is the resistance for 27 degrees and the higher end of our range

so T= 26+ (9612-9560)/(9612-9224)
T= 26.134 degrees !!
<p>Nice idea but what about power dissipation it will incrase temperature on resistor so what would be margin for error ? </p>
<p>is there any way in which the reading from the thermistor can be directly converted and displayed on LCD. If there is any circuit diagram could anyone help me out.</p><p>any help will be duly regarded. </p>
<p>Wow! It behaves to know that I can make my personal thermometer. It's extremely interesting since nowadays, illness is all around and a thermometer can help us to know. Thanks for sharing this one!</p><p>____________________</p><p>http://thermee.net</p>
Wonder if the sensor from an old pc case fan would work?<br>
probably depends on the rating of the resistance as the one used is 10k and the one in the pc case might be 100k etc hope it works :)
i looked on futurelec to find a cheap multimeter seeing as mine died \$4.90<br>but as a looked a the better module 1 dollar extra it had a thermometer with it
Yes, there are multimeters with built in thermometers, this was just for those who already have one without it, or just for informational or creative ideas.. :)
thanks :) it's funny though because the evil red glowing eyes make it look like your getting really angry but then you see the smiley face at the end and laugh
LOL.. ;)
This is very interesting, but:<br><br>1) You should better explain step 3 for those who are not engineers, i.e, me.<br>2) This device will be useful if it has a cable. What type of cable can be used?<br>3) In order to reduce reading time, can be &quot;peeled&quot; the thermistor?
1. I have made an alternative approximation method as the original formula involved solving an exponential equation which is difficult even for engineers. This method is plain addition and division. :)<br><br>2. I have revised the steps with the details on that. I didn't add it initially to make it look as simple as possible :)<br><br>3. This also has been added to the steps, and I wouldn't suggesting peeling a thermistor as the pins are closer to the edges internally and you might end up with an open thermistor. Try the other method :)<br><br>And thank you for your comments :)
Thanks for the response!<br><br>I suppose this method is only valid up to 100 &ordm;C. &iquest;There are some methods as easy as this (or so) to measure forge temperatures? It is to say, over 1000 &ordm;C. May be using thermo-pairs. But I don't know if the resulting voltage will be measurable using a cheap tester.
Yes, this will work till a max 120 &ordm;C. And to measure temperatures above 1000&ordm;C, you either need to build an IR thermometer, or use a thermocouple along with a multimeter. But thermocouples need to be calibrated for such uses and devices to measure above 1000&ordm;C are very costly. You could try the IR thermometer type, but I doubt if its easy to build. Also how about those tiny needle type thermometers which come on old ovens?
I thought so. It's really not so urgent my need, I will find out near my house if the need arises. Thank you very much.
Cool, I love when things are made simple :-)
Thank You :)