Патент USA US2065997

Dec. 29, 1936.
D. v. EDWARDS ET AL
GASEOUS DISCHARGE TUBE CATHODE
2,065,997
-
Filed June 22, 1934
Fig.‘ /.
INVENTORS
ATTORN EYS
Y
52,065,997
Patented Dec. 29, 1936
UNITED STATES
PATENT OFFICE
2,065,997
GASEOUS DISCHARGE TUBE CATHODE
Donald V. Edwards, Montclair, and Earl K. Smith,
East Orange, N. J., assignors to Electrons, Inc.,
of Delaware, a corporation of Delaware
Application June 22, .1934, Serial No. 731,860
11 Claims. (Cl. 250-275)
This invention relates to gaseous discharge
devices, and is particularly applicable to such
devices having thermionic cathodes of the heat
shielded type. It is also useful in connection
5 with tubes employing gaseous pressures of a low
order.
The art has experienced dif?culty in the use of
gaseous discharge devices, especially those em
ploying pressures below one millimeter of mercury
10 and down to about ?ve-thousandths of a milli
meter (.005 mm.) and those in which the shields
and glass walls are subjected to ion bombard
1
ment. The starting voltage and are drop in
such devices have been found to increase with
use and if the voltage is increased to force start
ing the difficulty is aggravated until ?nally only .
a glow discharge occurs. This amounts to a
complete failure to function inasmuch as a glow,
discharge has high resistance and cannot carry
20 load currents.
Such failures occur even though
the electrodes are apparently in good condition
and the generally accepted explanation has been
that the gas has disappeared; hence the difliculty
has been called “hardening”.
We have found that this so-called “hardening”
25
is not caused by gas disappearance but, on the
contrary, is due to cathode conditions brought
about by a lack of sufficient free space for ioniza
tion immediately adjacent the cathode emissive
30 surface, or surfaces, whereby the same cannot be
utilized effectively. When there is actual dis
appearance of gas it is a secondary effect.
The object of the invention is to avoid the
di?iculty above-mentioned.
To this end We ar~
35 range the cathode according to the pressure and
kind of ionizable medium in a gaseous discharge
device so that there will be sufiicient free and
unobstructed space adjacent the emissive surface
to utilize its emission e?ectively for the genera
40 tion of ions.
Further and related objects of the invention
are the provision of a cathode which will operate
satisfactorily in low-pressure tubes, and the pro
vision of a high-voltage gaseous recti?er having
45 longer life than has heretofore been practical.
The invention will be described with reference
to the accompanying drawing in which Fig. 1
shows, partly in section, a gaseous discharge tube
employing the invention; and Fig. 2 shows a
50 modi?ed form of cathode.
In Fig. l of the drawing, l is an anode, and a
control grid 2 may or may not be provided for
timing the starting of the discharge. 3 is the
cathode which, in the form shown, consists of a
55 cylindr cal hollow body open at the top and hav
ing its inner surfaces treated to render them
electron emissive. The cathode is indirectly
heated to operating temperature by a. heater 4.
Heating energy is conserved by a heat shield sur
rounding both the heater and cathode, and con
sisting of a number of nested, heat-re?ecting cans
5 having perforated covers 1.
The cathode 3,
heater 4, covers 1, and cans 5 are connected to
gether near the top of the cathode. The above
described electrodes are contained in an enlelope 10
8 which is provided with the usual re-entrant
stems and exterior caps or bases. The anode l
is supported from the upper stem by lead-in wires
sealed therein and connected to a terminal 9 in
the upper base. Connection is made to the grid! 15
through an additional lead-in wire which is con
nected at one end to a terminal I!) and at its
other end to a cross bar ll (shown in section)
this cross bar being secured to two of four grid
support wires I4 which are fastened at their 20
upper ends to a collar (not shown) around the
upper stem. A similar collar (not shown)
around the lower stem, and wires l5 support the
outer can 5 and the other parts secured thereto.
Lead-in wires l6, secured to the outer can, pro 25
vide additional support and also provide a com
mon, cathode-heater-shield connection to a ter—
minal I2 in the lower base. The lower end of
heater 4 is welded to a disc 6 in such manner as
to allow for expansion of the heater, the disc 30
being connected by insulated lead-in wires II to
terminal l3.
The distance (1 represents the free space for
ionization adjacent the emissive surface of cath
ode 3; it is the distance an emitted electron may 35
travel before it strikes the surface of a solid body,
such as the opposite emissive surface. In the
form of cathode shown in the drawing, d is the
inside diameter of the cathode 3 and this distance
is available to electrons emitted from any part of 40
the cylindrical surface. An equal or greater dis
tance is available to electrons emitted from the
bottom surface of the cylinder. Sometimes a
cathode is provided with vanes to increase the
emissive area in a given space, or the cathode 45
may have some other form or arrangement of its
surfaces.
In such cases the distance contemplated by
the present invention is that of the unobstructed
space between oppositely disposed emissive sur 50
faces, or it is the distance from an emissive sur
face, and perpendicular thereto, to another sur
face whether emissive or non-emissive, the latter
including neutralizing and shielding surfaces 55
2
2,005,997
which may or may not be connected to the cath
ode.
The art has endeavored to provide a maxi
.
.
distance will not exceed 0.1 centimeter-for any
load that would cause disintegration of the oath
ode.
-
.
"
mum of emissive surface in a minimum of space
The electron mean-free-path for ionization is
in order to take advantage of the ability of
electrons to travel in curved paths in an ionized
medium. However, we have found that suffi
cient space should be provided for generating the
ions and that there is a minimum value for the
the more important part of the distance d. After
an electron has been accelerated to or above the
velocity necessary for ionization in the p'articu
lar medium, it usually travels a relatively large
distance before it ionizes an atom, notwithstand
10 distance d in a given tube if the aforementioned
difficulties are to be avoided. This minimum
ing that it may collide with a number of atoms l0
within this distance. Hence the mean-free-path
depends upon the kind of ionizable medium and
for ionization should not be confusedqwith the
“mean-free-path of an electron” at the same
velocity. The latter is the distance such an elec
tron travels before it collides with an atom, but, 15
its. pressure under operating conditions, and to
some extent upon the cathode emissivity. It is
.15 preferable for practical reasons, such as conserv
ing the heating energy and making the tube as
small as possible consistent with other factors,
to design the cathode for the minimum value of
d plus a factor of safety, inasmuch as no ad
20 vantage is gained by making d too large.
25
30
35
40
45
55
on the average, only one out of many such col
lisions generates an ion. In a sense the ratio of
the mean-freespath of an electron to the mean
free-path for ionization is a measure of the ef~
?ciency of the emitted electrons in ionizing the
gas. It is this low efficiency which requires suf
According to the invention we make the dis
tance d not less than the sum of the distance ficient unobstructed space immediately in front
an electron must travel to attain ionizing velocity of the emissive surface to give the electrons ample
plus the average distance an electron must travel opportunity to ionize atoms before they waste
after that to produce an ion. For brevity, we their energy by striking a solid body.
Some data as to the mean-free-paths for ioni
shall call the former the “accelerating distance”
zation have been given in publications but values
and the latter the “mean-free-path for ioniza
tion”. Their sum is a distance which allows the for gas pressures greater than .05 mm. are'scarce.
Values obtained by exterpolating such data vary
average electron to travel freely from the cath
considerably from the values which we have de
ode surface to the ?rst ionizing collision.
The accelerating distance varies with load and termined experimentally. This discrepancy is
due to the increasing probability of cumulative
cathode activity and may be estimated mathe
ionization as the pressure increases. Our ex
matically. Its magnitude need not be known ac
curately as it is a small part of the above sum periments with argon at 0.1 mm. pressure would
for low pressures and substantial load currents. indicate a mean-free-path for ionization of about
The accelerating distance may also be de?ned as 1.6 cm. for ionizing electrons of the velocity usu
ally encountered in hot cathode discharges,
the perpendicular distance from an emissive sur
face to the outer boundary of its electronic space * whereas exterpolation of the published data
would put the path at approximately 3.0 cm. for
charge. Inasmuch as- the space charge is al
most‘ completely neutralized at full-load current, argon at the same pressure. However, both val-.
the boundary with such current is very close to ues are several times greater than the mean
the cathode surface. The space charge increases free-path of an electron which, for argon under
and its boundary recedes from the cathode as the same conditions, is about 0.16 cm.
The mean-free-paths above-mentioned should
the current is decreased, from which it will be
not be confused with the kinetic mean-free-path
apparent that the accelerating distance ap
proaches in?nity as the current approaches zero. of the argon atom, which pathis also small rela
However, the currents which are small enough tive to the mean-free-path for ionization at the
same pressure. The kinetic mean-free-path is
to make the accelerating distance large can usu
ally be supported safely by a small portion of the average distance an atom moves due to ther
the cathode surface, such as the edge portion mal agitation before it collides with another
nearest the anode. In determining the acceler ' atom.
The foregoing explanation of the distance d
ating distance for the type of tube shown in the
drawing, it is sufficient to take the current at treats its two parts separately and is based on
about 20% of full-load current. In some tubes present electron theory. The sum or total value
a dark sheath is visible adjacent the cathode for d may be determined empirically by either of
surface at low loads. If the emissive surface is the following methods.
The ?rst method consists in constructing a
visible through the tube envelope and if such a
sheath appears, then all loads less than that series of similar tubes having the same perpen
at which the sheath becomes perceptible may be dicular distance d but with different mean-free
60 disregarded in determining the accelerating dis
tance for that tube.
As an example of the length of accelerating
distance which may be expected with a cathode
of ordinary activity, let us assume a tube ?lled
65 with argon at 0.1 millimeter pressure. The in
stantaneous peak or crest current in such a tube
at full load would give an accelerating distance
of about 0.01 centimeter, whereas at 10% load
the accelerating distance would increase to 0.1
70 centimeter. It will be understood of course that,
during each cycle of full-load current, the ac
celerating distance will vary from the above min
imum of 0.01 centimeter to a maximum value
corresponding to minimum current for the cycle.
75 However, in the example given the accelerating
20'
25
30
35
40
45
paths obtained by ?lling the respective tubes
with the same gas at different pressures. These
tubes are put on life test and the pressure found
below which some of the tubes develop the effect
heretofore referred to as “hardening”. The said
distance d is then the minimum distance for that 65
gas and pressure and it may be used for future
designs, employing the same gas, pressure and
cathode emissivity.
>
The second method is preferably employed
after some experience has been gained with the 70
‘above life tests. It consists in operating a given
tube on the pump at various gas pressures and
plotting a curve of gas pressure versus arc drop.
The are drop should be fairly constant with de
creasing pressure down to a point at which it be 75
2,065,997
gins to increase rapidly. At this point the pres
3
sure is such that the distance d in the tube under
heat shield, the effective cathode surface for a
large range of load will be the same as if the
test equals the mean-free-path for ionization
emissive surface were a continuous‘ cylindrical
plus the accelerating distance.
surface passing through the axr s of the spirals.
Such a cathode is illustrated in Fig. 2 which
It is sometimes difficult to obtain de?nite and
accurate results by the above methods unless the
proper load current is employed, which in gen
eral should be relatively light, and some expe
rience may be required in selecting the proper
10 current. It should be large enough to shorten
the life of the tube by “hardening” but not so
large that an abundance of ions is generated by
the small percentage of electrons which succeed
in having ionizing collisions with atoms in less
15 than the average distance. For this reason a
tube which is operated at full-load current, or
one in which the current wave has‘ normal av
erage value but a high peak value, may develop
“hardening” less readily than if it were operated
20 most of the time at about 30 to 50% of full-load
current. The relative e?'ects of load currents
vary in different tube constructions, and depend
'to a- great extent upon the amount of shielding
and whether the distance d is only slightly below
shows a horizontal section of a cylindrical heat
shield 20 containing a number of spiral emissive
?laments 2| disposed in a circle of diameter d
with their axes parallel to each other and to the
cylindrical surface of the shield. The ?laments 10
may have their upper ends connected to the
shield and their lower ends connected to an in
sulated disc, such as 6 in Fig. 1. According to
the invention the distance d which, as stated, is
the diameter of the eiTective cathode surface, 15
should be determined as described above. Under
normal loads no ionization goes on between ad—
jacent spirals, and there may be other arrange
ments whereby part of the emissive surface is in
effective, hence the effective cathode surface will
be smaller in such cases than the total emissive
surface. However, the effective surface should
be sui?cient for the operating requirements of
the tube and should have su?icient free space for
25
ionization adjacent thereto.
be. In the latter case “hardening” may develop
Where a directly heated ?lament has a mate
at any load greater than that which can be sup
rial voltage drop along its length, it is possible
ported by the small percentage of cathode sur
to reduce the free distance somewhat provided
face which is near the top and therefore substan
the ?lament voltage is phased, relative to the
30 tially open to the tube atmosphere.
anode voltage, so that the more distant end of 30v
We believe that “hardening” is due funda
the ?lament is negative, relative to the end
mentally to a shortage of ions for the needs of nearer the anode, on the half-cycles when the
the arc discharge and that the discharge current anode is positive.
is uniformly distributed over the cathode surface
The invention is particularly advantageous in
so long as the free space meets the above-de
making high-voltage tubes, for it is known that 35
scribed minimum. If the distance d is less than the permissible anode voltage for most recti?ers
this minimum for the particular gas and pressure increases as the pressure of the gas ?lling is
used, there will be a shortage of ions under some - reduced. The permissible anode voltage remains
load conditions. This'causes the current to con
fairly constant with decreasing pressure down to
40 centrate on the top part of the cathode and dis
a critical pressure below which it increases
integrate it. The current then concentrates, if rapidly. This critical pressure can be raised by
possible, on another part of the cathode with decreasing the physical size of the tube. How
the same result, until the tube fails by, loss of ever, former low-pressure tubes intended to take
emission instead of by loss of gas as has gen
advantage of these relations have had very short
45 erally been supposed. With this type of failure
life. We have found that this is due to a lack 45
the cathode is not physically destroyed as in the of a suf?cient space for ionization which, ac
case of failure due to long life with load or due cording to our invention, may be made large
to overloads. On the contrary, a cathode which enough so that the pressure at which the arc
has failed due to “hardening” has a low emission drop suddenly increases is below the critical pres
which, if measured at a plate voltage below the sure at which the permissible anode voltage sud 50
ionization potential of the gas ?lling, will prob
denly increases.
ably have the same value as that of a new tube,
The following is a speci?c example of the re
but the saturation emission of such a cathode is
sults thereby obtainable. A tube having only
25 the minimum or very much smaller than it should
so low that it cannot carry load currents.
55
A‘peculiar effect of making the distance 11 less
than the minimum as above described, in a tube
having a heat-shielded cathode, is a pronounced
tendency to generate parasitic oscillations in arc
drop voltage at a frequency of approximately 10
60 kilocycles per second. This is probably due to
the inability of the cathode to generate su?icient
ions inside the shield whereupon ions are gen
erated outside thereof and, due to the high mo
bility of ions in this region, they suddenly diffuse
65 into the space within the shield thereby momen
tarily lowering the arc drop. When the electri
cal charges on these ions are dissipated there is
another shortage and the process repeats itself
at the frequency above-mentioned. The pro
70 vision of a proper free space for ionization within
the cathode shield avoids this di?iculty.
In the event directly-heated emissive ?laments
are used, such as where a number of spiral ?la
ments are arranged with their axes parallel to
75 one another about a circle inside a cylindrical
1.2 cm. of free space may be ?lled with not less
than 0.4 mm. of argon or roughly 0.2 mm. of 55
xenon, and have a permissible plate voltage of
110 volts per anode. By merely increasing the
free distance for ionization to 2.8 cm. the same
tube may be ?lled with 0.05 mm. of argon or
0.025 mm. of xenon and have the same current 60
rating but seven times greater plate voltage rat
ing. Thus with slightly more than two-fold in
crease in the cathode free space and without in
creasing the size of the tube, a seven-fold in
65
crease in permissible tube output is obtained.
From the foregoing it will be apparent that
the invention is particularly applicable to tubes
containing metal vapor, such as mercury vapor
tubes.
In such tubes the vapor pressure is not
constant but varies over a wide range due to
variations in load and ambient temperature.
The cathode in such a tube should be designed
to have a free distance for ionization, as above
described, for the lowest ambient temperature
70
2,066,997
I
and the smallest load for which the tube is’
designed. If the distance is large enough under
these conditions it will be ample for higher tem
cathode, and another electrode, said cathode
peratures and greater loads.
-
their entire area, a distance at least as great as
The~invention may be employed with other
the sum of the electron accelerating distance for
substantial load currents in said medium plus the
rare gases and vapors than those speci?cally
mentioned and also with common gases if they
are inert to the envelope and the electrodes.
We
claim:
.
.
10 I 1. A discharge device comprising ‘an envelope
containing an ionizable medium, a hollow ther
mionic cathode having a ?xed electron emissive
surface therein, and another electrode outside
the cathode, said cathode having an unob
15 structed space adjacent its effective emissive sur
having oppositely disposed ?xed,'electron emis
sive surfaces spaced apart, over substantially
electron mean-free-path for ionization therein,
the space between said emissive surfaces having
free communication with the remaining space in
said envelope.
'
10
'7. A gaseous discharge tube comprising'an en
velope, an ionizable medium at a pressure of about
1 millimeter of mercury or less and a pair of co
operating electrodes in said envelope, one elec
trode being a thermionic cathode having a ?xed 15
electron emissive surface and an unobstructed
face for a distance which allows the average
emitted electron to move from said surface to its
space, immediately adjacent its emissive surface,
?rst ionizing collision in said medium ‘without
striking a solid body.
2. A discharge device including a gas ?lling,
an anode, shields, and a thermionic cathode hav
ing a ?xed electron emissive surface, said cath
erating distance plus the mean-free-path for ion
ization in said medium, said unobstructed space 20
having free communication with the remaining
ode having an unobstructed space in front of
its effective emissive surface for a distance which
bears the same ratio to 1.6 centimeters as the
mean-free-path for ionization in said ?lling
bears to the mean-free-path for ionization in
argon at 0.1 millimeter of mercury pressure.
3. A gaseous discharge tube having an anode,
a thermionic cathode having a ?xed electron
emissive surface, and a gas or vapor ?lling at a
lower pressure for the particular ?lling than the
pressure which gives a mean-free-path for ion
ization in said ?lling corresponding to the mean
free-path in argon at 0.2 millimeter of mercury
pressure, and a heat-shield cooperating with
said cathode, the cathode having free perpen
dicular distances from substantially all of its
effective emissive surface to other surfaces of
said cathode and shield, said distances being
greater than the average distance an electron
must travel from the cathode to its point of
"ionization for all loads down to that at which
the cathode has a perceptible dark sheath.
4. In a discharge device containing an ionizable
medium at low pressure, a hollow cathode struc
ture having a ?xed electron emissive surface
and unobstructed space therewithin, measured
.perpendicularly to substantially all of the effec
tive emissive surface, ‘sufficient to allow the
average emitted electron free travel to its ?rst
‘ionizing collision in said medium at any load
great enough to affect the cathode life, said
cathode structure having a discharge opening
which maintains the ionizable medium inside the
hollow cathode at the said low pressure.
5. The combination in a gaseous discharge tube
of an anode, a thermionic cathode having a ?xed
.electron emissive surface, a ?lling of gas or vapor
at low pressure, and shields, said cathode having
. a space in front of substantially all of its effec
tive emissive surface for a distance greater than
the electron mean-free-path for ionization in the
said ?lling, said space being unobstructed by the
cathode, shields or anode, and having unobstruct
ed communication with the anode whereby the
pressure of said ?lling is substantially the same
throughout the discharge path from cathode to
anode.
70
6. A discharge device comprising an envelope
containing an ionizable medium, a thermionic
a.
eh
at least as great as the sum of the electron accel
space in said envelope.
_
8. A gaseous discharge tube comprising an en
velope, an ionizable medium and a pair of coop
erating electrodes in said envelope, one electrode 25
being a hollow thermionic cathode containing
?xed electron emissive surfaces, the distances
between said surfaces being ?xed according to
the kind and pressure of said medium so as to
provide su?lcient free space for the majority of 30
the emitted electrons to ionize said medium with
in the cathode, said cathode being open to said
medium whereby the pressure within the cathode
is not increased relative to the pressure outside
35
thereof.
9. A discharge device comprising an envelope,
an ionizable medium therein, and electrodes in
cluding a thermionic cathode having a ?xed elec
tron emissive surface, a heat-shield therefor and
another electrode, said electrodes and shield be 40
ing so disposed within said envelope relative to
the effective emissive surface of the cathode that
there is su?icient distance to permit the average
electron to travel perpendicularly from said sur
face to its ?rst ionizing collision with an atom 45
of said medium before striking one of the said
electrodes, the space between the cathode and
said other electrode being unobstructed.
10. A discharge tube comprising an envelope
containing argon at approximately .05 millimeter 50
of mercury pressure, an anode,‘ a thermionic cath
ode having ?xed electron emissive surfaces and
a free distance for ionization of about 2.8 centi
meters between said emissive surfaces, and a heat
shield surrounding said cathode, said tube hav 55
ing a permissible anode voltage of at least 750
volts.
~
11. A discharge tube comprising an envelope
containing a rare gas, an anode, a thermionic
cathode having ?xed electron emissivesurfaces, 60
and a heat-shield surrounding said cathode, the
said emissive surfaces being spaced apart a dis
tance greater than the electron mean-free-path
for ionization in said gas, the pressure of said gas
being above the pressure at which the arc drop 65
in said tube increases rapidly and below the crit
ical pressure at which the permissible anode volt
age suddenly increases.
DONALD V. EDWARDS.
EARL K. SMITH.
to