the task is to find the 96% response time for a thermistor. By plotting a graph of resistance against time for that thermistor, when i find the 96% point for that graph, will that be the same time for the 96% point on the temperature-time graph.

i would have thought that if the two times were the same that was presuming a linear relationship between temperature and resistance for a given thermsitor. However from the equations that i have found, this doesn't seem to be the case? When i plot the resistance against time graph, this is exponential decay graph, is that also the same graph for the heating of a thermistor graph (temperature-time)

Any thoughts on this would be appreciated or a website that might explain the releationshiop between temperature and resistance

 

> i would have thought that if the two times were the same that was presuming a linear relationship between temperature and resistance for a given thermsitor. However from the equations that i have found, this doesn't seem to be the case? <

Since a thermistor is a non-linear device, temperature versus resistance does not change linearly.

 

I have handled this by simply doing a single point conversion from the 96% temperature to resistance using the lookup chart or equation
from the manufacturer, and timing until this resistance is crossed. To do this with a scope (voltage) either a simple resistive divider
with a resistor equal to the target resistance (measure to 50% Ein) or a constant current source (measure to Rtarget X Iconstant) Actually any values can be used but excitation could be chosen to minimize self-heating as a cause of error. For very slow response you can use a meter. This way the shape of the curve doesn't matter. The information should be available from the manufacturer, or Omega has a good general
reference on thermistors. Fenwal and YSI less so at last look.

Hope this helps.

Regards

cww

 

Since the thermistor is nonlinear, the answer depends very much on the change in temperature you are working with.

If you make only a very small change in temperature (so the resistance change is < say10%) there will not be a significant difference between the two results. However, if you are talking about a change in resistance of about 50% then there could well be a measurable difference - but whether it is "significant" or not depends on the context.

If you can manage to get a controllable temperature difference of 1 or 2 degrees, make a test and see what you get. Then repeat with say 5 , and 10 deg changes, and look for differences. (But, dependng on the setup, you may be seeing changes in the thermal resistance from fluid to thermistor which will also have an element of temperature dependency, and will also affect the time constant ...)

Safe answer - Carefully plot resistance vs temperature for your particular thermistor. Measure resistance vs time and enter into eg an Excel spreadsheet. Do this for a number of different values of temperature difference. Convert the resistance values into temperature from your conversion chart, Plot results and determine the required answer. Report back here so we all benefit from your newly-acquired expertise. (Besides, I'd like to know the answer as well.)

Bruce.

 

As you have observed, the resistance/temperature relationship is nonlinear. Therefore, the response is defined by the temperature/time relationship not the resistance/time relationship.

For typical thermistor properties see http://www.ysi.com/temperature.htm

Vince Dooley

 

Hopefully these sites will clear up some of your questions:

http://www.vishay.com/docs/33028/geninfo.pdf
http://www.betatherm.com/resistance.html

 

i spoke to my physics teacher and he says that the times will be the same as when the voltage is at 4% of its original the temperature will be 96%.

This is a plan for an experiment so i'll post back when i have done it

thanks for the advice so far

Chris