Models 947, 948 and 960 Leak Detectors Operations and Service Manual
G-7
A sampling probe is connected to the leak detector. The object to be tested is filled with
helium at the desired test pressure and the probe is moved over its surface. Some of the
helium escaping from a leak is captured through the probe and enters the leak detector, thus
locating the leak.
Sensitivity of this type of testing is limited to about 10
−
7
atm cc/sec, since most of the
escaping helium diffuses into the surrounding atmosphere. The sensitivity is also limited
by operator technique and variation in ambient helium concentration in the vicinity of the
testing.
An alternative to probing is to enclose the object and probe the enclosure for a change in
helium content.
G.6.3 Test Piece Already Sealed
Figure G-4 is an illustration of a bell jar used to determine overall leak rate of a sealed test
piece.
Figure G-4
Test Piece Sealed with Helium and/or Other Gases –
Bell Jar Used to DetermIne Overall Leak Rate
Sometimes it is necessary to leak check a completely sealed object. This can be done by
placing helium inside the object (either 100% or mixed with other gas used for backfilling)
before sealing it. The object is then placed in a vacuum chamber connected to the leak
detector. Helium escaping from the object into the vacuum chamber is detected by the
spectrometer tube. Sensitivity depends on the partial pressure of helium in the object.
If the presence of helium in the finished object is undesirable, units already sealed can first
be placed in a container that is then pressurized with helium for a specific time at a known
pressure. Helium enters the object through leaks that are later detected, as described in the
previous paragraph. Gross leaks are sometimes not detected, since all the helium entering
through a large leak can be lost prior to testing. Also, spurious signals can be given by
helium not entering the object, but entering surface fissures and remaining long enough to
be detected.
Summary of Contents for 948
Page 175: ......
