|
by
Dr. Theo P. Schaad
Subject:
Long-term Stability of S/N 34260(PA4), Model 760-15A
This
unit was originally calibrated here Feb 1989, tested for
stability in 1990 by Det Norske Met Inst. (KNMI study), and was
later used as demo equipment all over the world (including
Erpatec Switzerland, Micromeritics Georgia, Ammo Eng. Israel,
and Canagrad Surveys Canada).
I tested the unit on 8 July 1994 at nominal 14.5 psia for
zero stability. The
unit was reconfigured to the original calibration coefficients.
Time
elapsed:
5.4 years
Measurement
1:
The
unit was tested with the CEC Dead Weight Tester,
Piston/Cylinder, Weight Set 2.
This set was last calibrated 2 Feb 1993 with a traceable
accuracy of 0.01% of reading, or +/-0.00145 psi at the tested
pressure value.
Applied
pressure:
14.50449 psia
Indicated
pressure:
14.50336 psia
Difference:
-0.00113 psia
Relative
diff:
-0.0076% fs
The
original calibration had been done with the CEC Dead Weight
Tester, Piston/Cylinder 1, Weight Set 1.
The calibration was based on the original CEC values
(1987?) with the masses remeasured in 1988.
The accuracy had been stated as 0.015% of reading, or +/-
0.00218 psi at the tested value. This accuracy is largely
independent of the current value.
Realistically, the two dead weight tester error should be
added in this measurement to +/-0.00363 psi.
The typical repeatability achieved on measuring a
barometer is 20 ppm or +/-0.0003 psi.
Hence the total stability should be expressed as:
Stability
in 5.4 years:
-0.00113 +/-0.0003 +/-0.00363 psi
where
the first error expresses the precision of the measurement, and
the second the known accuracy.
This gives a total range of between –0.0051 and +0.0028
psi.
Relative
slope:
between –0.0063% fs and +0.0034% fs per year.
Measurement
2:
The unit was tested with the CEC Dead Weight Tester,
Piston/Cylinder 1, Weight Set 1.
The old calibration coefficients and weight values (1989)
were used.
Applied
pressure:
14.49936 psia
Indicated
pressure:
14.49705 psia
Difference:
-0.00231 psia
Relative
diff:
-0.0154% fs
The
success of this measurement depends how much the applied
pressure of the dead weight tester has changed over the years.
Since we mostly used another piston set since 1990, I
expect that the piston area of Piston 1 is essentially
unchanged. Again,
as above, a repeatability error of 20 ppm should be considered.
Stability
in 5.4 years:
-0.00231 +/-0.0003 psi
Relative
slope:
between –0.0032% fs and –0.0025% fs per year.
|
