RTD Sensor Quality

Comments

• grim

  27 March 2015 01:43PM edited March 2015

  RTDs come in different flavors, each flavor represents the accuracy. These are defined standards you can look up:

  • Class C - Pass (you rarely see these anyway, this is where rejects go)
  • Class B - OK (Generally used in CIP or non-critical applications, still better than thermocouples)
  • Class A - GOOD (Pretty much the defacto standard when talking about RTD accuracy)
  • 1/3 DIN or Class AA - Better (These are rarer, as they are costly)
  • 1/10 DIN - Best (Very rare to find these out of specialized applications)

  If they are advertised as Class A RTD, they should meet the standard of Class A, since that's actually an internationally defined standard. Now, coming from a questionable source, off-shore, maybe they aren't quite Class A.

  At a minimum, as long as you are able to perform a calibration, even a Class B sensor will perform well.

  The thing to keep in mind is accuracy and interchangability are not the same. Plug in a different RTD and you'll get a different reading. Also realize that the performance and accuracy of the overall system (sensor and meter) is only as good as the calibration.

  I've been meaning to pick up a few of those Korean units to give them a try, I agree with you that the price is very compelling, especially given they have the connection heads.

  I would highly recommend you do not use thermowells with RTDs, reaction time will be significantly impacted. This is especially so if you are using a very "meaty" RTD sensor (1/4"). A 1/4" sensor in a thermowell might have a 10 second or greater response time. For condenser temperature control, I'd suggest 1/8th max, no thermowell. In the application a heavy duty sensor isn't necessary.

  Would also recommend that at a minimum, you are using a 3-wire RTD setup. This approach eliminates any error being introduced by the wiring.
• grim

  27 March 2015 02:01PM edited March 2015

  Since we generally work between 0-100c, the difference in accuracy between the classes is small, and gets even smaller when you reduce the temperature under 100c. You'll probably also see that paying the premium for a 1/3rd DIN sensor probably isn't going to net you any real benefit. We're still talking about better than 1C.

  

  

  

  rtd classes.jpg

  583 x 626 - 34K
• DocPorter

  27 March 2015 02:50PM

  I think I may pull the trigger on these. I just need to find out from my equipment supplier how deep my thermowells are.

  I've definitely had the same concern about the response time due to using a thermowell above the dephlegmator. My thought is that since I'm using a proportional valve the slow response will "smooth" out the curve and I'll be able to account for it in tuning the PID hopefully.

  If that doesn't work I'm drilling into the still (although I'm trying to avoid this).
• grim

  27 March 2015 02:59PM edited March 2015

  Yeah it'll work just fine, if it's a constraint, just work around it. It's not like these applications have radically swinging/variable temperatures that need to be compensated for. They generally run steady state until the operator makes a change.
• CothermanDistilling

  28 March 2015 01:14PM

  FWIW - I have a 5mm OD thermocouple in a 1/4" od compression tube thermowell in my dephlegmator, and it woks fine with a proportional valve.

  I have on my list to make a ferrule to eliminate the thermowell and test, but low on the priority list at the moment