The Story of the Vale

Part 3

C. H. Gardner

In the third article of his 4 -part series, C. H. Gardner deals with

what is probably the most important part of the valve -its cathode - the necessary high vacuum was the
heater assembly result of considerable improvements
in the pumping equipment together
with the firing of "getters" to
remove the last traces of gas as
pumping was being completed and
just before the valve was sealed off.
NTHESE DAYS OF TRANSISTORS WE pumped. The gases released during Also, the internal components of the
I may be apt to forget that the
valve, unlike the semi- conductor,
has "consumable" parts which limit
this chemical change are sucked
away by the evacuating pump. The
valve is then "aged" by running the
valve were raised to a high tempera-
ture to drive off any occluded gases,
the components being raised to the
its useful life. Such things as broken heater at a higher than normal necessary temperature by eddy cur-
heaters, insulation breakdown and temperature with a positive potential rents induced in them from alterna-
misplaced electrodes could all cause applied to the other electrodes. This ting currents of suitable frequency
failure of the valve, but perhaps the "ageing" causes a certain percentage passed through a coil surrounding
greatest problems in practice have of metallic barium and strontium to the bulb. The getter was fired in a
been concerned with the life of the be formed in the emissive coating similar manner just before the sealing-
cathode. Many of these problems and it is the structure of the aged off process took place. These pro-
are common to both directly and coating, containing atoms of these cesses occurred in sequence on a
indirectly heated cathodes, and the metals, which provides the effective rotating machine, the assembled valve
latter will be dealt with here in order emissive material. The chemical being fed in by an operator and the
to illustrate the complexity of the process is involved, and an important pumped, gettered and sealed valve
matter. The overcoming of these point to note is that, should any being removed on the completion of
problems must be considered as a oxygen remain in the valve after the cycle.
major development in the history of evacuation or should it be released Various materials have been used
the valve. from the glass or metallic com- for getters, the necessary piece of
ponents whilst the valve is in material being fixed at a suitable
Indirectly Heated Cathode operation, this will reoxidise the point of the assembly. This vital
An indirectly heated valve employs all- important barium and strontium process has many complications, as
a heating filament which is insulated metallic atoms. The result will be a it is not easy to ensure that the fired
from the cathode proper by a coating reduction in the emission of the getter reaches all the gas molecules
of, generally, aluminium oxide. The valve. or that the final residue is not
cathode consists of a nickel tube into It will thus be seen that an deposited in undesirable places.
which the insulated heater is in- important stage in valve development The amount of chemical, phy-
serted. This nickel tube may be was given, firstly, by attaining a very sical and engineering research and
coated with a mixture of strontium high vacuum and, secondly, by
and barium carbonates which are ensuring that this high vacuum
converted into oxides whilst the would be maintained whilst the valve Depending upon the processing used,
the getter might alternatively be fired as
valve is being processed after was in operation. a separate operation after sealing off, with
assembly and when it is being The ability to obtain and maintain equivalent results.

NOVEMBER 1965 257

 from the heater to the cathode was
of importance in order to keep
wattage low and heating time within
reasonable limits. The insulation
had, also, to be of such a nature that
it would withstand the heating and
cooling of the heater as the set was
switched on and off.
Alumina (aluminium oxide) was
the material generally used for this
insulation. It coated the heater and
obviated the earlier arrangement in
which the heater was contained in an
insulating tube of magnesium oxide.
Gradual improvement over a con-
siderable number of years in the
technique of coating the heater and
in the purity of the alumina has made
leakage or breakdown of the cathode-
heater insulation virtually a thing of
the past, in spite of the increased
potentials now often present. Heat -
ing-up time has also been consider-
ably improved.
Breakdown of the heater itself
was a fairly common trouble with
the earlier valves. Considerable
strain was put on the heater wire by
the frequent heating and cooling
caused by switching the set on and
off. With the reduction in valve size
and, consequently, the size of the
cathode tube, heater design called
for some ingenuity. Basically, the
heater consisted of a spiral of wire.
Two popular battery valves of the early 1930's -The PM2HL and PM2A. Later, the coiled coil heater came
These had 2 volt filaments and were widely used in portable receivers into general use, the heater being
bent back on itself to form an
inverted V. More recently the single
wire has been replaced by a stranded
development necessary to improve and a big step was given by the wire providing greater flexibility.
the plant and processing in these evolution of a cathode with flat sides The later constructions have also
directions whilst, at the same time, and coated with the emissive material been effective in the reduction of
speeding up production may be well by a new method. This provided a hum.
imagined. The development has hard smooth surface which allowed Rather a high proportion of space
been continuous ever since the a cathode -grid spacing of less than has been devoted to the development
1920s. a twentieth of a millimetre. The old of the cathode -heater assembly, but
method of spraying on the emissive it must be pointed out that without
Design Difficulties material provided a random rough- a considerably high performance in
The design and manufacture of the ness of such a nature that very fine this component not only would a
cathode presented many difficulties. cathode -grid clearances caused far reasonable degree of reliability and
A typical example was given by the too great a spread of valve charac- life for the valve have been un-
development of resistance between teristics. The older type of cathode obtainable but the development of
the nickel tube and the emissive construction sometimes caused the remainder of its assembly would
material as the valve was in use, this cathode material to be carried over have been impossible.
being often due to minute traces of to the grid, resulting in this electrode
impurity in the nickel core. The itself emitting. In the case of rectify-
growth of such interface resistance ing valves such material could cause General Assembly
would gradually affect the perform- sputtering between cathode and A major development which has
ance of the valve. anode, often with dire consequences been of general application has been
In order to obtain maximum to associated components. the change in the general assembly
efficiency, the spacing between grid We cannot leave the subject of the towards greater rigidity, reduction
and cathode should be as small as all-important cathode without refer- in size and improvement in r.f.
possible. In the very early valves ence to the heater itself and to the performance. At the same time,
this spacing was quite considerable insulation between cathode and alteration in construction has re-
and, even some twenty years ago, it heater. There were conflicting quired much modification of the
was usual to find a spacing of a requirements for this insulation. manufacturing plant.
seventh of a millimetre or so. Series operation and circuit require- This major development in con-
Reduction of this spacing has ments resulted in many cases, in a struction will be dealt with next
gradually taken place as the result considerable difference of potential month, in the concluding article in
of improved forms of grid construc- between cathode and heater. At the this series.
tion and greater rigidity of assembly, same time, good conduction of heat (To be concluded)

258 THE RADIO CONSTRUCTOR

 IN LAST MONTH'S ARTICLE IN THIS SERIES WE but too low a value in this capacitor is liable to
continued our discussion on the triode valve, result in undesirable changes in performance.'
and started by examining the operation of It is desirable that the anode of the infinite
grid current bias. We then desçribed the manner impedance detector should have a low impedance
in which an a.f. amplifier triode may be coupled path to chassis, and this may be provided by con-
to the output of a diode detector, and finished by necting a bypass capacitor between anode and
discussing the grid leak detector. chassis, as represented by C2 in Fig. 324 (a). Such
We shall now deal briefly with another type of a capacitor should offer a low reactance both at
detector -the infinite impedance detector -after audio and at radio frequencies (whereupon it
which we shall proceed to the characteristics and "bypasses" these frequencies to chassis) and a
constants of the triode. representative value would be 0.21i.F. A series
resistor, R2, is also introduced in this diagram.
The Infinite Impedance Detector This resistor forms, with C2, a decoupling filter
The basic circuit of the infinite impedance detector which ensures that r.f. and a.f. voltages on the
appears in Fig. 322. In this diagram an amplitude anode of the triode do not appear across the h.t.
modulated r.f. signal appears across the tuned supply lines, where they might become coupled
circuit and is applied to the grid of the triode. into other stages in the equipment. R2 and C2
The detector load is in the cathode circuit, and also prevent undesired r.f. and a.f. couplings
consists of R1 with C1 in parallel. R1 could have, from other stages to the detector. R2 is not essential
typically, a value of 10ok12, and C1 a value of if there is no risk of such couplings. A secondary
200pF at frequencies in the range occupied by the function for R2 is to reduce the h.t. voltage applied
medium and long wavebands. to the triode, should this be desired.

understanding
INTRODUCING MUTUAL CONDUCTANCE
AND

ANODE A.C. RESISTANCE

By W. G. Morley

The operation of the infinite impedance detector

is quite simple. It can be considered initially as a
cathode follower, in which the cathode goes positive
when the grid goes positive. The presence of
Cl modifies the cathode follower action, however,
because this capacitor becomes charged by the
radio
The waveform of Fig. 323 shows that there is
positive peaks of the r.f. signal appearing on the both an r.f. voltage and an average direct voltage
grid. Between positive peaks the capacitor discharges across the cathode load of the infinite impedance
into R1, with the result that the voltage across this detector. This is the same situation as occurs
resistor takes up the form shown in heavy line with the diode detector load, and we remove these
in Fig. 323. This is similar to the waveform given two unwanted voltages in much the same manner
by a diode detector, and an average a.f. voltage as we did with the diode detector. The final circuit
appears in the waveform which corresponds to is shown in Fig. 324 (b), in which R3 and C3 form
the a.f. which originally modulated the r.f. signal. a low-pass filter to attenuate the r.f. voltage, whilst
The infinite impedance detector draws negligible C4, which has a low reactance at audio frequencies,
current from the tuned circuit which connects to prevents the direct voltage from being passed to
its grid and offers, in consequence, a very high t If the value of C, is made too low, the valve may present
a negative
and
input impedance. It is from this factor that the resistance to the tuned circuit, with a consequent increase in Qwhen
a tendency towards oscillation. (Negative resistance is given
term "infinite impedance" derives. The impedance an increase in voltage causes a decrease in current. Oscillation will
may be increased by reducing the value of C1, be discussed in a later article.)

259
NOVEMBER 1965
 Hi+ the purpose of valve characteristics and constants
after a basic knowledge of the manner in which
Tuned the valve is to be used has already been acquired.
circuit It may be added that a valve constant is a figure
which represents a particular attribute of the valve.
Modulated
Detected The two constants we shall discuss this month are
R F input
output closely connected with sets of curves which illustrate
the valve's characteristics, and the sets of curves
will be dealt with at the same time.
The characteristics and constants we shall now
consider may all be found by means of the circuit
Fig. 322. The basic infinite impedance detector shown in Fig. 325. This is similar in principle to
Fig. 299 in the August issue, which we employed
the following stage. R4 ensures that the average for obtaining TaVg (anode current -grid voltage)
voltage on the right hand plate of C4 is at chassis curves, with the exception that we now have a
potential, and may serve as the grid resistor for a variable anode voltage provided by a potentiometer,
following a.f. amplifier valve. and the grid voltage available cannot go positive
As with the diode detector, it is necessary to
keep the a.c. /d.c. cathode load ratio close to unity HT+
if distortion at the higher modulation depths is
to be avoided.2 This requirement may be satisfied
by giving R4 a value which is four or more times
that of R1.
The main advantage of the infinite impedance Tuned
detector is that it has a high input impedance. circuit
Because of this it offers very little damping on
any tuned circuit to which it is connected, whereupon
the Q of the latter remains virtually unaltered. Detected
output
Modulated
Valve Characteristics And Constants
R F input
Apart from the fact that we looked at the IaVg
characteristic curve of the triode in the August
issue we have not yet discussed in detail the
characteristics and constants of the triode. There
are three important constants, and they provide
(a)
a measure of the valve's capabilities. As we shall
see later, these constants also apply to amplifying
valves having more complicated electrode structures HT+
than has the triode, but we shall discuss them in
terms of the triode only at this stage. Triode
characteristics and constants were not dealt with
immediately after we introduced the triode valve
because the writer feels that it is more helpful
for the beginner to gain an understanding of simple
applications and circuits first. It is easier to grasp
Detected
2 See "Understanding Radio" in the June 1965 issue. output
Modulated
C1 discharges into R1
R.F input

Average (b)
Cathode
voltage
relative
'dig! 116 direct
Itage
Fig. 324 (a). In a practical infinite impedance
detector, it is desirable to have a by-pass capacitor
veroge between anode and chassis. C2 carries out this
to chassis function here. R2 gives decoupling, and may be used
A.F volt-
age also to provide a lower voltage on the anode that
0111111111111111111 111111111114 that available on the h.t. positive line
(b). For practical operation, the infinite impedance
detector needs a filter to remove the r.f voltage across
Fig. 323 The waveform appearing at the cathode of the cathode load. This is provided here by R3 and C3.
the triode in Fig. 322 is shown here in heavy line. Also needed is capacitor C4, which has a low reactance
The positive r.f half - cycles of the signal are added at audio frequencies and which prevents the average
to assist the explanation direct voltage appearing at the output

260 THE RADIO CONSTRUCTOR

 current to change of grid voltage) would have been
0.5 =2mA per volt, just as before.
0.25
The lower ends of the curves of Fig. 326 (a)
are noticeably curved and we would find that if,
on the Va =100 curve for instance, we attempted
to find the mutual conductance by measuring the
change in anode voltage for a change in grid voltage
between X and X' in Fig. 326 (b), we would get a
different answer than we would for a grid voltage
change between Y and Y'. This is entirely due
to the pronounced curvature at these points. Nor-
mally, mutual conductance measurements are
taken at the straighter, or more linear, parts of
the IaVg curve where the same ratio is more likely
to result for both large and small changes in grid
Fig. 325. A circuit which allows the characteristics voltage. If mutual conductance measurements
of triode to be found. Sources of direct voltage
a have to be taken for a grid voltage which appears
other than batteries may, of course, be employed at a pronouncedly curved portion of the IaVg
characteristic curve, it is better to use as small a
of the cathode. As wé saw in the previous article, 30
the grid is not normally allowed to go positive
of the cathode in practical applications and so
characteristics are not given for positive grid
voltages.3 The circuit given in Fig. 325 enables
us to vary both the anode voltage and the grid
voltage, and to measure the anode current which
flows.

Mutual Conductance
Since we have already examined the IaVg curve,
this provides a useful starting point for the first
constant which we shall consider. Fig. 326 (a)
shows the IaVg curves for a typical voltage amplifier
triode at anode voltages of 100, 150 and 200.4
These curves tell us that, as grid voltage goes
more positive, the anode current increases. They Grid voltage - Vg
also tell us, obviously enough, that a change in
grid voltage causes a corresponding change in (a)
anode current.
The latter attribute is defined by the first constant 15
to be discussed, this being mutual conductance.
The mutual conductance of a valve is the ratio
of change in anode current to change in grid voltage,
the anode voltage remaining constant. Mutual lo
conductance is expressed in milliamperes per volt. Anode
If, at a constant anode voltage, we change the current
grid voltage of a triode by volt and find that
1 la(mA)
this causes the anode current to change by 2mA,
then the triode has a mutual conductance of 2mA
per volt. This may be abbreviated to 2mA /volt
or 2mA /V. It is not necessary to find the ratio -2 0
by using a change in grid voltage of exactly one -8
volt. If we had measured the change in anode
current for a change in grid voltage of 0.25 volt,
and found it to be 0.5mA, then the mutual con- Y

ductance (which is the ratio of change of anode Grid voltage -Vg

This statement may appear to conflict with the fact that, in last
the
grid current bias and grid leak detector circuits we examined cathode
(b)
of the
month, the grid was allowed to go very slightly positive operation was
during positive signal peaks. However, this method ofshould function Fig. 326 (a). laVg curves for a voltage
Typical
due to the requirement, in these circuits, that the grid amplifier triode
as the anode of a diode, whereupon small positive excursions at the (b). Measurements of mutual conductance at a curved
grid become inevitable. The positive grid voltage and grid current
in these two instances are kept to a low figure by reason of the as- section of the laVg characteristic may give different
sociated circuit design. the August results according to the magnitude of the change in
4 These are the same curves as appeared in Fig. 300 in grid voltage
issue, but without the section corresponding to positive grid voltage.

261
NOVEMBER 1965
 we have to go back to our basic electrical quantities.5
Resistance is measured in ohms, and is equal to
Iwhere E is the applied e.m.f. in volts and I is
the current, in amps, which flows. Conductance
is the reciprocal of resistance, and the basic unit
of conductance is the mho (the word "ohm" spelt
backwards). Thus, conductance (in mhos)=
1 If a circuit component has a
resistance (in ohms).
resistance of 3 ohms, it has a conductance of } mho.
Since conductance is the reciprocal of resistance,
it is equal to . Now, our mutual conductance is
É
dIa
which is also a current divided by a voltage,
dVg
and we may see, in consequence, how the term
V9 "mutual conductance" derives. "Mutual" means
"conjointly shared" and so "mutual conductance"
Fig. 327. Two IaVg curves drawn on the some axes. refers to a conductance which is conjointly shared
Curve A has a steeper slope that curve B and represents between grid and anode. A mutual conductance of
o higher mutual conductance 2mA per volt may be expressed, in terms of amps
grid voltage change as practicable, in order to and volts, as 0.002 amps per volt or 0.002 mhos. It
obtain an accurate answer for the particular grid
voltage concerned. In valve manufacturers' litera- is more convenient to use the micromho (one
ture, mutual conductance is normally quoted for millionth part of a mho) than the mho, whereupon
specific values of anode voltage and grid voltage, our mutual conductance becomes 0.002 x 1,000,000
and these figures will almost inevitably apply to a
linear part of the IaVg characteristic curve. =2,000 micromhos. It may be seen from this
Mutual conductance may also be referred to that, to express mutual conductance in micromhos,
as slope. This is because the mutual conductance it is only necessary to multiply the mA per volt
figure defines the "slope" of the IaVg curve. In figure by 1,000. Similarly, a mutual conductance
Fig. 327, curve A has a steeper "slope" than has figure in micromhos may be changed to mA per
curve B, whereupon it represents a higher mutual volt by dividing by 1,000.
conductance. A final point is that a set of IaVg curves, such as
The abbreviation employed for mutual conduct- those shown in Fig. 326 (a), are referred to as the
ance is gm, and the definition for mutual conductance mutual characteristics of the valve concerned.
may be expressed in the following equation: They may also be described as the transfer
characteristics.
gm--
dIa
dV (Va
constant).
Anode A.C. Resistance
Non -mathematical readers should not worry about The second important constant we shall consider
the letter "d" which appears before "Ia" and "Vg ". is the a.c. resistance, or anode a.c. resistance, of a
In the fraction this letter merely signifies that the triode. This is not a new concept to us because,
denominator (dVg) is "change in Vg" and that the in the July issue, we examined the a.c. resistance
numerator (dIa) is "corresponding change in la ". of the diode. We saw that the resistance to an
To quote our previous example, if the change alternating current offered by a diode depends
in grid voltage (dVg) is 1 volt and the corresponding upon the slope of the Iaya characteristic curve.
change in anode current (dIa) is 2mA, then the The a.c. resistance offered was demonstrated by
equation tells us that gm=-2 per volt. We considering the alternating current which flowed
at a known alternating voltage, and it was pointed
would get the same answer with the grid voltage out that this a.c. resistance had no connection
change of 0.25 volts and the corresponding anode with the d.c. resistance of the diode.
current change of 0.5mA which was also mentioned. The same applies to the triode, and we may
The term "Va" appearing in the equation stands, define anode a.c. resistance as the ratio of change
of course, for anode voltage. in anode voltage to the corresponding change in
In American terminology, "mutual conductance" anode current, grid voltage remaining constant.
is usually referred to as transconductance, or grid- Anode a.c. resistance is expressed in ohms. As
plate transconductance (the term "plate" being was pointed out when we dealt with the diode,
employed instead of "anode "). Mutual conductance the term "anode a.c. resistance" is abbreviated
is also expressed in micromhos, instead of in mA
5 Resistance and conductance were covered by "Understanding
per volt. To understand why this unit is employed Radio" in the October 1961 issue.

262 THE RADIO CONSTRUCTOR

 V9=0 -25 -5 -75 -IO
io
to ra. If a change in anode voltage of 10 volts
corresponds to a change in anode current of 2mA, 8
or 0.002 amps, then the anode a.c. resistance in
6 -15
ohms is .= 5,000.
0. (mA) 4
We may express anode a.c. resistance in the form
2
ra_
dia (Vg constant). 300
In this equation, the letters "d" carry out the same 100 200
function as with the mutual conductance equation. va
The term "dia" means "change in Ia" and the
term "dVa" means "corresponding change in Fig. 328. A set of IaVa curves for a typical voltage
Va ". The similiarity of the equation to that for amplifier triode. Each curve corresponds to a

resistance, R Ì , is readily apparent.

different grid voltage

and anode current, by checking the slope of the

We explained anode a.c. resistance with the Iaya curve at the corresponding point. In valve
diode by drawing its IaVa characteristic curve and manufacturers' literature, anode a.c. resistance is
we shall now do the same for the triode. However, normally quoted for specific values of anode
the grid of the triode introduces a complicating voltage and grid voltage, these applying to a
factor, and it is necessary, if a full picture of the relatively linear section of the IaVa curve.
performance of the triode is to be given, to draw Anode a.c. resistance can be looked upon in
not one Iaya characteristic curve, but a relatively the same manner as the internal resistance of an
large number of curves, each corresponding to a a.c. generator. It may also be referred to as anode
different grid voltage. A typical example is shown impedance or anode incremental resistance, and, in
in Fig. 328, in which we see a "family" of IaVa American literature, as plate resistance.
curves with grid voltage as parameter.6 A set of IaVa curves, such as those shown in
We can find anode a.c. resistance at any grid Fig. 328, are described as the anode characteristics
voltage and anode voltage, or at any grid voltage of the valve.
Next Month
6 A parameter is a constant in a family of
functions which can be
lta In next month's issue we shall describe the
given arbitrarily chosen values. In this instance, the8grid
is a parameter because curves are drawn at V=-40, S
=- and remaining important constant of the valve, this
so on. In the IaVg curve of Fig. 326 (a), anode voltage is the para-
meter.
being amplification factor.

Club (vents
The Slade Radio Society Northern Heights Amateur Radio Society
Hon. Sec.: D. Wilson, 177 Dower Road, Sutton Coldfield. Hon. Sec.: A. Robinson, G3MDW, Candy Cabin, Ogden,
Halifax.
November 12th -Film Show presented by Don November 10th -Tape Recorders, by F. Thistle-
Wilson. The programme will include some of the latest
instructional films on radio subjects in addition to films thwaite.
of more general interest. November 16th -Mullard Film Show, Bradford.
November 26th-Annual General Meeting. November 24th-Electronic Logic, by G. E. Craven.
The Roding Boys' Society
An invitation is extended by the Society to any local
boy, interested in radio or science generally, to visit
Derby and District Amateur Radio Society Wanstead Community Centre, The Green, London, E.11,
any Tuesday at 7 p.m. More distant boys can write
to
Hon. Sec.: F. C. Ward, 5 Uplands Avenue, Littleover, Derby.
R. J. Lipscombe, 187 Markhouse Road, Walthamstow,
November 20th and 21st-Second R.S.G.B. Top London, E.17.
Band Contest.

263
NOVEMBER 1965