Interpreting Thermistor Tables 1

[roydm]

I am trying to substitute a thermistor which gives me resistance values in steps of 10 Ohms from zero to 100.
0=9750. 10=5970, 20=3747 - - - - - -100=204 Ohms


When I look at the thermocouple parts lists they rate the devices in a totally different fashion Rt/R25 (related to 25°C and use some very complicated formulae to describe the slope
Typical 0°C = 2.813 Rt/25C 100°C = 0.95690 Rt/25C

How do I convert these values into Ohms so that I can compare Apples to Apples?

Thanks

Roy

 

[Skogsgurra]

I'm afraid you cannot substitute thermistors for thermocouples.
Or is it RTDs you want to use? Doesn't work either. If you do not change the electronic system as well.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[roydm]

My mistake, I meant thermistor

I am trying to source a replacement for an existing Thermistor but having trouble.
Nearly every document I find refers to a complex Steinhart-Hart thermistor equation and the tables don't say what the resistance is at any given temperature just refer to the Rt at 25°C
I'm no math whiz but how do you know what Rt is in Ohms from that?

I gather different materials have different slopes, no I don't know what mine is made of

Other lists give the resistance at 25°C and a ratio


I don't need an exact match but I would like to get close at least.

 

[IRstuff]

Unless you do get a near exact match, your results may will be garbage. The Rt is a function of what value was originally specified, i.e., you can get 2.2k, 20k, or 100k. For the particular material, every one of those thermistors will scale the same way:

http://www.veris.com/docs/support/faq/RTD-Thermistor_Z202030-0N.pdf

the table is quite enlightening, as it lists the thermistor value changes for a 10k Rt with different materials. If you're only operating in the positive temperature ranges, then you might get something like a 5°C error if you have the right Rt, but the wrong material.

TTFN
faq731-376

 

[mkeith]

Double-check the data for the thermistor you have. Even if it has a table, they often will specify R25 and a "Beta" or "B" value. R25 is the resistance at 25C. B or Beta is a parameter you can plug in to a log equation to solve for the resistance at any temp (you also need to plug in the R25 value).

Check out the thermistor entry on wikipedia. In particular, look at the B or Beta parameter equation:

R=R₀e^B(1/T-1/T0)

Note that R₀ in that equation is the resistance at T₀. Temp must be in Kelvin. Basically any R and T that match will do. So you can use R25 for R₀ in that equation, and 298.15 for T₀.

So to recap, R0 is R at 25C. T0 is 25C, converted to Kelvin. B comes from the datasheet for the resistor. If you can find R0 and B for your current thermistor, you should be able to find a cross easily. If not, you should use the equation to generate a table for some possible crosses, and compare the table to your existing part.

In principle, you could even enter the values from your table and solve for B, which would also make it easy to find a cross.

Good luck!

 

[roydm]

The sensors are for an automotive application air and head temperature for the ECU.
After looking for days I just happened on the exact match from Vechay 3000 Ohms at 25°C
The table I found listed the Ohms in 10° increments, I'm not sure why the other vendors list in such a complicated manner

Thanks for your help

 

[IRstuff]

Is doing a lookup and then an interpolation really easier and more accurate than calculating it directly?

TTFN
faq731-376

 

[roydm]

It is for me, thats for sure
I can't figure out why they make it so difficult unless it's to cover every different 25 C value for a particular material
Vechay had a table showing resistance for each 10 degrees of C.

 

[IRstuff]

Tables are inherently inaccurate for many applications, since they're based on a single thermistor. Many applications require both calibration and direct calculation to achieve the accuracies that I typically want, which is < 0.1°C.

At a minimum most tables would require an offset calibration.

TTFN
faq731-376