U.S. patent number 3,555,346 [Application Number 04/702,951] was granted by the patent office on 1971-01-12 for vacuum tubes.
This patent grant is currently assigned to National Research Development Corporation. Invention is credited to James Dwyer McGee, Robin Wyncliffe Smith.
United States Patent |
3,555,346 |
McGee , et al. |
January 12, 1971 |
VACUUM TUBES
Abstract
An improvement is made in a known vacuum tube arrangement in
which information is stored as a modulated electron stream
traveling to and fro through a drift region between reflecting
meshes, information retrieval being accomplished by gating pulses
applied to one mesh to allow electrons to pass through it. Analysis
of the gating mechanism shows that satisfactory high-speed
operation requires a short electron transit time through the
decelerating field associated with the gating mesh. One practical
means of achieving this involves the provision of a high potential
electrode between the drift region and the gating mesh.
Inventors: |
McGee; James Dwyer (London,
EN), Smith; Robin Wyncliffe (London, EN) |
Assignee: |
National Research Development
Corporation (London, EN)
|
Family
ID: |
9816467 |
Appl.
No.: |
04/702,951 |
Filed: |
February 5, 1968 |
Foreign Application Priority Data
|
|
|
|
|
Feb 10, 1967 [GB] |
|
|
6546/67 |
|
Current U.S.
Class: |
315/16; 315/10;
315/12.1 |
Current CPC
Class: |
H01J
31/52 (20130101) |
Current International
Class: |
H01J
31/08 (20060101); H01J 31/52 (20060101); H01j
029/46 () |
Field of
Search: |
;315/14--16,10,12 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Bennett, Jr.; Rodney D.
Assistant Examiner: Hubler; Malcolm F.
Claims
We claim:
1. A vacuum tube arrangement incorporating:
A. a vacuum tube having;
a. means, comprising an electron-emissive cathode, for generating
an electron stream of variable intensity,
b. an output device spaced from the generating means and responsive
to electrons from the stream, and
c. an intermediate electrode system disposed between the generating
means and the output device and including,
1. a tubular electrode disposed so as to be coaxial with a straight
electron transmission path extending between the generating means
and the output device,
2. a first apertured screen disposed across said path between the
generating means and the tubular electrode, and
3. a second apertured screen disposed across said path between the
tubular electrode and the output device;
B. means for providing a focusing magnetic field extending along
said path;
C. means for maintaining the tubular electrode at a positive
potential with respect to the cathode to provide in said path an
elongated drift region along which electrons may travel with
substantially uniform velocity;
D. means for selectively maintaining the first screen at a positive
or negative potential with respect to the cathode;
E. means for biassing the second screen negative with respect to
the cathode potential to establish a decelerating field for
electrons emerging from said drift region in a direction towards
the second screen, the potential at any point in the decelerating
field relative to the cathode potential being V; and
F. means for applying to the second screen positive going gating
pulses of magnitude V.sub.p greater than the magnitude of the
negative bias on the second screen; the arrangement being
characterized by that improvement consisting of so dimensioning
said decelerating field that the value of the parameter ##SPC3## V,
V.sub.p, V.sub.z, m and e being expressed in a self-consistent
system of units.
2. A vacuum tube arrangement according to claim 1, in which said
intermediate electrode system further includes a second, relatively
short, tubular electrode disposed coaxial with the first-mentioned
tubular electrode between the latter and the second screen, the
arrangement including means for maintaining the second tubular
electrode at a much higher positive potential than the first.
3. A vacuum tube arrangement according to claim 2, in which there
is disposed across that end of the second tubular electrode nearer
to the second screen a metal plate maintained at the same potential
as the second tubular electrode and having formed in it a long
narrow slot.
4. A vacuum tube arrangement according to claim 1, in which there
is disposed between the tubular electrode and the second apertured
screen a further apertured screen spaced by a very small distance
from the second apertured screen and maintained at a small negative
potential with respect to the cathode.
5. A vacuum tube arrangement according to claim 1, in which the
spacing between the second apertured screen and the adjacent end of
the tubular electrode does not exceed one centimeter, there being
disposed across this end of the tubular electrode a further
apertured screen maintained at the same potential as the tubular
electrode.
Description
This invention relates to vacuum tube arrangements of the kind
incorporating: a vacuum tube comprising means for generating an
electron stream of variable intensity, an output device spaced from
the generating means and responsive to electrons from the stream, a
tubular electrode disposed between the generating means and the
output device so as to be coaxial with a straight electron
transmission path extending between the generating means and the
output device, a first apertured screen disposed across said path
between the generating means and the tubular electrode, and a
second apertured screen disposed across said path between the
tubular electrode and the output device; means for providing a
focusing magnetic field extending along the transmission path;
means for maintaining the tubular electrode at a positive potential
with respect to the cathode of the tube so as to provide in the
transmission path an elongated drift region along which electrons
may travel with substantially uniform velocity; and means for
individually switching the two screens between potentials which are
respectively negative and positive with respect to the cathode so
as to control the flow of electrons along the transmission
path.
Arrangements of this kind are described, for example, in a paper by
J. D. McGee, J. Beesley and A. D. Berg, published in the Journal of
Scientific Instruments, Volume 43, pages 153-- 159 (Mar. 1966), and
may be used for various purposes involving the storage of
information. Thus, in operation of such an arrangement the electron
stream is modulated in intensity in accordance with information
which is to be stored, and the modulated stream is admitted into
the space between the screens by maintaining the first screen at a
positive potential. The electron stream may then be trapped in the
space between the screens by maintaining both at negative
potentials, the electron stream making repeated transits of the
drift region in opposite directions and being reflected on each
approach to one or other of the screens by virtue of the
decelerating field extending from the relevant screen in the
direction of the drift region. When it is desired to retrieve an
item of information from the tube, an appropriately timed gating
pulse of short duration may be applied to the second screen so as
to drive it positive, a corresponding part of the electron stream
thereby being caused to emerge from the space between the screens
and travel on to the output device.
Detailed analysis of the operation of such an arrangement has now
shown that when a gating pulse is applied the sampling time (that
is the period corresponding to that part of the original electron
stream which is caused to pass through the second screen by the
application of the gating pulse) is generally longer than the
duration of the gating pulse. Moreover a dispersion effect is
encountered due to variation of the length of the sampling time for
different values of the velocity of entry of electrons into the
decelerating field in front of the second screen, which will of
course occur in practice because of the spread of emission energies
of the electrons at the cathode. In order to make high speed
operation feasible it is essential that this dispersion effect
should be reduced as much as possible, and the present invention is
concerned with the provision of means by which this may be
achieved.
Broadly speaking the invention is based on the realization that the
dispersion effect referred to can be reduced if the arrangement is
made such as to obtain a relatively high value for the ratio of the
magnitude of the decelerating field in front of the second screen
to the distance over which this field extends; this corresponds to
a relatively short electron transit time through the deceleration
region.
The invention will be further explained with reference to the
accompanying drawings, in which:
FIG. 1 is a diagrammatic illustration of one vacuum tube
arrangement of the kind specified, used as a framing camera;
FIG. 2 is an explanatory diagram; and
FIG. 3 is a diagrammatic view of part of the electrode structure of
a vacuum tube for use in an arrangement of the kind specified,
modified in accordance with the invention.
The arrangement shown in FIG. 1 incorporates a vacuum tube 1 having
a tubular glass envelope 2 which is closed at its ends by glass
plates 3 and 4 respectively carrying a photocathode 5 and a
phosphor screen 6; in operation the screen 6 is maintained at a
high positive potential with respect to the cathode 5. The tube 1
is disposed coaxially within a solenoid 7 which provides a strong
axial magnetic field such that electrons travelling along the
length of the tube 1 will be constrained to follow helical paths of
such small diameter that they may be regarded virtually as moving
in straight lines parallel to the axis of the tube 1. The flow of
electrons from the cathode 5 to the screen 6 is controllable by
means of two apertured screens 8 and 9, which may suitably be in
the form of wire meshes, the screen 8 being disposed adjacent the
cathode 5 and the screen 9 being disposed at an intermediate
position along the length of the envelope 2. In operation, the
screen 8 is biassed slightly positive with respect to the cathode 5
but may be driven to a negative potential by pulses from a pulse
generator 10, while the screen 9 is biassed slightly negative with
respect to the cathode 5 but may be driven to a positive potential
by pulses from a pulse generator 11, the generators 10 and 11 being
controlled by a timing circuit 12 as explained further below.
Between the screens 8 and 9 is disposed a tubular electrode 13,
constituted by a conductive coating on the wall of the envelope 2,
which in operation is maintained at a positive potential (suitably
about 100 volts) with respect to the cathode 5 to provide an
elongated drift region for electrons. Between the screen 8 and the
electrode 13 is disposed a series of annular electrodes 14 which in
operation are maintained by means of a potentiometer 15 at graded
potentials such as to provide a substantially uniform electrostatic
field between the screen 8 and the nearer end of the electrode
13.
Between the screen 9 and the screen 6 are disposed in succession an
electrostatic deflection system 16 in the form of a pair of
parallel plates, and an accelerating system comprising a series of
annular electrodes 17 which in operation are maintained at suitably
graded potentials by means of a potentiometer 18. A deflection
waveform, of either "ramp" or "staircase" form, is arranged to be
applied to the deflection system 16 from a deflection wave form
generator 19 which is also controlled by the timing circuit 12.
Light from a high-speed event which is to be recorded is arranged
to be imaged on to the photocathode 5 by means of a suitable
optical system (not shown), the resultant electron stream emitted
by the cathode 5 flowing through the screen 8 by virtue of the
initial positive bias on the screen 8. Occurrence of the event is
detected by a detector 20 having a rapid response, the output from
which is arranged to initiate operation of the timing circuit 12.
The operation of the latter is such that after a short delay a
relatively long negative going pulse is applied to the screen 8
from the generator 10, so that the electron stream is effectively
trapped in the space between the screens 8 and 9, making repeated
transits of this space in opposite directions. The timing circuit
12 also actuates the generators 11 and 19, the former being
arranged to apply to the screen 9 a series of short positive going
gating pulses so timed that a different sample from the electron
stream will be allowed to pass through the screen 9 each time the
stream approaches the screen 9. Each such sample will subsequently
be accelerated to the phosphor screen 6 to produce an output image,
the successive images (representing successive phases of the event
to be recorded) appearing at different positions on the screen 6 by
virtue of the deflection waveform applied to the deflection system
16 from the generator 19. The images appearing on the screen 6 are
arranged to be photographed by means of a conventional camera (not
shown). By using an arrangement as shown in FIG. 1, it is possible
to accomplish high-speed photography with framing rates of up to
10.sup.9 frames per second, with the camera synchronized to the
event to be recorded.
As indicated in the general discussion above, the operation of the
arrangement is vitally dependent on the manner in which sampling is
effected by the gating pulses applied to the screen 9, and this
point will now be considered in detail with reference to FIG. 2,
which is an idealized diagram showing graphs of the variations of
electrostatic potential and electron kinetic energy in the
deceleration region between the electrode 13 and the screen 9, it
being assumed for simplicity that there is a uniform electric field
within this region. In the diagram, distance is measured from the
end of the electrode 13 nearer the screen 9, the latter being
disposed at a distance d; zero potential is taken to be that of the
cathode 5 and the electrode 13 is assumed to be maintained
throughout at a potential V.sub.D. The scales of electrostatic
potential and electron kinetic energy are such that one volt on the
former corresponds to one electron volt on the latter. The line a
indicates the decelerating field which exists when the screen 9 is
biassed negative by an amount V.sub.B, while the line b indicates
the decelerating field which exists while a gating pulse is applied
to the screen 9 so as to drive it positive by an amount V.sub.G, it
being assumed that the fields are perfectly linear; in the analysis
which follows it is assumed that all potentials change
instantaneously, that is the gating pulse is assumed to be
perfectly rectangular and any effects due to retarded potentials
are neglected.
In general, electrons will enter the decelerating field from the
drift region with kinetic energies slightly in excess of eV.sub.D
(where e is the electronic charge), due to their emission energies;
we denote the voltage corresponding to the excess energy as V.sub.O
(which will typically have a value in the region of zero to one
volt). As they traverse the decelerating field towards the screen
9, the electrons will lose kinetic energy linearly with increasing
distance at a rate dependent on the instantaneous value of the
decelerating field; an electron will be reflected back into the
drift region if its kinetic energy is reduced to zero before it
reaches the screen 9 but will pass through the latter if it reaches
it with finite kinetic energy, it being assumed that an
accelerating field exists beyond the screen 9.
Consider first electrons which enter the decelerating field shortly
before the application of a gating pulse. Where the conditions are
such that the electron transit time through the deceleration region
is relatively long, the variation of kinetic energy with distance
as the electrons travel towards the screen 9 may be represented for
three cases of electrons entering the decelerating field at
successively later instants by paths such as ABCD, AEFG and in the
diagram, the lines AHEB, CD, FG and KL being parallel to the line a
and the lines BC, EF and HK being parallel to the line b. It will
be appreciated that the lines BC, EF and HK indicate the loss of
kinetic energy for the relevant electrons during the gating pulse
and that they increase in length in the order stated because of the
different values of kinetic energy for the relevant electrons at
the beginning of the gating pulse. The path ABCD corresponds to a
case in which the electron enters the decelerating field
sufficiently early to be reflected while the path AHKL corresponds
to a case in which the electron enters the decelerating field
sufficiently late to pass through the screen 9. The path AEFG
corresponds to the limiting case of the first electron to enter the
decelerating field which will pass through the screen. Where the
electron transit time through the deceleration region is
sufficiently short, however, an alternative possibility arises,
namely that the limiting case is represented by the path AXG, where
the line AX is parallel to the line a and the line XG is parallel
to the line B; in this case, the line XG will in general correspond
to only part of the duration of the gating pulse. For either
possibility we denote by t.sub.1 the time which elapses from the
instant at which the electron in the limiting case enters the
decelerating field until the beginning of the gating pulse.
Consider now electrons which enter the decelerating field during
the application of a gating pulse. The variation of kinetic energy
with distance as the electrons travel towards the screen 9 may be
represented for three cases of electrons entering the decelerating
field at successively later instants by paths such as AMN, APG and
ARS in the diagram, the lines MN, PG and RS being parallel to the
line a and the line ARPM being parallel to the line B. The path AMN
corresponds to a case in which the electron enters the decelerating
field sufficiently early to pass through the screen 9, while the
path ARS corresponds to a case in which the electron enters the
decelerating field sufficiently late to be reflected. The path APG
corresponds to the limiting case of the last electron to enter the
decelerating field which will pass through the screen 9; we denote
by t.sub.2 the time which elapses from the instant at which this
electron enters the decelerating field until the end of the gating
pulse.
The total sampling time T is therefore given by the equation:
##SPC1## where t is the length of the gating pulse.
Assuming that (as will normally be the case in practice) V.sub.O,
V.sub.B and V.sub.G are all small compared with V.sub.D, it can be
shown, by substituting for t.sub.1 and t.sub.2 in equation (1),
that ##SPC2## m being the electronic mass; equations (2a) and (2b)
are respectively applicable when T.sub.o is greater and less than
t/.sqroot.Y. It should be noted that T.sub.o is approximately the
electron transit time through the deceleration region which would
be obtained if the screen 9 were maintained at zero potential.
Consideration of equations (2 a) and (2b) indicates that as V.sub.O
increases from zero, T will increase monotonically, giving rise to
the dispersion in sampling time. Clearly, the dispersion effect may
be reduced, for given values of t, V.sub.B and V.sub.G, by making
the quantity as large as possible, and therefore the quantity
T.sub.o as small as possible. With the arrangement described above
it is not normally practicable to utilize very high values for
V.sub.D, the potential of the electrode 13, since this would unduly
decrease the transit time of electrons in the drift region for a
given length of this region. It will, therefore, normally be
preferable to utilize a small value of d in order to bring about a
reduction in the dispersion of sampling time. The improvement
obtainable in this way may be illustrated by the following example,
in which V.sub.D, V.sub.B and V.sub.G respectively have values of
100, 2 and 8 volts, and t has a value of 5 nanoseconds. In this
case, where d has a value of 5.5 centimeters, the value of T when
V.sub.O is zero is 12.3 nanoseconds, while its value when V.sub.O
has a value of one volt is 16.9 nanoseconds, thus giving a spread
in the sampling time of 4.6 nanoseconds; when d has a value of 0.55
centimeters, the value of T when V.sub.O is zero is 5.9
nanoseconds, while its value when the value of V.sub.O is 1 volt is
6.1 nanoseconds, thus giving a spread in the sampling time of 0.2
nanoseconds.
An approximation to the conditions discussed theoretically above
may be achieved in practice by modifying the arrangement shown in
FIG. 1 to incorporate a third apertured screen disposed across the
end of the electrode 13 nearer to the screen 9, the third screen
being maintained at the same potential as the electrode 13 and the
spacing between the screen 9 and the third screen being made
relatively small; from the point of view of obtaining a
satisfactorily small dispersion of the sampling time, this spacing
should not appreciably exceed one centimeter when the potential of
the electrode 13 is 100 volts which would correspond to a value of
T.sub.o of about 3.4 nanoseconds. The provision of the third screen
however, has the disadvantage that the electrons must pass through
it on each transit of the drift region, so that unless it has a
very small shadow ratio the absorption of the electrons by the
third screen severely restricts the number of transits which may be
achieved in practice.
It is therefore desirable to consider alternative arrangements
which do not involve the use of a screen extending across the drift
region, and since in general these will not involve a linear
decelerating field, it is convenient to define a figure of merit
(comparable to the parameter T.sub.O) which is applicable generally
to any given electrode system and set of operating conditions in a
vacuum tube arrangement of the kind to which the invention relates.
Thus denoting by V the potential (relative to cathode potential) at
any point in the decelerating field when the second screen is
biassed negatively, we define a point Z as that point on the axis
of the electrode system at which the potential is increased to V +
0.1 V.sub.P when a gating pulse of magnitude V.sub.P is applied to
the second screen; the figure of merit T.sub.Z is then defined as
equal to , where V.sub.Z is the value of V at the point Z and
d.sub.Z is the distance from the point Z to the second screen. It
will be noted that for the linear case discussed above the value of
T.sub.Z is approximately equal to 0.95T.sub.0. In general it would
appear that in order to achieve a satisfactorily small dispersion
of the sampling time the value of T.sub.Z should not exceed 3.5
nanoseconds.
One arrangement which has proved satisfactory in practice for
obtaining a low value of T.sub.Z involves the provision between the
second apertured screen and the tubular electrode which defines the
normal drift region of a second, relatively short, tubular
electrode disposed coaxial with the first and maintained at a much
higher potential in order to provide a high value for the
decelerating field in front of the second apertured screen. With
this arrangement, the use of a plain tube as the high potential
electrode will normally be satisfactory only if the electron flow
is confined to a relatively small part of the cross section of the
vacuum tube adjacent the axis, since although a relatively low
value of T.sub.Z can be obtained it will be accompanied by a spread
in the sampling time over the cross section of the tube due to the
bulging of equipotentials into the interior of the second tubular
electrode. This limitation can however readily be overcome by
disposing across that end of the second tubular electrode nearer to
the second apertured screen a metal plate maintained at the same
potential as the second tubular electrode and having formed in it a
long narrow slot which will restrict the bulging of the
equipotentials.
One such arrangement is illustrated in FIG. 3, in which for
convenience only that part of the electrode structure is shown
which is modified as compared with the arrangement shown in FIG. 1.
The apertured screen 9' and the tubular electrode 13' correspond
respectively to the screen 9 and electrode 13 shown in FIG. 1, but
between and spaced from them is disposed a further tubular
electrode 21 having a diameter equal to that of the electrode 13'
and having a length approximately equal to its diameter; the
electrode 21 is disposed coaxial with the electrode 13', and the
electrodes 13' and 21 may conveniently be constituted, as is the
electrode 13 shown in FIG. 1, by conductive coatings formed on the
wall of the envelope (not shown in FIG. 3). Across the end of the
electrode 21 nearer the screen 9' is disposed a circular metal
plate 22 connected to the electrode 21 and having formed in it a
narrow slot 23 whose length extends along the major part of one
diameter of the plate 22. In one particular example, where the
diameter of the electrodes 13' and 21 is six centimeters, the
spacing between the plate 22 and the screen 9' is two centimeters
and the dimensions of the slot 23 are four centimeters by one
centimeter, the electrode 13' being maintained at a positive
potential of 100 volts and the electrode 21 and plate 22 being
maintained at a positive potential of 2,000 volts. With this
arrangement the figure of merit T.sub.Z has a value of about 1.7
nanoseconds, the spread in the sampling time being less than 0.2
nanoseconds and exhibiting virtually no variation over the area of
the slot.
As an alternative to the arrangement just described the arrangement
shown in FIG. 1 could be modified by incorporating a further
apertured screen disposed immediately in front of the screen 9, the
spacing between these two screens being very small (say a fraction
of a millimeter) and the further screen being maintained throughout
at a small negative potential. In this case, when the screen 9 is
biassed negative the electrons will be reflected before they reach
the further screen and so no absorption will occur. When a gating
pulse is applied to the screen 9, the positive equipotentials will
penetrate through the further screen so as to allow electrons to
pass through the two screens. With such an arrangement it would
appear possible to achieve very small numerical values of the
figure of merit T.sub.Z, with a corresponding spread in the
sampling time of the order of 0.1 nanoseconds; the arrangement
would however have the disadvantage of involving a relatively high
capacitance between the screen 9 and the further screen, which
would restrict the use of high-speed pulses.
* * * * *