Lake Michigan environmental survey : final report / [edited] by John C. Ayers.

petri dishes on the 5" x 5" NaI crystal. A point source of Cs-137 containing
approximately 3 microcuries was placed at the top of the tower and counted for
100 seconds. The percent dead time and location of the photopeak were recorded. The tower was then shortened by one or more of the petri dishes and
another 100 second count taken. In this way the dead time of the spectrometer
was increased slowly from 1.2% to 31.1% with a noticeable induced gain shift.
The second experiment determined the effect of sample activity on counting efficiency. Two towers of petri dishes were built on the 5" x 5" Nat crystal. Two point sources of Cs-137 were used, each with an activity of about 3
microcuries. The first source was placed on top of one tower and counted for
100 seconds. Then the first source was removed, and the second source was
placed on top of the second tower and counted for 100 seconds. The first
source was then replaced on the first tower, and both sources were counted together for 100 seconds. The counts under the photopeak were recorded for the
three counts. If the live timer were accurate the counts detected with both
sources in place should equal the sum of the counts detected with each source
counted individually. The dead time resulting from having both sources in
place was also recorded. The towers were then shortened by one or more petri
dishes, and another set of measurements was taken.
Figure A-5 shows the data so obtained, and indicates that as dead times
exceed about 5% the fraction of counts lost by overloading the live timer increases linearly. Therefore, it would be desirable to keep the activity of the
calibration planchets below 5% to insure that no counts will be lost due to
the live timer. Figure A-5 also indicates that the gain is not constant with
increasing dead time and that the change in gain with dead time is most significant at dead times around 2%.
From the information displayed in Figure A-5 it was decided that calibration Dlanchets could be prepared that had activities which would result in dead
times between 4% and 5%. Within this range the spectrometer would not be losing any counts due to overloading the live timer, and the gain shift induced
by a sample with 4% dead time would be within 0.1% of the gain shift induced
by a sample with a 5% dead time.
Spectrum Analysis Computer Program
A spectrum analysis computer program which uses the least squares method
has been developed by Dr. J. I. Trombka of the Goddard Space Flight Center.

A-33