THE PROSPECTIVE GREAT INCREASE 
in the use of nuclear reactors and radia- 
tion sources will increase the need for 
rapid and highly sensitive methods of 
monitoring for radioactivity. As more 
and more people are potentially ex- 
posed to contamination, genetic con- 
siderations indicate lower levels for 
the maximum permissible exposure. 
There appears to be a growing need 
of an instrument capable of detecting 
radioactivity in human beings and 
in foodstuffs at levels well below the 
recommended maximum _ permissible 
concentrations (1). 
The Human Counter was designed to 
meet these requirements for gamma- 
emitting nuclides. The sensitivity of 
the detector is 5 X 107! c for a nuclide 
emitting one gamma ray per disintegra- 
tion and its sample capacity is 300 lbs. 
For large samples, concentrations of 
gamma emitters as low as 3 X 10°” 
c/kg can be detected. The sensitivity 
is such that the natural K*° content of 
the body (about 1078 gamma-ray 
curies) can be measured to a precision 
of 5 percent in two minutes of counting 
time. 
Description 
The Los Alamos Human Counter is 
essentially an enlarged and improved 
version of the 47 gamma-ray detector 
(“dog counter”) previously descriked 
(2), and of the original human counter 
which was a temporary modification of 
a neutrino detector (3). It consists of 
a cylindrical tank 72 in. in length by 
30 in. in diameter, with an 18-in. 
diameter thin-walled tube (0.019-in. 
stainless steel) running down the axis 
(Fig. 1). The annular space between 
the tube and tank wall is filled with the 
scintillation solution 4 gm/1 terphenyl 
and 0.1 gm/l 1,4di[2-(5-phenylox- 
azolyl)]-benzene (POPOP) dissolved 
in reagent-grade toluene. The capac- 
ity of the detector is 140 gal (about 
1,000 lb). The inside of the tank is 
painted with a special reflective coating 
of anatase TiO, in an epoxy resin that 
withstands the attack of the solution.* 
Scintillations are detected by 108 
photomultiplier tubes (DuMont 6292) 
inserted through the outer wall of the 
tank in 12 rows of 9 tubes each. The 
central cavity accommodates the sub- 
ject or the sample within the sensitive 
shell so that the effective counting 
* Glidden’s Zopaque SD in Shell Chemi- 
eals Epon Formulation XA-200. 
angle is nearly 47 and the counting 
efficiency can be made nearly inde- 
pendent of source position. The 
detector is enclosed in a 5-in.-thick 
20-ton lead shield. Figure 1 shows 
the detector rolled out of the shield for 
servicing. It is mounted on heavy 
rails and ball-bearing rollers, so that 
one man can easily move it. Figure 2 
shows how the subject is put into the 
detector. A metal trough is placed on 
the track on which the tank rolls so 
that it forms an extension of the central 
cavity. The subject lies down in a 
canvas sling which is drawn from the 
trough into the detector by means of a 
windlass. 
Tubes are mounted as in the K-9 
dog counter (2). Each tube views the 
solution through a glass window so that 
tubes can be changed without draining 
the tank. The window is cemented 
into the end of a steel tube with epon 
resin. An O-ring seal provides a tight 
joint between this tube and another 
welded into the tank wall. 
Figure 3 is a schematic diagram of 
the method of wiring the photomulti- 
pliers. The bleeder circuits, one of 
which is shown, are conventional using 
470-k resistors between all stages. 
The last three dynodes are by-passed 
by 5,000-yuf capacitors for stability at 
high rates. The shield is kept halfway 
between the cathode and first-dynode 
potentials. The tubes are balanced by 
means of an additional resistor R in 
series with the bleeder string. This re- 
sistor adjusts the voltage drop per 
stage so that the output pulses of all 
tubes are equal for events of the same 
energy. An unequal voltage drop 
results between the anodes and tenth 
dynode, but the tube characteristic is 
independent of this voltage over a very 
wide range (4). 
To eliminate paralleled anode load 
resistors and output capacitors, the 
anodes are tied together at the high- 
voltage-supply potential. (A positive 
supply is used to prevent spurious 
pulses which arise through the glass 
envelope when the cathode is operated 
at a high negative potential.) Provi- 
sion is made for dividing the tubes into 
two banks, a convenience in testing and 
a necessity when counting very soft 
gammas. Coincidences can then be 
observed between the two banks to 
eliminate tube noise. Tube noise and 
spurious after-pulses are negligible for 
events with energies above 1 Mev, so 
that single-channel operation is used 
Background Determinations 
Deviation 
Time of from 
day Background (cps) average 
L210 447.7+ 2.0 +2.0¢0 
12:50 443.3 —0.2 
438.8 —2.5 
13:20 441.4 —1.2 
445.2 +0.7 
444.3 +0.2 
15:20 439.5 —2.1 
18:30 445.2 “On? 
446.2 +1.2 
445.8 +1.0 
444.9 +0.6 
450.6 (+3.4) 
20:30 449.1 (+2.7) 
21:00 452.3 (+4.2) 
Average: 443.8 (excluding last three 
values) 
in this case. The tubes in a given row 
are alternated between the two chan- 
nels. Each channel has a single 10- 
Meg load resistor for its entire 54 tubes 
and one output capacitor of 500-yuf 
capacity. 
The electronic circuits used to 
amplify, analyze and record the pulses 
are conventional (5). The energy 
resolution of the detector is sufficient to 
permit considerable background reduc- 
tion by proper pulse-height selection, 
so that independently variable upper 
and lower discriminators are used in 
each channel. In addition to the 
scaler used for actual measurements, a 
precision rate meter driving a strip- 
chart recorder is in operation at all 
times. While the rate meter is not 
accurate enough for a precise activity 
determination (which requires a preci- 
sion of 2 eps in 400), it can be used to 
detect nonstatistical fluctuations in the 
average rate due to electrical inter- 
ference, slow drifts in the electronics, 
etc. In practice, the apparatus has 
been shown to be stable to 0.5% of the 
background rate over periods of several 
hours. Because of the high counting 
rates (background 400 cps, average per- 
son 50 cps of K*°), statistical accuracy 
is obtained in very short times. A 
100-see count on the average subject 
will determine his activity to a sta- 
tistical precision of better than 5%, so 
that long-term stability is not an 
essential requirement. 
The table on this page shows the 
