is a third point between 2 going to define the curve ? (sub-zero does not help RTDs, thermistors may be better ?)
sure, but with not so much confidence as if there were 6 points
the cal is done with a reference system having presumably a 10 fold greater resolution/accuracy (= somewhat expensive)
some typical RTD inst and probe cal data (to less than the optimum resolution !!)
true . . . .inst . . . . probe
temp . . .corr . . . . corr
20.00 . . 0.00 . . . . -0.10
25.00 . . 0.00 . . . . -0.12
30.00 . . 0.00 . . . . -0.14
35.00 . . -0.01 . . . . -0.16
40.00 . . -0.01 . . . . -0.17
45.00 . . -0.02 . . . . -0.18
50.00 . . -0.03 . . . . -0.20
55.00 . . -0.03 . . . . -0.22
60.00 . . -0.04 . . . . -0.24
as they were done separately, both corrections must be added to arrive at the 'actual' temp (ignoring uncertainty, etc.)
- a better procedure is to cal the inst and probe together to greater precision, now std practice w/me
I am suggesting that the assumption of linearity is such until demonstrated as fact, particularly in view of the resolution being described
Incoherent
I recognize, and admire, your doing all this yourself, but I question the ability of others to replicate your work - not stuff for a tyro
|