Good morning,

I am currently in the process of building a circuit to excite an PRT (100 ohm @ 0 degC) with a constant current.

This current should be < 1mA and remain constant regardless of load resistance.

I have found two circuit designs;

• I've seen the use of a LM317 but this seems to have a minimum current output of around 2 mA - do current regulators go lower than this?
• Otherwise I found a design which used two op-amps with a precision referance voltage (so a voltage regulator would be used instead) to generate the constant current.

My question is really should I use an integrated current regulator with a low output current or regulate a voltage (~ 2.5 V) and convert this to constant current?

The common method of using a bridge circuit will not provide adequete accuracy for the scope of the measurements.


Thanks in advance,


Peter

 

The common method of using a bridge circuit will not provide adequete accuracy for the scope of the measurements.

What kind of accuracy do you require?

 

Take a look at an LM334 -- that might suit your needs.

 

What kind of accuracy do you require?

It's really just to satisfy my project requirements, I need the transducer to be as linear as possible over a 0 - 100 deg C range.

I am performing dynamic compensation on the circuit to try to negate the delay in heat transfer through a thermowell. The more linear the response, the less error in my estimations.

I expect it to be non-linear to some extent; ways of correcting that error will be discussed, but they do not really come under the scope of the project.

Hope that makes sense

@ someonesdad - Thank you for the suggestion, I am reading through the data sheet at the moment.

Thanks again,

Peter

 

You may find that the Omega website has good information - http://www.omega.com/temperature/Z/TheRTD.html

 

A current regulator diode might work for you. There are lots below 1mA such as 1N5283 to 1N5296. You will still need to provide a voltage to bias everything.

Regards,
Ifixit

 

Thanks for the note, ifixit -- I had never heard of one of these. Here's a link where you can see the i-V curve.