What easily acquirable material changes its resistance when heated?

I'm trying to build a Resistance Thermometer that varies current/voltage output when heated. It has to be DYI, I don't want to buy anything done, only something like a light bulb, and this has to be easily changed, as the ideal temperature range I want to provide to it is 0-100 oC. I understand all resistors change their resistance with temperature, but I'm looking for something that changes it a lot more than the regular hardware store resistor.

• Your question contains an answer: A tungsten filament light bulb - if you can figure out how to heat that filament without breaking the bulb :-) – Anindo Ghosh Mar 3 '13 at 4:55
• @AnindoGhosh I thought about using it, actually, that's why I wrote lamp. But Bobbi's answer made me look up the table of materials and turns out iron is more appropriate. Now I'm wondering where will I get low impurity iron. Thanks, though. – Henrique P. Mar 3 '13 at 5:09
• For a reasonable alternative to high purity iron wire, consider steel guitar strings, the thinnest ones i.e. high notes. There are several materials used for these strings, you would want the ones with highest change in resistance pre degree. Also worth considering are titanium steel guitar strings, they have a higher base resistivity, so less current consumed per Volt. – Anindo Ghosh Mar 3 '13 at 6:47
• Consider nichrome en.wikipedia.org/wiki/Nichrome – Optionparty Mar 3 '13 at 14:38
• What's wrong with a normal thermistor? That's exactly what they are for. – Olin Lathrop Mar 3 '13 at 16:30

You are looking for the temperature coefficient of resistivity. Iron is pretty high, I just looked up 0.0147 at 500C. Copper, in my table, is 0.0042

I use iron wire for a high current load resistor, and it is annoying how sensitive it is to heating.

Here is a link explaining it all.

• Thanks, I did some reading and loved the table you lead me to. I'll look up where can I get low impurity iron or something like that. If you have any ideas, let me know ;) – Henrique P. Mar 3 '13 at 5:10
• I was just using fencing wire. No idea what was in it. – Bobbi Bennett Mar 3 '13 at 5:39

You might be able to make a homemade thermocouple. If the range you really want to measure is 0-1000C, that's about the only way to do it. This is really more of a materials science question.

• I think his range is 0 to 100 degrees C. But he used an "o" in place of the degree symbol. – The Photon Mar 3 '13 at 15:33

It's not obvious where you draw the line on "buy anything done". Even if you buy a piece of iron wire, it's been mined, refined, alloyed, drawn, heat-treated, etc. That's a lot of "done".

So why not just buy a linear thermistor like this one or this one? You will still gain the experience of designing the signal conditioning, ADC, and power circuits to use with it.

If the parts I linked don't quite meet your needs, be aware that there are several types of thermistors. Some are designed to have a linear response for sensing purposes like yours. But others are designed to have a strongly nonlinear response for particular control purposes (like over-temperature protection or inrush-current limiting).

And of course there are both negative tempco (NTC) and positive tempco (PTC) thermistors. You could use either one with some adapation of your software or other circuits.

Also I found that many of the linear thermistors available have a maximum operating temperature of 60 C or so, but the ones I linked is spec'ed for 125 and 175 C.