Single-Side-Band Techniques at Microwave Bands, Chapter 1
@ OK1AIY
____________
During the 1970s, the technology and materials became available to start
experiments with SSB also in microwave bands. The lowest band was 23 cm (1296
Mhz), so the first designs were tested there. The wide band had assigned a
narrower section for international use, 1296-1298 Mhz.
Design tinkering was demanding on time and materials, and any available source
was used. Each designer utilized his own approach and ideas, a mutual
consultation was done by talks during regular seminars as well as by papers in
technical magazines. Like in my earlier papers on SSB in VHF bands, I would
like to present how I fought myself, and others can compare my approach with
theirs. After many years the memory fails on details- there were many
interesting events. During the related era, the progress was significant in
the „professional“ as well as in the amateur designs and related fields. To
offer a better explanation I would give examples only indirectly linked to SSB
but illustrate the development and our attempts. Some actions may seem funny
today but would not distort the colorful old times.
How we worked at 23 cm back then
In the introduction, let us get back to 1960s and show the readers our
technology and what communication was then possible. In the surrounding Europe
and in our country (Czechoslovakia) one could find several stations with a
modern equipment using quartz-crystal controlled oscillators and vacuum-tube
frequency multipliers. Favorite tubes in power multipliers and amplifiers were
LD11, output power of up to tens of Watts. Our favorite design manual was the
book by Antonin Rambousek „Amateur very-short-wave techniques“. The book was
were well written and author's designs were simple and gave good results. Most
receivers were conceived as down-converters for known German war-surplus
receivers like „Emil“, Fug16 or E10AK. The front ends used bare mixers like
1N21, 1N23 and similar silicon diodes. RF preamplifiers were not yet
available. During contests the solo-oscillators were mostly used, the designs
were well done and reliable. Let us remember the Kolin group, with a genius
designer Vraťa Poula (later OK1WGO, Fig.1), and also exciting descriptions of
advanced designs in Amateur Radio magazine, by a group of Ostrov nad Ohří. The
designers Václav Vachuška, OK1YN, and Mirek Klusák, OK1VMK, are admired even
today for their skills (Figs. 2,4,5). In Moravia there was an active group of
experimenters that during contests climbed quite unaccessible hilltops, and
had to power their rigs by rechargeable batteries only. Their solo-oscillators
were advantageous for their low power demands, and their operations continued
into 1970s (Fig.6). After one contest was evaluated, OK1VAM claimed „the Field
Day was again won by the Moravians with their ugly solo-oscillators...“
Fig.1 Vráťa, OK1WGO, and Jára, OK1AEW, with their rig for 23 cm band, 1954:
Fig.2 Schematic diagram of Václav Vachuška's (OK1YN) 23-cm transceiver, with
RCA 5794 triode. As a transmitter it operated as a solo-oscillator, as a
receiver as a super-regen detector:
Fig.3 RCA 5794 (left), the most-used vacuum triode in solo-oscillators for
23.cm band. The western winds brought the meteo-sondes into Western Bohemia
since 1950s. On the right, a similar YD1100 triode, suitable for up to 7 Ghz
(Valvo, Siemens, 1970s):
Fig.4 Mirek's OK1VMK transceiver for 23-cm band, 1959:
Fig.5 Transceivers by OK1YN and OK1VMK for 23 and 13 cm bands of 1960:
Fig.6 Field Day 1962 at OK2KEZ , Vysoká Hole, Jeseniky Mountains, 23 cm. Milan
OK2BFF stands at the antenna. Nowadays he has probably the most modern systém
on 2320 Mhz, 10, 24 and 47 Ghz in the Czech Republic:
Operation with Solo-Oscillators
If the signal was strong enough, communication was possible by AM voice
modulation. Also the FM mode was possible as the super-regenerative detector
can handle it, too. Back then nobody was concerned if his output was AM or
FM... For weaker signals, modulated telegraphy was used (ICW). I remember I
have made my QSO with OK2KEA by whistling into my microphone (Fig.8). A signal
without modulation is heard as „silence“. Otherwise it is with superhet
receivers and quartz-controlled transmitters. Some of us were lucky to have
used this new technology. Instead of 100-200 km, the range was extended to
several hundreds km by using the classical CW. My first QSO with a fully
solid-state equipment (described later) was with Karel, OK1BMW (Fig.9), and
OK1KTL. Note: In the 1950s, Rafena Radeberg in East Germany has manufactured
RF generators for higher bands. They were enclosed in heavy metal boxes, and
rectangular dials allowed to set exactly the frequency as well as output power,
from zero to units of Watts. There were several versions, one for 70 cm,
another for 23 cm. An AM modulator can be connected to front-panel terminals
(KZ50), and another pair of terminals for a telegraph key for CW operation. A
nice equipment!
Varactor- a Breakthrough in Design
In the early 1970s, after a successul introduction at 432 Mhz (Figs.11, 13), a
new element entered the 23-cm field which for a long future changed the game
for most microwave frequency bands. First it was adopted in professional
designs, soon also was available for amateurs. It was a variable-capacitance
diode, named „varactor“. Mirek, OK2AQ, introduced it as follows: Following the
discovery of a non-linear dependence of P/N junction capacitance upon the
backward voltage, special diodes were developed, named varicaps or varactors.
Their use was to tune resonants circuits in receivers. Another application was
in circuits with time-varying parameters. In the era when the receiver
front-ends used diode mixers, varactors allowed to design parametric
amplifiers. Their pump oscillator operated at a very high frequency, drove
varactor capacitance, and the resulting amplifier or converter amplified
signals without adding much noise. Varactor variable capacitance found also an
application in frequency multipliers (including higher-power multipliers),
with good conversion efficiency. Also up or down converters with good
parameters can be built with varactors. In radio-amateur world, varactors
allowed to build very simple multipliers. Another advantage was that VHF/UHF
frequency bands were assigned to radio amateurs as integral multiples. New
designs followed soon, and allowed to generate good power levels for very good
AM, FM and CW communication.
Fig.7 In the front, a transceiver by Mirek Klusák, OK1VMK, at VHF techniques
exposition, Prague, 1959 (photo by OK1PM):
Fig.8. Solo-Oscillators were favored also by OK1AIY. The first QSO with OK2EA
at 23 cm band. Field Day 1967 at Černá Kupa Hill, no tripod was available. On
the right, antenna is held by Láďa, OL5AHS, now OK1AUB:
Figs.9, 10 Today a historical equipment for 23 cm by OK1BMW. Transmitter as a
power tripler with LD12 , and a receiver-converter, 1297 to 27 Mhz. Included
was an 8W exciter with QQE03/12, starting from 24 Mhz out of a VFO or a VXO.
Antenna switching by translation of a cable coax connector. The
receiver-converter starts from a 23.5 Mhz quartz oscillator, doubler and
tripler with E88CC, and a symmetrical tripler with 6CC31 to 423 Mhz.The
following tripler used BA110 varicap in a double half-wave resonator, and the
third resonator is coupled to a silicon mixer diode. Originally there was a
1N21 diode, replaced with Tesla 35NQ52, later a Schottky diode HP2800, a
better was HP2350. The IF preamplifier used E88CC. The complete „combine
harvester“ was used for years with a 1m diameter parabolic dish, later Karel
used long Yagi antennas. His record QSO was made on September 3, 1967, with
OK3CDB on Velká Javorina, Slovakia (290 km):
Fig.11 Varactor multiplier, 2 m to 70 cm, a „consumer version“, 1960s:
Figs 12, 13 Varactor multiplier, 2 m to 70 cm, with 1N4387 (Motorola), built
by OK1AIY, 1972:
Beginning With Varicaps
At the beginning, varicaps developed for TV tuner circuits were used which
were encapsulated in glass. They were used from VHF through 4.-5th TV bands.
The first types were BA110, BA149 and BA121, there were other suitable types,
too. Multipliers had to be built and optimized. One of the best was BAY70 from
„psi“ company, Pacific Semiconductor Industry. We also tested components from
Czechoslovak Tesla industry: the best was KA204 manufactured in Tesla
Piešťany. The easy availability stimulated experiments to optimize multiplier
operation and thermal dissipation. Even though we could not achieve the
performance of „professional“ varactors, outputs of up to hundreds of
milliwatts was a respectable level for small portable equipment (Figs.14,15).
Chip availability and and even the support of Tesla Piešťany director allowed
to solder chips to the base substrate and optimize heat dissipation. I have
worked in Tesla Vrchlabí where e.g. KT507 thyristors were being developed, and
those had a chip size similar to KA204. Some effort and support was needed,
also to overcome personal resistance. Even though several tens of modified
devices were made, see Figs. 16 and 17. The cooperative and Smaragd Radio Club
(OK1KNH) was using those devices in frequency multipliers to 70 and 23 cm for
the wide amateur community. As the designs were simple and easy to made,
varactor multipliers covered the gap for years before expensive and
unavailable RF power transistors became domesticated. The positive approach of
Smaragd Radio Club was responded negatively. Antagonisms ceased, fortunately,
over time.
Figs. 14, 15 With this transmitter we achieved many QSOs in 1970s at 70 and 23
cm bands. Its volume was as small as as shoe box even with two sets of
batteries. A very successful design:
Figs.16, 17 A varactor multiplier from 70 to 23 cm with Ka204S varicap (laid
before):
Professional Multipliers With Varactors, 1960s to 1970s
The data transmission requirements were gradually more demanding, so higher
frequency bands had to be adopted. Lower transmitter power needed to be used,
and the interference was rare. Then common transistors were efficient as
amplifiers up to 300-500 Mhz, above that varactor multipliers were used.
Cascaded stages, up to four, were needed for multi-channel telephone and TV
radio links.
Some Notes to Varactor Multiplier Design
As mentioned before, designs are essentially quite simple. In addition to a
varactor and a DC bias resistor, 50-100 kOhms, tuned circuits are need for
varactor matching for the best efficiency and the highest output power. If the
multiplier generates the third or a higher harmonic, another idler circuit(s)
are needed at the second or the third harmonic. Professional varactors were
manufactured with various electric parameters and cases designed for a
particular frequency band and input power. (Figs.21, 23). One of power
varactors suitable to multiply from 2 m to 70 cm was BAY 96. Valvo catalog
specified an allowed input power of 40 W, and efficiency of up to 64%. Ondrej
OK3AU connected the multiplier to the output of his REE30B power amplifier,
and without any overload concerns operated through several satellite
transponders. A similar multiplier is shown in Figs. 12 and 14 in PE-AR/15,
p.40. A varactor is a versatile device that can multiply the frequency as well
as mix a SSB signal from 2m or 70 cm band. This feature was used in microwave
designs of the „first generation“ and will be described later.
Fig.18 A set of varactor multipliers to generate the local oscillator signal
in 24 Ghz band (section 445 Mhz to 2671 Mhz), a): mechanical design, b):
schematic diagram:
Fig.20 Comparing varactor multipliers NEC:
Fig.21 Detail of TESLA varactor multipliers for a TV radio link at 8 Ghz.
Heavy body design is important for mechanical and thermal stability:
Fig.22 Another view of TESLA varactor multipliers: in the front, two stages
connected by a semirigid coaxial cable. Output is a 8 Ghz waveguide:
Fig.23 A NEC varactor multiplier for 5.6 to 5.8 Ghz band. The NEC radio link
replaced in 1979 the old RVG958 systems (now an updated system might have been
installed):
Fig.24 Amateur varactor multiplier, design by OK1AIY, from 445 to 2671 Mhz,
for a 24 Ghz transverter , 1987:
Generating a SSB signal at 23 cm band (1.3 Ghz)
As the SSB signal cannot be simply multiplied, we used mixers and vacuum tubes
on the top of available selection (Figs. 25, 26). Some tubes were designed for
such demanding application. Their design corresponded to the required small
internal capacitances and inductances. Amplifier triodes utilized
grounded-grid operation. One available tube was PC88, manufactured in the late
GDR. Other suitable types were EC88, EC8010, or long-life E88C were
manufactured by Western companies. Such triodes were used in TV receiver front
ends, and were better than PC86. One of the first good designs with this tube
was published in DUBUS (as a blue copy) by Claus, DL7QY, of Berlin.. Later
also by DC8NR in UKW-Berichte. A partial schematic shows the mixer section.
The first tube amplifies LO signal at 1152 Mhz, the second mixes this LO
signal with the SSB signal at 144 Mhz. The feed-through capacitor is only 18
pF, so it grounds the cold end of LO coupling circuit. For 144 Mhz this
capacitor forms a part of the pi-network with the 5-turn coil. In the plate
circuit we have a sum and difference of the above frequencies plus LO. In the
described case, L4 is tuned to the required 1296 Mhz. Following stages amplify
the signal to the desired higher power. Tubes shown in Fig. 25 could be used
but none was then available. In 1970s, the German Radio Club, DARC, awarded
successful stations for taking part in BBT, Bayerischer Berg Tag (Bavarian
hilltop Field Day), with valuable prizes. Through the Czechoslovak Central
Radio Club the designers were awarded some of so needed components. One such
prize was a 3CX100A5 UHF power triode for high-power applications. Another
prize was 2C39 BA which was popular for UHF amateurs worldwide. After 1980s
when the old radio links RVG958 were replaced, we had enough HT323 UHF radio
tubes. They were a close equivalent of 2C39BA. A 10 W amplifier using this
tube is shown in Figs 30-32.
Fig. 25 Power tubes for 23 cm band:
Fig.26 Tubes used for QRP at 23 cm:
Fig.27 A schematic diagram of a 23 cm mixer with a PC88 triode:
Figs. 28, 29 Mixers and amplifiers for 23 cm, top and bottom views:
Fig. 30 A 10 W power amplifier after 30 years in a nicotine den of OK1KZN:
Fig.31 Top view of the amplifier:
Fig.32 Bottom view of the amplifier:
Microwave Design Work
Describing the SSB designs for 23 cm, it was humorously stated that they were
made by „bare hands“. Experts did not like it or they refused to accept it.
But yes, they were right: no more test instruments were needed than an AVO
meter, a GDO and a diode power indicator. For 23 and 13 cm work there was a
good wave meter, by RAFENA, RVG 935, for 0.86 to 3.1 Ghz. Five units were
available in a service set. Its frequency range was the highest, no comparable
device existed. RVG microwave radios for multi-channel telephone transmission
required a permanent assistance, often there were radio amateurs in the crews.
Thanks to Aleš, OK1AGC, we could loan such wave meter from the radio link from
Trutnov to Černá Hora. For years we were lucky to use it in our work. The lack
of test equipment was not and is not so substantial. More important, a
designer had to decide what is important and what is rather „cosmetic“ in his
design. Mechanical and electrical design is more demanding at microwaves, so a
rational approach was needed, to do only important things. Separate circuits
were installed in separate „boxes“, or modules. Some dislike it and are right.
Today a modern Chinese microwave link has a small box joined with antenna feed
in parabolic dish focus. Everything is installed in this box, and DC power and
data are going along the cable. In 1970s such approach as well as systém
manufacturers did not even exist. All this shows the progress and a practical
demonstration for the readers how the technology is changing. Now after
several decades we can demonstrate how radio amateur work followed the
technical development, components and systems, too. Designing modules in boxes
was a practical approach as any improvement or a change or repair only
required to swap connectors and DC power, and to test another module in the
setup. A similar modular design was back then also adopted by famous
manufaturers. Surplus modules from the old times are now available as a
valuable source of reliable parts and are being used around the world.
In the 1980s and 1990s, the possibility to communicate by SSB at microwave
bands like 23 cm has opened really „new horizons“. Long-distance communication
became possible as easy as on 2-m band, and amateur microwave contests became
quite popular. As taking the power from the mains was not often available,
battery-powered transvertors had to be developed with transistors.
Fig.33 Block diagram of a 1296 Mhz transvertor:
Legend to Figure:
Zesilovač - Amplifier, Směšovač – Mixer, Násobič – Multiplier, Oscilátor – Oscillator, Vstup k 2 m přijímači – Input to 2m receiver, Výstup z 2 m budiče- Output from 2 m exciter
Fig.34 Schematic diagram of a 144/1296 Mhz transvertor:
Fig.35 Front panel, transvertor for 144/432/1296 Mhz:
Fig.36 Interior view of the transvertor:
Solid-State Transvertor Design for 23 -cm Band
The goal was to design a transvertor that can operate outdoors on a
rechargeable battery. Also important are a small size and mass, to meet the
specifications of the Bavarian Hilltop Contest. The frequency plan utilized
intentionally 288 Mhz frequency that also suited the 70-cm transvertor. For
simplicity the 70-cm transvertor is not shown in Figs. 33 and 34, only its
schematic is shown in Fig.37. For the frequencies up to 500 Mhz there already
were available suitable low-cost transistors offering up to 3 Watts. On the
transvertor, Fig.35, to switch the frequency band it was required to swap two
connectors from the 2-meter transceiver and flip one switch. Antenna
connectors are located on the side, not visible in the figures. To indicate
output power, there was a small reflectometer installed in each antenna
output. A section of the schematic diagram, Fig.38, shows the mixer in which
the SSB signal is generated at 1296 Mhz. Transistor base receives the LO
signal, 1152 Mhz, from the preceding amplifier. The SSB signal at 2 meters is
fed to the base. Emitter blocking capacitor is small and acting upon both LO
and 1296 Mhz. It does not affect 144 Mhz and this mixer outputs a good SSB
signal. The collector circuit is tuned to 1296 Mhz, and following stages
amplify the desired signal. Suitable types of transistors may look funny today
but as the BFR series was introduced later than when the design was attempted,
it was found that BF357 and BF378 were suitable after all. Although designed
for IF amplifiers of TV receivers, they offered a bit of gain at 1296 Mhz.
Five stages were needed to get the desired output but the amplifier was stable,
no oscillations. On antenna relay (QN59925) output connector we even observed
a glowing filament of the popular 6V, 0.05 A lamp which caused an excitement.
Vlaďa OK1FBQ also loved it, so he made another unit with comparable results.
(He worked in TESLA Votice company where there were professional experts for
RF technology). In the evaluation of next-year Field Day there was a notice
that already two stations ran a successful QSO by SSB at 23 cm...The other
station was OK1KJB. Output power of 0.1 W was nothing of much pride but it was
good for the Bavarian Hilltop Contest. A 10 Watt amplifier was later added for
„fireplace operations“. (Figs.30-33).
In 1978, CQ-DL magazine introduced another transvertor version with a
symmetrical mixer (DF8QK an DC0DA of SSB Electronics). The design was
„printed“ on a double-clad PC board to accelerate the installation.
Low-capacitance trimmers (e.g. SKY, quite costly) were not available, so I
used our glass tubular trimmers. Cost limitations did not allow to make a
2-Watt final stage (3x BFQ34, Figs.39,40). It was finished later but not used
in the transceiver. Instead it has served till today in OK0EA beacon on Černá
Hora. As the time passed by, „nicer“ components were available, DD9DU had
described in CQ-DL magazine (1986) a new generation transverter with GaAs FETs,
with one or two gates. Two DPS allowed a small and technically modern design.
More data will be presented in future descriptions.
Fig.37 Block diagram of a 2m/70cm/23cm transvertor:
Legend to Figure:
Vstup - Input, Zesilovač - Amplifier, Směšovač – Mixer, Násobič – Multiplier, Oscilátor – Oscillator, Buzení - Exciting
Fig.38 A section of the schematic of 1296 Mhz transvertor:
Fig.39 Design of the final 2W amplifier for 23 cm. The massive dissipator on
bottom is not visible. SSB Electronic company has offered this amplifier as a
kit in 1978:
Fig.40 Final power amplifier, 2W, for 23 cm :
Fig.41 Cover page of the Proceedings of Radio Amateur Meeting in Chrudim,
September 1975, where the above transverter details were published :
This paper was also published in print in Practical Eectronics magazine, with
permissione (PE/AR Magazine – Practical Electronic and Amateur Radio, Czech amateur magazine, in Czech).
Related links:
Česká verze