Teledyne 8800A, Operating Instructions Manual

Page: 53 / 72

8800 Series Trace Moisture Analyzer Instruction Manual Appendices
46
Appendix G: Dewpoint Response time Analysis
Manufacturers of dewpoint sensors will make unbelievable claims of speed. But without support-
ing data revealing the conditions under which the measurements were made, such as flow rate,
temperature, the dewpoint at which the test was performed, and most importantly the units of
measure. This last aspect can be extremely deceiving, consider the following example.
We have performed a variety of tests to determine the response time of ours as well as the compet-
itor’s sensors. For this discussion let’s consider a step change of -40°C to -60°C dewpoint, in
other words a challenging but typical dry-down. An XTR-100 sensor w/SS sintered filter was
placed in an electro-polished stainless steel manifold, the temperature was 23°C and the sample
flow rate was 3 LPM. The dewpoint was generated with a computer controlled mass flow valve
gas dilution dewpoint generator, and was verified with a NIST and NPL calibrated MBW chilled
mirror. The system was allowed to achieve equilibrium at -40°C dp overnight, then the dewpoint
was abruptly changed to -60°C, the change is depicted below as time 0.
The graph below shows the percent change of recorded dewpoint (°C), the same data is also
shown in percent change of the recorded dewpoint converted to ppmV. It is immediately obvious
how a sensor that has response times exceeding ten minutes can be claimed to have response time
of 3 minutes, simply by manipulating the units of measure. The logarithmic relationship between
dewpoint and ppmV makes the ppmV reading appear faster. Since the response time is stated in
percent change, the units of measure are automatically hidden from the audiance.
30
40
-50 C = 50%
60
70
80
90
-60 C = 100%
110
1
2
3
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
% chan
ge of pp
mV convert
ed measurement
%
cha
nge
of
C
mea
sure
ment
Time (minutes)
Percent Change