Graph of Ln(R) against 1/T

1. Mar 30, 2013

lionely

The resistance,R, of a thermistor varies with absolute temperature according to the following equation:

R= R0 ek/T

I was asked to find the constant k using the graph.

I put the equation given in the form

ln R= k/t + lnR0

k would be the gradient. So I found the gradient from the graph and I got 2968.80

then I was asked to find Ro

so I used a point off the graph and the gradient

point (2.88,4.3)

4.3 = lnRo + (2968.80/2.88)

=> lnRo = -1.031 x 106

then I took anti logs and got Ro = 0 . Which i think is wrong...

Guidance appreciated.

2. Mar 30, 2013

TSny

Should the 2.88 be multiplied by 10 to some power?

Do you really want to divide here?

3. Mar 30, 2013

lionely

Oh yeah the 2.88 should be 2.88 x 10^-3

and yeah wouldn't I divide? Isn't it k/T?

4. Mar 30, 2013

TSny

k/T says to divide by T. But you're graphing 1/T, not T.

5. Mar 30, 2013

lionely

oh I see thank you.

6. Mar 30, 2013

lionely

I got Ro to be 0.014

7. Mar 30, 2013

TSny

That looks ok. But, I find the graph has a slope (gradient) closer to 2850. That would change the result for Ro somewhat.

8. Mar 30, 2013

lionely

Oh what did you use for your values to find the gradient? I used (3.52*10^-3 , 6.2)

(2.88 * 10^-3, 4.3)

9. Mar 30, 2013

TSny

When finding the gradient, it's best to pick two points on the line rather than picking two data points. The data points generally don't lie directly on the best fit line. Also, it's best to pick two points fairly far apart.

I picked the two points shown in the attached figure.

Attached Files:

• Graph.jpg
File size:
18.4 KB
Views:
107
10. Mar 30, 2013

lionely

ohh okay thank you

11. Mar 30, 2013

TSny

Also, note that you cannot expect to get a gradient more precise than 3 significant figures for this graph.

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted