# Help Understanding Thermistor Datasheet

Joined Apr 15, 2009
133
Good Day All.

I am trying to use a program called EUREQA to help plot the non-linear curve for my thermistor's.

But due to this being the 1st time of my using thermistor's I don't fully understand what is being said in the datasheets.

The thermistor's I have are Epcos B57861S0103J40, and here is the datasheet.

Best Regards.

#### MikeML

Joined Oct 2, 2009
5,444
Are you trying to use EUREQA to fit an equation through the resistance vs temperature data in the spec sheet? If so, I have had good luck using the curve fitting built-into EXCEL to do this...

Joined Apr 15, 2009
133

I am trying to use EUREQA to (1) plot the temperature vs resistance curve, and (2) show me the best equation that explains what is going on.

But I am having grate problems just understanding the datasheet.

What I have been able to work out so far is my B57861S0103J40 is a 10k +/-5% Resistance tolerance.

I am looking at the table on page 6 of the datasheet, under R/T No 8016.
I can see that I have all the temp ref there but don't understand, RT/R25, a (%/K).

Can someone spare the time so explain what they mean?

Best Regards.

#### boriz

Joined Jul 16, 2009
48
I think...

R25 is the resistance at 25 degrees c, the temperature at which thermistor values are usually quoted. IE: A 4K7 thermistor is one that has a resistance of 4K7 at 25 degrees c.

Rt is the resistance at temperature t.

So if t=25, then Rt/R25=1 (You can see this on the chart)

%/K Im less sure about, but it looks like an error or deviation from the ideal curve in percent Kelvin. Not particularly useful information I suspect.

So the Rt/R25 curve gives you the shape of the temperature verses resistance graph without offering any specific resistance information. For that, you would need to know the R25 for the thermistor. Then you can calculate the predicted resistance for any particular temperature.

For your Eureqa task, (I think) you only need a list of temperatures and a list of the corresponding resistance values. Once you know the R25, you only need the numbers from the first two columns: T and Rt/R25.

EG:
If the thermistor is a 4K7 type (at 25c), then the R25 is 4700. So using the first column we find a temperature, say 55c, then its corresponding Rt/R25 value. Using the first table, thats 0.3417. Now multiply this by 4700, giving 1601.29. So the resistance of that thermistor at 55 degrees c is 1601.29 ohms.

Choose the appropriate table for your thermistor, work out a table of temperatures versus resistance, using all the data from the table or just a sub set, put that into Eureqa.

I think.

#### thyristor

Joined Dec 27, 2009
94
Thermistors follow an exponential curve where R = Ae^(B/T)

R is the resistance at any temperature T

B (the Beta factor) will be given in the datasheet, usually some figure like 4000 say.

T is the temperature in degrees Kelvin and

A is a constant that can be calculated by filling in R25 (the given resistance at 25degC ie: 298degK)

You will end up with an equation like R = 0.02e^(4000/T)

Plot this and you have your thermistor curve

Joined Apr 15, 2009
133

The datasheet now makes sence.

Best Regards.

#### Darren Holdstock

Joined Feb 10, 2009
262
It's the Steinhart-Hart equation you're looking for. It's a very good model for a thermistor in its 4-coefficient variant, and still pretty good with the more widely used 3 coefficients.

You can make your working range more linear by judicious choice of biasing resistor; a physicist colleague proved this to me once, but alas I don't have his method or the time right now to work it out from first principles. But it's eminently do-able.

Thank-you for the EUREQA link, that looks really handy. MS Excel has a handy function for producing equations from the trendline facililty in the graphs, but you have to select the type of equation first (linear, polynomial, exponential etc.) for the best fit. It does, however, give an R2 value for the confidence of the trendline fit, so it's possible to gauge the accuracy of the mathematical model. I look forward to giving EUREQA a run, it looks even better.

#### MikeML

Joined Oct 2, 2009
5,444
Here is a thermistor I fitted just the other day. The columns of resistance vs termperture (degF) was taken right off the data sheet. I only needed a good fit at 55 degF +- 10degF, so I asked it to fit a simple exponential instead of the more complex Stienhart.