L
leeenglock
good reply and very comprehensive.
also to note there are other practical difficulties at the site, eg. the sensor is in thermowell which is fairly massive and
introduces time constant issues and also conductivity issues and also thermal leakage , ie the ambient where the probe is used is say 40degc in the plantroom while the probe tip is to be in area of say 5degc, then the conduction of the thermowell will add to the inaccuracy of the sensor. i normally use a thermowell with a hole for the sensor to protrude into the fluid
stream, ie not a regular thermowell.
another issue with high accuracy readings is the self heating of prt and thermistors---low currents must be used which then introduce noise issues, cleanliness of the contacts, etc.
for high resolution and accuracy the fast sampling rate is also inadequate since there is settling time for the a/d converter, contact bounce if rel= ays are used, no time for the
averaging/integrating system to cancel out noise, etc. definitely a non trivial problem.
rgds
leelock
also to note there are other practical difficulties at the site, eg. the sensor is in thermowell which is fairly massive and
introduces time constant issues and also conductivity issues and also thermal leakage , ie the ambient where the probe is used is say 40degc in the plantroom while the probe tip is to be in area of say 5degc, then the conduction of the thermowell will add to the inaccuracy of the sensor. i normally use a thermowell with a hole for the sensor to protrude into the fluid
stream, ie not a regular thermowell.
another issue with high accuracy readings is the self heating of prt and thermistors---low currents must be used which then introduce noise issues, cleanliness of the contacts, etc.
for high resolution and accuracy the fast sampling rate is also inadequate since there is settling time for the a/d converter, contact bounce if rel= ays are used, no time for the
averaging/integrating system to cancel out noise, etc. definitely a non trivial problem.
rgds
leelock
