expect to see them or as we would like to see them and when the memory is
vague as to the exact figure, the subconscious is in a good position to
juggle it a little bit.
(2) Environmental errors.
(a) Unless they are carefully controlled or allowed for, the
environmental conditions can introduce extraneous variables into the
measurement process. This leads to errors that are difficult to recognize
and to evaluate.
The reading obtained is of the wrong value, or it is a
reasonably true measure of something other than the item being measured.
(b) At the time a measurement is made, all environmental factors
that could have a bearing on the accuracy of the reading should be recorded.
The nature of the ground light intensity would have little effect on a
voltage measurement, although it could be a factor during the calibration of
a photomultiplier tube.
(c) When measurements are made to five or six decimal places,
several factors must be taken into account: temperature, humidity,
barometric
pressure,
local
gravitational
constant,
vibration,
and
contaminants. Every laboratory should be able to determine and/or control
all of these factors.
Temperature control requires heating and
refrigeration systems, suitable blowers, and air ducts.
Humidity control
involves a process to remove water from the air.
Barometric pressure is
difficult to control but it can be easily measured.
The gravitational
constant depends on altitude and latitude (and to a lesser extent), local
geodesic anomalies, see paragraph 4.C.(2)(m)); this can be computed to the
necessary degree.
Dust can be controlled by the use of filters, dust
precipitators, special clothing, and shoe cleaners. The laboratory should
be located away from vibration causes such as punch presses, railroads, and
main highways. The floor should be rigid and massive (thick). Special work
areas requiring additional isolation can be established with pneumatic shock
mounting.
(d) The one item that requires the most control and compensation
is temperature. It is next to impossible to maintain a constant temperature
level with personnel entering and leaving a laboratory.
Each human body
radiates about 100 watts of heat.
If, for each person working in a
laboratory, there is a 100-watt bulb turned on when he is absent and off
when he is present, it would compensate for changes in the heat due to his
coming and going.
(e) High
accuracy
standards
and
measuring
instruments
are
calibrated to be within their rated tolerances at a specified temperature.
If the temperature is some other value, the readings will be different. The
amount of this difference is obtained from the temperature correction chart
that accompanies the standard or instrument.
140
Previous Page
