Page 9 of 33
7.1
Sample Pipe
The SADPminiEx is supplied with 2 connections: a “fir tree” on one side and a
Swagelok
®
compression fitting for
1
/
4
in,
1
/
8
in, or 6mm OD pipe (as specified on your
order) on the other. Either can be used as the inlet connection (6).
PTFE and stainless steel are recommended materials for sample pipes. Some specialist
flexible tubing may be suitable as sample pipe, but common flexible tubing, such as pvc,
nylon, polythene or rubber, should not be used, as they are permeable, and water vapour
from the atmosphere will diffuse through them into the sample gas.
A 2m PTFE sample pipe is supplied for use with the “fir tree” connection. The first time
this is used it is advisable to warm the end that will go over the “fir tree” for a few
seconds with a heat source to soften it slightly, which helps to get a gas tight seal.
7.2
Sample conditioning
The main principle when considering the presentation of the gas sample to a portable
dewpoint meter is to keep the sample system as simple as possible and of minimum swept
volume, so that it has the least possible influence on the gas sample.
7.3
Pressure and flow control
In most cases the gas to be sampled will be at a positive pressure, which will need to be
reduced to atmospheric pressure. At low pressures, up to around 10 barg, this can be
done by including a simple needle valve (5) to control the flow and thereby, the pressure
dropping across it. Ideally this would be the isolation valve (1) on the sample point, thus
keeping the number of joints and glands, which are points of potential moisture ingress,
to a minimum.
At higher pressures it is advisable to reduce the pressure across a pressure regulator (4).
Any regulator used should be stainless steel and of a low swept volume. Typical
laboratory or welding type regulators are not suitable because they have elastomeric
diaphragms, which are permeable to water vapour, and the high swept volume will be
slow to purge.
7.4
Particulate contamination
If the sample gas is likely to contain particulate material this should be kept away from the
sensor by including a small in-line particle filter (3) in the sample system.
If the surface of the sensor becomes partly coated with particles it will not affect the
calibration of the sensor, but will restrict the movement of water molecules across the
surface, so increasing the time it takes for the sensor to come to equilibrium with the
