Product guide

93
Humidity and Dew-Point Instruments
Please note: Michell Instruments adopts a continuous development program which sometimes
necessitates specifi cation changes without notice. Please contact us for latest version.
www.michell.com
The deviation from the ideal calibration line will have a
maximum or minimum, usually expressed in the form:
‘Non-linearity ±0.5% RH’, at a given temperature. It
should also be noted that linearity characteristics can
be signifi cantly affected by temperature. When having
an instrument calibrated, linearity is a factor to bear
in mind when deciding how many measurements are
needed and at what intervals throughout the range of
measurement.
Hysteresis
Hysteresis, in general terms, is the dependence of a
reading upon whether the condition is approached from
above or below the value of interest. For example, given
a humidity cycle of say 10% RH to 50% RH to 90% RH
and back to 50% RH, most hygrometers would not give
an identical reading on both occasions at 50% RH.
Hysteresis is related to repeatability, but includes any
‘directional’ effect.
Hysteresis should be considered if the measurement is
to be used for control purposes, for example to activate
on-off control of air conditioning.
Response time
Response times are included in specifi cations to provide
an indication of how long the instrument takes to react
to changes in the applied condition. While the sensor
itself may have a particular response time, any screening
of the sensing element, e.g. by a protective fi lter, will
slow down this response. Effective air movement will
speed it up. (Constant and specifi ed air fl ow is required
for a response time test.) Response times are practically
always slower for falling humidity than for rising
humidity.
Quantitatively, response times are usually quoted in
terms of the time taken to register 63% of a step
change in the applied condition (although other
conventions are sometimes used, e.g. 90%).
Response times are usually quoted for the hygrometer
alone, at constant temperature. However, response
times of the associated sampling systems may be much
greater. Optimistic specifi cations of response time
may raise false expectations for the user to see stable
measurements after very short time periods. However,
for most relative humidity measurements, the time taken
for the sensor and nearby materials to equilibrate with
respect to temperature is by far the most signifi cant
factor infl uencing the response time. For measurements
of dry gases, equilibration of moisture in the sampling
system is usually the key infl uence.
Long-term stability
The measurement characteristics of any instrument will
change with respect to time, due to gradual changes
in electrical or material components. Estimates of long-
term stability or drift, refer to the likely change in the
instrument’s measurement performance with respect
to time. Regular checks of calibration should be made
to quantify this potential problem. Although it may
sometimes be desirable to adjust the hygrometer reading
in the light of drift, performance could be compromised
by the adjustment process, and the overall drift
characteristics could be masked by frequent adjustments.
Quantitatively, drift may be expressed in terms of a
time span and a fi gure. However, if an instrument was
subject to drift of less than 4% RH per year, it would
not follow from this that the drift over six months would
be less than 2% RH (though it could be expected to be
somewhat less than the annual fi gure). Drift is not always
consistent, and measurements of drift always include
some contribution from short-term variability.
Temperature coeffi cient
Temperature variation has a most signifi cant infl uence
on relative humidity itself. In addition, every humidity
sensor has a temperature coeffi cient, which can be simply
explained as a change in measurement characteristic at a
different temperature. A temperature coeffi cient might be
expressed, for example, as 0.1% RH per °C, which could
result in an additional 5% error when measuring at a
temperature 50°C away from the calibrated temperature.
5 What is calibration?
Calibration is the process of comparing a measuring
instrument against an authoritative reference for the
same type of measurement, to identify any bias or
systematic error in the readings. The outcome of a
calibration is normally a certifi cate listing any corrections
that need to be applied to the values indicated by the
instrument, together with an estimate of the uncertainty
in the calibration, and other relevant information. For
example, a calibration of a given instrument at, say, 50%
RH, might show it to read too high by 1% RH. If so, a
required correction of –1% RH would be shown on the
certifi cate.
Calibration is often taken to mean ‘adjustment of the
instrument to read correctly’. This is not true. Calibration
and adjustment of an instrument are quite separate
concepts, and the two should not be confused. Thus,
when arranging for the calibration of any instrument,
it is important to establish clearly whether or not the
instrument is to be adjusted as well as calibrated. If so,
it should be specifi ed whether calibration information is
required only after adjustment, or whether results are
also required for the initial or ‘as found’ condition.
A Guide to the Measurement of Humidity