Having worked in the petrochemical industry for a number of years, I was very familiar with thermocouple and RTD temperature sensors. For the past 16 years I have worked with thermistors in cold storage warehouses, where fruit is stored at temperatures very close to its freezing point (30-31.5 °F). I have found thermistors to be very stable and extremely accurate, with superb resolution. They are also much cheaper than RTDs.

I don't understand why thermistors aren't widely used in industrial applications. They aren't suited for high temperatures, but they are very useful from -70 to +100°C. I understand that they have a non-linear response, but that is easily dealt with through software or microprocessors.

I'm advocating the use of thermistors for industrial applications.

 

I'm in full agreement. Within their temperature range, the high sensitivity makes a lot of things simpler. And with the processor horsepower available these days the math shouldn't be a problem. They should standardize one of the general sets of equations for linearization and vendors could provide the coefficients for their particular units. This would avoid lookups and other methods that discourage their use. It should even be possible to make an ASIC to do the conditioning.

Regards

cww

 

A lot of industrial processes are over 100 C so that is why they are not used.

James Bouchard

 

The two main reasons are probably standardisation and lack of knowledge.

RTDs are well established, well known and have an easily understood characteristic. They are relatively easier to calibrate. They also have a much wider usable range and better stability in most applications although thermistors have improved a great deal. In most process plants RTDs would still be required to cover some applications so the use of thermistors would require more spares, more knowledge etc. Standardisation is much simpler with RTDs and the cost difference in process plant applications is negligable. It might even be more for thermistors due to the extra spares and training requirements. Thermistors are also produced with a range of curves so one has to be careful in selection and interchangeability.

Having said that I agree that there are applications where I would prefer to use thermistors and it is unfortunate that there are not many industrial transmitters produced for thermistors. Their advantages are their sensitivity and in some applications their ruggedness. I am not sure how aware some people are of the fragility of RTDs.

Vince Dooley

 

hi, I was wondering if you would explain to me how exactly does a thermistor work? see, I am given a project from lecturer to design a temperature measuring device from some given apparatus, i.e. thermistors, copper wire, bla bla... do u knwo any good websites where I could find the 'correct' circuit to set up to measure the temperature???

thanks

Regards
Sutejo

 

Exactly would be another lecture or two. But a good working definition would be that it is a device with a relatively large and non-linear change of resistance with temperature. This originally was negative i.e. resistance decreases as temperature increases, but positive tempco
(short for temperature coefficient of resistance) devices soon appeared. Circuits for thermistors tend to be simple as the change needs little amplification and no compensation in most cases. For circuits, I would peruse the Fenwall and YSI sites and google "thermistor circuit". Omega may still offer lots of free info as well on their site. I tend to use a simple bridge and linearize in software.

Regards

cww

 

Industrially, thermistors represent a relatively late arrival on the temperature measurement scene. When they first appeared they had the major disadvantage of being extremely nonlinear. That is not a problem now when it is normal to have processing capability in an instrument/transmitter. They still have the problem of limitations of temperature operation for many industrial processes.

However, within their operating range and particularly for very small ranges they are the most economical selection available and excellent precision can be achieved at very low costs. They are underused in industry because a lot of instrument practitioners just don't know anything much about them and most large companies standardise where possible and RTDs are an excellent choice in that respect for precision, stability and interchangability.

Within their range thermistors have the added advantage of being extremely rugged and robust if the right ones are selected. Not many people appreciate just how fragile RTDs are.

Check the archives of this list. (How do you do that?) There have been discussions on this subject before.

YSI is a good site.

Cheers

Vince