Go Down

Topic: Reading a 1kohm Thermistor (Read 3722 times)previous topic - next topic

WalterW

Feb 16, 2011, 05:09 pm
Hello,
New to arduino, i try to do some experiences and i have a problem:
Is it possible to read temperature from a 1kohm thermisor like this?

Code: [Select]
`5v--------ww Thermistor (1kohm) ww----|---------------------------ww10kohmww----------------GROUND                                                                |                                                                |                                                           Analog in`

I know i have to use the Steinhart-Hart equation but is my 10kohm resistor too huge ?

Thanks

liuzengqiang

#1
Feb 16, 2011, 05:13 pm
Yes the circuit is correct in theory but in practice you choose the "10K" resistor to be the value near the center of your temperature measurement to minimize digitization error. Say your measurement is around 25DegC, so your 1K thermistor is near 1KOhm most of the time. Then you should use 1KOhm series resistors instead of 10K. If you work at a different temperature range, use a different known resistor in series. As a matter of fact, my electronics students just finished their project measuring resistors and I have error analysis and curves if you're interested. We're using 10K known resistors.

retrolefty

#2
Feb 16, 2011, 05:27 pm
The main disadvantage of lowering the value of the series fixed resistor too much is that the thermistor will exhibit more 'self-heating' due to the increased current flow through it. How much this will effect the accuracy of the thermistor at the temperature of interest is a exercise good for a student to determine.

Lefty

WalterW

#3
Feb 16, 2011, 05:33 pm
Thanks, i think i'm gonna use a 10kohm thermistor for the beginning instead of a 1kohm thermistor

liuzengqiang

#4
Feb 16, 2011, 05:50 pm
You could have used a 10K thermistor though
My student project is sorting random resistors so there's no heating thermistors. In a temperature sensitive circuit, you can use thermistor to do some negative temperature feedback or compensation I suppose. you can also pass "large" current through thermistors and dip them in liquids to measure liquid levels as liquids quickly dissipate heat while air does not.