2001 last Monthly Radio FAQ part 3
F.C. Trevor Gale
All the FAQ sections can be downloaded from the Web at this URL:
http://www.xs4all.nl/~tgale/oradio/ .
Part 3 - Dating / judging the age of Old Sets.
Original FAQ editor: Hank van Cleef, Gerard Tel (ger...@cs.uu.nl); now hosted on Trevor
Gale's site (email: tg...@xs4all.nl ).
This part of the FAQ is still under completion: however I include some of the information that
was in other sections prepared by the previous editors of this FAQ.
Set Types and Examples of Old Tubes/Valves
As is mentioned in the original section below, there are many cases where an old radio or like
equipment can be roughly dated by examining what valves/tubes they used (I use both
terms, since the U.S. "Tube" is referred to in U.K. English as "Valve" (except for picture or
other display tubes).
From an old publication in the 60's, long forgotten, I have a picture from an article showing
some older tube types together with brief descriptions.This is viewable under the following
URL: http://www.xs4all.nl/~tgale/oradio/tubetype.html and is not too clear, but it should give
an impression of some of the earlier forms that you might encounter. In addition to this, it is
worth noting (in Europe, at least) that radios with push-button, or 'piano-key' type band-
selection controls normally stem from the post 1948 years although the may be rare
exceptions. Another useful pointer is the tuning arrangements: early models (e.g. pre-1935)
often had no band-change switch since there were not many broadcast bands to cover.
Further, the actual tuning condenser / dial-knob drive can be a clue: in really early sets the
dial knob directly drove the tuning condenser, with no reduction mechanism - there simply
were less stations around so therefore selectivity and accuracy of tuning was not what was
required in later sets.
Dating old radios by the Tube/Valve Complement.
The development of vacuum tubes, both electrically and mechanically, advanced at a rapid
pace between about 1925 and 1950. The vast majority of radios sold for home entertainment
between 1920 and the late 1950's were built to various standard circuits. In most cases,
checking out what tubes are used in the radio will place it's date of manufacture within a few
years, identify which of the standard circuits it used, and give a some indication of the quality
of the set. Most radio repair technicians in the 1930-60 era did not need to look at schematics
most of the time, even when the problem was not a burned-out vacuum tube heater or
filament.
The tube complement is not always an accurate guide, except insofar as the presence of a
given tube indicates that the set was built after that tube was placed in production. You won't
find any 1932 radios using tubes with octal bases or 6.3 volt filament heaters, and you won't
find any prewar radios with 7-pin miniature tubes. But you may find a 1946 table radio built to
a 1935 design. There are also a few other design features that are very obvious on casual
inspection; I'll mention some of them as we go along.
In the following discussion, there are references to the example circuits shown in the RCA
Receiving Tube Manual RC-19, dated 1959. This manual is available in reprint from Antique
Electronic Supply. Examples 19-1 through 19-4 in particular show examples of four standard
circuits that were used, either identically or with minor modifications, in the majority of the
smaller "collectible" radios built from the mid-1930's on.
1. The five or six-tube AC-DC radio with 150 ma. tube heaters wired in series. Example
circuit 19-4 shows one of these radios, using 7-pin miniature tubes. This design is colloquially
called the "All-American Five" by some of us. The design was first built in 1939, using octal
tubes (i.e., 35Z5 and 50L6 in place of 35W4 and 50C5), so it is also called by some a "35Z5
radio" or a "50L6 radio." I list this design first, not only because it dominated home
entertainment radio production for over 20 years, but because it is a very simple
superheterodyne circuit. If you study this circuit and know what every component's function
is, and study an example radio of this design, you'll be prepared to trouble-shoot and repair
most post-1935 radios.
These sets do not have a power transformer, and could operate in places like mid-
Manhattan, which had 110 volts DC as its primary electrical service. Most of these were built
as table radios, although some were installed in small consoles and radio-phonograph
combinations. Virtually all clock radios use this circuit. These are generally AM-broadcast-
only. The tube set shown in the example is one of three common sets, having either octal,
loctal, or 7-pin mechanical design, but electrically equivalent.-Some sets, particularly in the
early postwar period, were built with mixtures of tube mechanical types, because of tube
shortages and availability,-and some sets used more than one configuration during their
production runs.
The six-tube version had an RF preamplifier, and was more sensitive than the five-tube.
Example circuit 19-3 shows the same basic design with an RF preamplifier stage, with tuned
output (three-section tuning capacitor). Many of the six-tube versions used resistance
coupling between the RF preamplifier and the converter stage (see Diagram no. 3, p. 339, in
RC-19, for a resistance-coupled pentode circuit). The six-tube version was often called a
"35L6 radio" because a 35L6, 35A5, or 35C5 was used, allowing connection of one more 12-
volt heater in the series heater string. In the fifties, some of these radios were built with a
selenium rectifier, omitting the rectifier tube. Also, a few manufacturers built a four-tube
version that omitted any IF amplification.
Several low-end "boatanchor" communications sets used this circuit, adding multiple tuning
coils and provisions for a beat-frequency oscillator. Notable examples are the Hallicrafters S-
38, S-41, S-119, S-120, and Ecophone EC-1 series; and the National NC-46 and SW-54.
The tube complements are:
a. First version, built primarily 1938-40. (note: this design is similar to the 19-4 example, but
is its immediate prececessor, so has a few substantia-l differences, noted below). 12A8 RF-
converter, 12K7 IF amplifier, 12Q7 detector-audio, 35L6 power output, and 35Z5 rectifier.
The first three tubes had small top caps for the signal grid connections, with either metal or
glass envelopes. The original glass tubes had a "G" suffix, indicating use of an ST-12
stepped bulb envelope. The major difference between this design and that shown in example
19-4 is the use of a 12A8, which uses a slightly different oscillator circuit than the 12SA7,
14Q7, or 12BE6. The other top-cap tubes are very similar to the single-ended octal tubes
which followed, varying primarily in mechanical construction. 12J8 and 12K8 were
sometimes used as converters as well. RC-19 unfortunately omits any circuits for these
converter tubes. This version uses a series resistor in the heater circuit because the heater
voltages do not add up to "near 120"). The proper place for this resistor, electrically, is
between the rectifier heater and the power amplifier heater.
b. Second version, built 1939-ca. 1960 12SA7 RF-converter, 12SK7 IF amplifier,-12SQ7
detector-audio, 50L6 power output, 35Z5 rectifier. This is almost the same radio, but using
single-ended tubes in the first three stages and a power output tube with a 50-volt heater.
The major difference is in use of a 12SA7 in place of the 12A8---these tubes are different
internally.-Note that the sum of the nominal heater voltages adds up to 122.8 volts, allowing
operation without need for any series resistor in the heater circuit.
c. Postwar version, 1945-mid '60's 12BE6 RF-converter, 12BA6 IF amplifier, 12AT6 detector-
audio, 50B5 power output, 35W4 rectifier. The only difference from b., above,is the use of
seven-pin miniature tubes. All are electrically identical to the octal versions above. Some
sets were built using a mix of seven-pin miniature and octal tubes, however, the presence of
seven-pin miniature tubes indicates that the set is postwar production.
d. Loctal tube version, 1940-ca. 1960 14Q7 RF-converter, 14A7 IF, 14X7 detector-audio,
50C5 power output, 35Y4 rectifier. Once again, the same radio as version b., using loctal-
base tubes in place of octal. Philco and GE were fond of using loctal tubes. Note that some
radios used a 14B8 converter, which is the same configuration in a circuit as the 12A8.
The six-tube configuration used the same tube type for both RF preamplifier and IF amplifier,
and the 35 volt heater version of the output tube. In most cases the RF preamplifier is
resistance-coupled to the RF-converter stage, and the radio used a two-stage tuning
capacitor.
Some later versions used movable slug tuning in place of a variable capacitor.-This variation
began around 1947, and became more common during the next decade.
2. Five or six tube AC-DC transformerless radios using 300 ma heaters wired in series.
These radios were the precursors of the 150 ma. series heater radios. Some of these radios
also included a tuning eye indicator, typically a 6E5. Total voltage drop of the series heater
string was 68-74-82 volts requiring an external voltage dropping resistor of some sort. These
radios often used "ballast" tubes or resistance wire in the line cord for this purpose.
a. Version using large-base 5, 6, or 7-pin tubes, 1935-50. 6A7 RF-converter, 78 or 6D6 IF, 75
detector-audio, 43 power output, 25Z5 rectifier. Most of these sets were built before 1938,
although a few manufacturers built them in the early postwar era. There are more variations
on this design than on the 150 ma. heater designs described above. As noted, some sets had
6E5 tuning eye tubes. Sets with shortwave often had a 76 triode as a separate local oscillator
for the 6A7.
b. Version using top-cap octal tubes, 1936-1950's. 6A8 RF-converter, 6K7 IF, 6Q7 detector-
audio, 25A6 or 25L6 audio, 25Z6 rectifier. This reflects the switch to octal tubes in 1936. The
first three tubes had small top caps for signal grid connection. The 25A6 is an octal version
of the 43; the 25L6 is a 25 volt heater beam power tube identical, except for heater, to the
35L6 and 50L6. The 25Z5 is a full-wave rectifier (two diode sections), and was usually
connected with the two sections in parallel. However, some manufacturers, notably Philco,
used the two sections to provide voltage doubling for B+. Radios with voltage doubler power
supplies are AC-only, as a voltage doubler requires alternating current to "pump" the doubler
circuit.
c. Version using single-ended octal tubes, 1939-50's. 6SA7 RF-converter, 6SK7 IF, 6SQ7
detector-audio, 25L6 output, 25Z6 rectifier. Once again, this is a "switch," this time to single-
ended octal tubes. Major circuit difference is in the 6SA7 circuit because of differences
internally between the 6SA7 and 6A8.
This version was generally not built as a "price leader" inexpensive table radio because of
the availabity of 150 ma. tubes that didn't require a dropping resistor in the heater circuit. It
was very often used as the basis for an upscale AC-DC radio. Some configurations that you
may run across:
1. Shortwave receiver using an additional RF preamplifier, separate local oscillator, and
second IF stage. The 6SK7 was used for the RF and IF stages, and a 6J5 as a local
oscillator.
2. Push-pull audio output, using two 25L6 tubes and a 6J5 as a phase inverter. This may be
combined with the RF-IF additions, above, and a tuning eye tube (6E5 usually).
Note that use of rectified line voltage gives a relatively low B+, a major limitation in the
transformerless design. The primary market for a "full house" receiver that had all of these
features would have been the DC service metropolitan areas, particularly New York City, and
that is the general area where most "odd-ball" configurations of transformerless sets can be
found today. In summary, all of the designs identified in items 1 and 2 above either used the
circuit shown in RC-19 example 19-4, or fairly simple variations of the design. There are very
few radios with these tube complements that vary markedly from the design, which was
established around 1932, and licensed to builders through Hazeltine and RCA patent
licenses. In general, the sets that deviate markedly from the standard circuit are a few
Philcos and Zeniths, and some off-brand sets that may have been marketed through chain
stores with chain store brand names.
3. Postwar AM-FM sets, 1945-up. These were made in two configurations: separate FM front
end, and common front end (i.e, RF, IF, mixer, and IF amplifiers. There are many variations
on both designs, using 7-pin miniature tubes, loctal tubes, or "hot" octal tubes. The 6SB7Y
was a "hot" 6SA7-type tube capable of self-exciting oscillation at FM frequencies, and the
6SG7 a "hot" replacement for the 6SK7. The presence of 88-108 MC FM in a radio always
means that it is a postwar set, as this band was not assigned to FM until April, 1945. Manual
RC-19 shows an example of an FM tuner in example 19-9. Many AM-FM sets "merged" AM
capability into the FM tuner design by using a bandswitch in the RF and converter stages,
and by connecting IF transformer coils for 455KC and 10.7 Mc. in series, the idea being that
the desired frequency will cause one or the other to resonate (high impedance) and the other
will appear as a low DC resistance. The bandswich would also select which IF fed the AM
detector, and which detector's output was used to feed the audio section. Example 19-9 also
shows both the limiter-discriminator and the ratio detector designs commonly used in FM-
capable sets.
This ends the "most common" AC-DC section. Now we will consider history, and some of the
other designs.
Example 19-1 in RC-19 shows a later battery-operated portable, using 7-pin miniature tubes.
This design was built after about 1934, originally using 5-6 pin tubes in ST-12 bulbs; later,
octal or loctal tubes. This circuit also is the basis for most later battery-operated "farm" sets,
some of which were built as floor consoles. Close study of the circuit will show its
resemblance to the 19-4 example. A very significant difference is the use of filament tubes,
and the method of using a back-bias resistor (R10 in the example) to develop grid bias
voltage for the output tube. Note also that a different local oscillator circuit is used for the
1R5. This circuit was often used in the "All American Five" design as well, and is not unique
to the battery design. Resistance values in example 19-1 have been chosen for operating
with a 67.5 volt B battery; otherwise, the circuit is suitable for operating with a 90 volt B
battery.
Example 19-2 shows a typical three-way portable. The term "three-way" may seem
confusing, when the radio can be operated either from the power line or from batteries.
However, the fact that it could operated from 110 volts DC as well as from AC lines was
considered noteworthy when DC domestic service was common; thus "AC or DC or internal
battery" are the "three ways." Note that a modern ricebox radio operating on an internal
battery or with an AC adapter is not "three way" as it will not operate from a DC line.
Once again, this is the Hazeltine-RCA standard circuit used in examples 19-1 through 19-5,
with specific provisions for the three way feature. Example 19-2 also shows use of a double-
tuned RF preamplifier. Notable are the use of series connection of the receiver filaments,
provision of a rectifier, and a changeover switch. In practise, many manufacturers provided a
dummy line-cord outlet inside the receiver. Plugging the line cord into this outlet would
mechanically actuate the changeover switch, placing the receiver on battery operation. When
studying this circuit, note in particular the order in which the tube filaments are wired, and the
use of an 1800-ohm resistor (R14) in the 3V4 filament circuit to provide a shunt-feed balance
current.
The order of connection of series-wired heaters and filaments is significant in series-string
sets. In this case, the 3V4 is connected to the high end to provide grid bias for operating, and
the shunt resistor provides some of the plate and screen currents for the tube. The rectifier
circuit shown is typical, although three way portables may use a 35Z5 or a selenium rectifier.
DC output from the rectifier is around 120 volts, depending on the rectifier used, which
requires a large dropping resistor to feed the receiver filaments. Note the use of two large
electrolytic filter capacitors, C28 and C29, connected to either end of the 3V4 filament. Small
filament tubes require "clean" DC power, thus these two capacitors filter out both residual
ripple from the half-wave rectifier and audio-frequency variations caused by varying power
draw of the power tube. This circuit arrangement is critical. If any filament opens, one or both
of those capacitors will charge up to the rectifier output voltage. Also, the design assumes
that the rectifier is part of the voltage-dropping string, and 1.5V filament tubes are limited in
their ability to handle out-of-tolerance filament voltage.
The circuit shown in figure 19-3 for an AC-operated receiver is the same as that in figure 19-
4, with several upscale features, and resistance values selected for operation at 250 volts B+
rather than 120. Note that the circuits for the 6BE6 converter, 6BA6 IF, and 6AV6 detector-
audio stages have the same configuration as those shown for those three stages in figure 19-
4. An additional 6BA6 RF preamplifier is provided for higher gain and better selectivity. A
pair of 6AQ5 tubes provides push-pull output. The second 6AV6 placed ahead of the lower
6AQ5 grid circuit inverts the audio signal for grid drive, with "approximately unity gain,"
determined by the tapped grid leak (470K/8200 ohms) in the top 6AQ5 circuit. This particular
circuit is a classic example of older home entertainment engineerin-g, and there is much to
criticize in its selection over the use of a twin-triode balanced paraphase using a 12AX7 or a
6SN7. Why was it chosen? Habit, probably---it was a good choice for 1932.
The main feature of this set which differs from AC-DC configuration is, of course, the use of
a power transformer and a 5Y3 full-wave rectifier. The configuration of the rectifier circuit
was one of the earliest and most durable circuits in the history of tube-type home
entertainment radio. This later configuration uses a 5Y3 instead of an 80, has larger filter
capacitors (20 mfd rather than 8 or 10 mfd), and a resistor in place of an inductance between
the two filter sections. Older radios most often used a speaker field coil between the two filter
sections, partly because Alnico magnets were not available until the late thirties, and partly
because inductance at this point compensates for using smaller capacitance values to get
good filtering.
Note the configuration of the screen circuit for the 6BE6 and two 6BA6's. All three screens
are connected together. This is poor design, and likely to cause parasitic oscillations. The
circuit in figure 19-4 also shows the screens connected together, but in this instance, there
are only two screen, in stages that operate in opposite phase, so any coupling between the
two stages has a negative feedback effect.
Older radios:
Home entertainment radio began in 1920. KDKA in Pittsburgh generally has gotten credit for
being the first commercial broadcast station. The two major receiving tubes available at the
time with the UX201 and the UV199, as they were called at the time. The UX201, later
revised and called 01A was a low mu triode. The V99, as the UV199 came to be termed, was
derived from a telephone amplifier triode, developed during WWI. Several manufacturers
built sets, but the most predominant in the collector market is the Atwater Kent neutrodyne
TRF set using 01A's driving headphones. A standard inexpensive set used regenerative
feedback to achieve gain. These were prone to oscillate, squawk, and whistle, and created
no end of radio frequency interference, and rapidly lost favor, particularly in high-density
metropolitan areas. The first commercially significant superheterodyne receiver was the RCA
"catacombs" receiver of 1924. This set used V99's, a 42 KC IF frequency, and a headphone-
driving-a-horn "loudspeaker." Both the A-K and the RCA sets required three DC voltage
supplies. The A supply (5 volts DC for 01A, 3.3 volts DC for V99) heated the filaments. The
B supply, typically 90 volts, provided plate voltage. The C supply, ranging between 9 and 15
volts, and connected as a negative supply, was used to bias the tube grids. RF gain was
controlled by a rheostat which controlled the filament voltage. These three voltages were
supplied by lead-acid storage batteries, with a Tungar bulb charger for charging the batteries
when the radio was not being used. All of the RF stages, and the catacombs superhet local
oscillator, were tuned by separate dial knobs.
If this sounds like the definition of a kloodge, it was. I had examples of both an O1A Atwater
Kent and an RCA "portable" (ran on dry batteries) catacombs set, complete with lead-acid
batteries and Tungar charger, at the end of WWII. These sets sold by the thousands, but
were obsolete by 1929, and most of them were discarded when their storage batteries wore
out. Worth noting that "Philco" is a contraction of "Philadelphia Storage Battery Company." It
is also worth noting here that RCA, or "Radio Corporation of America," was not a separate
company until 1929, but a patent pool and sales company owned by General Electric,
Westinghouse, and AT&T. The phonograph fans will, no doubt, describe how the Victor
Talking Machine Company and Radio Corporation of America became RCA Victor.
Automatic volume control methods were developed around 1925. AVC, which is synonymous
with the term "Automatic Gain Control" (AGC), allowed sets to operate at much higher input
sensitivity, and to reduce that sensitivity to prevent overloading in the presence of a strong
signal. Methods of tracking RF stages and a local oscillator operating at some difference
frequency were also developed in the mid-late 1920's. The final developments needed to
build a mains-powered single knob tuning "modern" superheterodyne radio were filaments
capable of working on AC without developing hum, a suitable high-voltage rectifier, and a
tube with high plate resistance. The first two appeared around 1928 in the form of the 26 and
71A tubes and the 80 rectifier. While these were not the actual "first" devices, they appear in
almost all of the early mains-powered radios. The third came about a year later in the form of
the UY224 tetrode, later known as the 24A. The 24 also had another recent innovation, the
indirectly-heated cathode, which allowed the cathode element of each tube to "float" at a
different voltage from the heater supply DC reference.
Problems with secondary emission from the 24 were "cured," more or less, by processing the
plate material to reduce this emission. This produced the 24A. However, a more permanent
fix was to include a third grid to "suppress" the reverse current resulting when plate voltage
was lower than screen voltage. The 57 and 58 pentodes were the result. Both have 2.5 volt
indirectly-heated cathodes. However, the 58 has a characteristic known as "variable-mu."
Actually, with pentodes, one considers transconductanc-e, and what "variable-mu" actually
does is to reduce the transconductance as the tube is more heavily biased. The feature is
desirable in circuits with AVC. These pentodes showed up around 1931. The pentode power
amplifier was also introduced around the same time, with the 47 replacing the 45 in many
designed of the 1932-34 era.
The last significant development in tube design for AM broadcast radios was the
development of a single tube with two control grids to serve as a self-exciting local oscillator
and mixer amplifier. The 2A7, quickly replaced by the 6-volt-heater equivalent 6A7, was the
predominant design, and the 6A7 was used very commonly until after 1940. The 6L7 also
was introduced fairly early. This is a mixer that is not designed to operate as a self-oscillator,
and was used, particularly in communications sets, with a separate local oscillator, until the
1950's.
Availability of a single tube for the superheterodyne oscillator-mixer function was essentially
the death-knell for TRF designs. Another contemporary-development which entered
production in 1933 was the 2E5 "tuning eye" tube, which varied a shadow area on a visible
target as an inverse function of the control grid voltage. TRF sets were built into the 1950's,
but are not very common. They tend to be either very cheap radios for use in metropolitan
areas with strong signals or in high end sets where the broad bandpass allowed "high fidelity"
(though the AM stations actually only transmit a signal that has 5KC as the 3db half-power
point in the modulation).
Availability of components for a vibrator power supply made automobile sets operating from
6 volts DC practical. There was a wholesale switch from 2.5 volt heaters to 6.3 volt heaters in
1934. The 2.5 volt heater series of tubes quickly became obsolete. The switch to 6.3 volt 300
ma. filaments was parallelled by development of a two-diode rectifier and an output tube with
25-volt 300 ma. heaters, making series string wiring of the heater circuit practical. These are
the 300 ma. heater transformerless sets described above, which date from about 1934.
Octal-based tubes enter the picture in 1936. Many of the original designs were built in self-
shielding steel envelopes. Metal octal tubes were built with a flat "button" glass seal, which
allowed much shorter electrode lead connections. Early glass octal tubes continued to use
the older "press" design, with relatively long leads. RF and AF tubes in the original octal
series had small top caps for connection to their control grids. It was not until about 1939 that
single-ended tubes entered production.
Development of a button seal that could be used with glass envelopes allowed manufacture
of metal-based "loctal" tubes. These entered production in 1939. At the same time, a
cylindrical bulb for glass tubes also entered production, allowing closer spacing between
tubes.
Experimental FM became a commercial broadcast enterprise in 1940. The original FM band
began at 42 megacycles, and production of home entertainment receivers to receive that
band began in 1941. The band originally overlapped the experimental television band (later
channel 1, 48-54 megacycles). The FM band was reallocated to 88-108 megacycles in the
spring of 1945, thus a set with 88-108 capability is postwar.
Another "strictly postwar" feature is the 7-pin miniature tube. The 9-pin miniature followed
around 1949.
A few tubes were "survivors" through the 1928-50 period. The standout among these is the
80 rectifier, which was still being used in new production in the mid-1950's. The 5Y3GT which
replaced it is nothing but an octal-based version of the 80. The 2A3 and 45 power triodes, as
well as the less-common 6A3 were all used from the early 1930's until well into the 1950's.
There remains today something of a cult that believes that these triodes are the only audio
power tubes worth considering. All of these tubes use filament cathodes, and the most
practical circuits for using them required a separate filament winding, elevated to the 40-60
volts needed to bias these tubes near cutoff.
You can E-mail the host of this FAQ (Trevor Gale) at tg...@xs4all.nl.