Lead shield 
(ee ee 
VZZZZ2 Lu ig 
[eresa’e Se ees “eee ete ese") 
aS Insulator 
Pix ep 
CZZ ZL EL 
Te Isotope 
a= 
“Aluminum cup 
+ 
“Lead shield 
FIG. 2. First experimental model of 
generator with solid insulator 
improved vacuum techniques and elec- 
trode design, a potential difference of 
365 kv was obtained. In the other 
device, a direct electrical connection to 
the source electrode was made through 
an external insulator that formed part 
of the vacuum enclosure. A some- 
what lower voltage was obtained with 
this device due to surface leakage on 
the external part of the insulator ex- 
posed to the atmosphere. A capacity- 
charging curve across a resistive load 
was measured by Linder for an equiva- 
lent circuit of a beta current source, a 
capacity, and an internal resistance in 
parallel with the external load. The 
internal resistance was a combination of 
surface conduction across the insulator 
bushing and some conduction through 
the vacuum by cold emission or residual 
gas. 
Solid Dielectric 
In 1951, a project was started at this 
laboratory to develop a solid insulator 
that could be substituted for the 
vacuum dielectric in the direct-conver- 
sion generator to increase reliability. 
The literature at that time stated, 
however, that insulators were both 
lowered in electrical resistivity (6) and 
physically degraded (7) by ionizing 
radiation. These effects are not limi- 
tations in the vacuum generators since 
insulator bushings can be shielded from 
the radiation. On the other hand, 
when a solid is substituted for a 
vacuum, the radiation must penetrate 
the solid dielectric itself for operation 
of the device. 
Thus, to determine the feasibility of 
a solid-dielectric type of radioactive 
battery, a test program was started on 
commercially available insulators. 
Experimental Data 
For geometrical simplicity, tests 
were made using flat sheets of insula- 
138 
2 
° 
= 
> 
4,000 
t (sec) 
FIG. 3. Charging voltage vs time for 
0.002-in. polystyrene after irradiation for 
few minutes (A) and several hours (B) 
tion between the faces of cylindrical 
source and collector electrodes, as 
shown in Fig. 2. As a beta source, 25 
me of Sr%-Y% were deposited in an 
electrode formed by an aluminum cup 
covered with aluminum foil. An alu- 
minum disk, opposite this source elec- 
trode, collected the beta particles after 
penetration of the insulator. Lead 
shielding was used around the alumi- 
num collector and aluminum source 
electrode to reduce the bremsstrahlung 
below tolerance at the surface. 
In operation, the beta current col- 
lected through a 0.002-in.-thick poly- 
styrene insulator was measured with an 
electrometer to be about 50 X 107 
ampere at zero voltage. The total 
beta current emitted from 25-mc 
Sr%-Y99 is 300 X 107-1? ampere; how- 
ever, one half of the current is emitted 
in an opposite direction from the col- 
lector and is absorbed in the aluminum 
cup. The remaining loss of current is 
due to absorption in the thin aluminum 
sheet covering the cup and in the 
insulator, itself. 
Charging voltage. The open-cir- 
cuit charging voltage was measured as 
a function of time by an electrometer 
FIG. 5. 
Equivalent circuit of battery using 
solid dielectric that increases in resistance 
with irradiation time 
R=8«10!2 ohms 
R= 4x 10!2 ohms 
50 100 150 
V (volts) 
200 
FIG. 4. Current-voltage characteristics of 
0.002-in. polystyrene sheet under condi- 
tions of Fig. 3 
voltmeter with negligible capacity. 
The resulting voltage, V, (the lower 
curve in Fig. 3) for the 0.002-in.-thick 
polystyrene insulator showed an expo- 
nential rise with time, t, to a maximum 
of 200 volts. 
As in the case of the vacuum gener- 
ators of Moseley and Linder, capacity 
charging of this type is given by the 
equation 
V = RI5(1 — e-/80) (1) 
where F is the total internal and exter- 
nal resistance in parallel, C is the total 
capacity, and J, is the beta current. 
The maximum charging voltage, Vimax, 
for t > RC is 
Vinx = RI pg (2) 
Next, the initial rate of charge is 
found by expansion of Eq. 1 fort < RC 
to be independent of R and given by 
I 
Vo= 7 (3) 
Thus, from Eq. 2, for Vinax of 200 volts 
and Js; of 50 X 107! ampere, R is 
4 X 10!? ohms. 
Irradiation effect on insulator. 
After several hours, a second measure- 
ment of the charging voltage as a func- 
tion of time was made, and V,,, was 
found to have increased to 400 volts, as 
shown by the upper curve of Fig. 3, 
indicating an increase in insulation 
resistance to 8 X 10!” ohms while the 
initial charging rate remained the same. 
Consequently, a direct measurement of 
internal resistance, R;, was made by 
applying an external voltage to the 
generator terminals and recording the 
resulting current with the electrometer. 
The current-voltage characteristics 
for an unused 0.002-in.-thick poly- 
styrene sheet is shown by the lower 
curve of Fig. 4. The straight line 
indicated an ohmic resistance of R; = 
