Single-Side-Band Techniques at Microwave Bands, Chapter 2
@ OK1AIY
____________
2304 (2320) MHz band
In 1970s, 13 cm band was the highest contest band. Stations capable of this
frequency operation were few and the number grew only slowly. Mostly vacuum
tubes were used, LD11 and LD12 could generate the needed power. Quartz
oscillators brought a good stability while the input mixer used a single
diode. As there was some experience from 23 cm band, the goal was also to
design a similar solid-state transceiver that could run on a battery. The goal
was tough, though. The BFR series transistors appeared only later and those
available could not do much a 13 cm. Signal levels were quite low, difficult
to detect. Moreover, the available wave meter (PE-AR 8/15, p.37) had a faulty
diode, so the low-level signals could not be detected. The faulty OA601 was
replaced with a GA301 from TESLA. A similar fault happened in another wave
meter. The authors are not inclined to publish negative results, so a lot of
time passed by. Designers had no good and reliable methods, so it took a lot
of time and work to get an acceptable result. Also, no experts for such
microwaves were available. One 13-cm band design come out later, see Figs. 42,
43. Varactor multipliers were used, so only AM, FM and CW was possible. During
Field Day 1976 there were 3 contest QSOs, with OK1KIR, OK1KTL and DL7QY. The
next Bavarian Hilltop Contest was complicated by a faulty hilltop stand, so
the contest was run from the ground, Figs. 44 and 45. Grown trees blocked the
line-of-sight, but two QSOs to Alps, with DL2AS/P and DJ8VY/P were done by FM
and with a good report. DJ8VY was very excited and a following cooperation
with him was very fruitful. Each following contest brought 2-3 QSOs, OK1KKL of
Kozákov was included, also DK0NA and DL7QY. The transvertor has served us well,
it was described in detail in AR 1-2/1977. More materials and components were
becoming available, and DJ8VY even offered hid new design for the use of SSB
mode.
Fig.43a,b Transvertor for 13 cm, with transistors and varactors:
Fig.44,45 Two views on OK1AIB and OK1AIY installations, Boubín hilltop, BBT,
1976:
Design of the Solid-State Transvertor for 13-cm band (2.3 Ghz), SSB
Like in 23 cm band, the goal was to design a transvertor with comparable
features in 13 cm band. The mechanical design was more demanding, resonant
circuits were made of a silver-plated pipe, 6 mm diameter, half-wave long. The
base plate and sides were soldered of one and double-clad laminate boards,
Fig.46. The design was described in AR 4/1981, the schematic is shown also
here, Fig.47. We can see that the SSB signal is generated in a mixer in a
different way than described earlier. The level-controlled SSB signal
generated in 2-m band is injected, together with the DC bias to mixer
transistor base. Via the pi-network, the LO signal at 2160 Mhz is introduced
to the same point. Both signals do not interfere in this structure. One can
see that transistor emitters are directly connected to ground even that some
resistors could be used. Back then we had problems with good RF blocking as
good capacitors were not available. Today, even amplifiers up to 15 Ghz, with
GaAs transistors, have the „emitter resistors“ perfectly blocked, see Fig.49.
Then no negative DC bias is needed. The BFR96 were then new and offered the
highest power. Their cost was high, we bought them at meetings for 160 Kčs
apiece, around 5 USD. The outpur power of 100-150 mW had to be fine, later
transistor or vacuum-tube amplifiers were added. The HT323 tubes could be used,
the difficult part were suitable coaxial resonators, difficult to
manufacture. Machining required a good lathe and a trained operator, not
accessible for a general public. Mechanically similar were EC560, 2C40 or
Russian 6C5D tubes. Suitable parts like bronze contact fingers were used to
design the first mixer and amplifier, see Fig.48. Around that time, QST
published a description of an aplifier and mixer for 13-cm band. Better
features were so attractive that I jumped in the new design, too. To operate
the systém, higher LO power was needed at 2160 MHZ. I used a coupler from 540
Mhz in the circuit shown in Fig.42, 43 ( PE-AR 10/2015, p.40), and a
multiplier using PC88 and HT323 was used. The described module could be
extracted from the panel unit and used outdoors powered by a battery. A new
unit was then built as shown in Fig.46. Both versions were developed at the
same time.
Fig.46 Top and bottom views of a 13-cm transvertor:
Fig.47 Schematic diagram of a transmitter section for 13 cm, 1980:
Fig.48 Amplifier using HT323 tube:
Fig.49 Amplifier module on my hand:
Amplifiers for 13 cm band, 2304 and 2320 Mhz
The mechanical design of vacuum-tube amplifier with 2C39BA (HT323, Fig.50) was
quite demanding.The output power was quite respectful compared with earlier
versions, but we had no means to measure it. The joy was always broken by some
annoying problem. Cathode and plate resonators had to be tuned in operation
and due to heating and cooling this tuning was difficult. Mainly during EME
communication, one of the operators had to continually keep peaking the power
output (as explained by SP6GWN during their first EME QSOs). Professional
manufacturers certainly used a low-dilatation materials. We have found the
closest material in a discarded oil capacitor. Much later, using power
transistors in amplifies overcame the problem. Power amplifier was built in a
panel rack with all DC-power and monitoring circuits including a high-power
cooling fan (named MEZaxial by the maker). See Fig.51. Also Ludvík OK2SLB
started building his 13-cm rig. The final stage of operations was done in a
collective station OK1KQQ, and the home including a kitchen and living room
was quite crowded. The operation ran well, and by midnight we could measure
some 20 W output. The open radiator caused many of us „rabbit red eyes“ in the
morning, but a bottle of rum was blamed as we drank a liter of it during the
group effort. We regularly had more QSOs with OK1KQQ from Lysa Hora.
0.8 W Transistor Amplifier
In 1981, TRW introduced a microwave A-class transistor capable of 0.8 up to 3
Ghz. Under 20 VDC and 120 mA, only 0.13 W was needed for input at 2 Ghz. The
data sheet showed also a schematic of DC bias circuit (Fig.52) to adjust 120
mA current, and a PC board pattern just for 13 cm band. Thanks to Michael
DJ8VY (Fig.53) I received one PC board. The result is shown in Figs. 54, 55.
The box is connected to rear connectors of the transvertor (Fig.20, AR 10/2014,
p.38). Antenna connector, T/R relay and a one-stage RX preamp can fit inside,
too. We intended also to build a 12 t o 24 V DC converter but back then the
switchers with a choke did not exist, so only 13.5 V DC was used. And it
stayed so...
Fig.50 Amplifier with HT323 tube:
Fig.51a,b SSB transceiver by OK1 AIY for 13 cm, in a panel unit:
Fig.52 Schematic diagram of a DC power supply for TRW 53001 transistor as a
linear amplifier for 13-cm band :
Fig.53a,b Michael, DJ8VY, during a winter Bavarian Hilltop Contest, 1992, at
JN67GS locator, 1675 meters above sea level. He is now the leading BBT
contests:
Figs. 54, 55 Side views on the RX/TX amplifier for 13-cm band:
Other components for 13-cm band (2.3 Ghz)
For a successful outdoor operation, one needs a rotary directional antenna fed
by a good cable. Over years all that was gradually developed given the
material possibilities here and in the world. Better components and materials
became available, and new designs appeared. War surplus still survived and is
still used. Antennas have developed the fastest.
First Experience with Parabolic Antennas
I have tested a parabolic antenna for a first time on 23-cm band by the end of
1960s. Its diameter was one-half meter, and I misadjusted the focal point. No
experience was available among hams. Later it was announced that in Svazarm
location of Chlumec nad Cidlinou we can see some parabolas removed from a
radio link tower. They were huge, more than 4 meters in diameter, and so heavy
we could not move them. (Today interested people would know what to do). I
could not bring one home on my motocycle... But there was another one, laying
in a corner. Three meters, made of duraluminium, so this attracted my
attention. The transport was managed by Eman, OK1 AIF. Local agricultural
cooperative transported hay from my location back down, so I found „my
parabola thrown down in a snowed road side. Years later I learned how it
happened. As the dish was oversized, before the truck my friend Jarda, OK1 HL,
had to ride on a motorcycle and waved a red flag to indicate the oversize
load. Unbelievable what could have been arranged with a „zero“ expense back
then. I painted the dish with a special aluminum varnish, made a dipole
radiator and mounted it in the focus. Supported by a hotel Panorama wall in
Benecko (Figs. 56,58) it pointed to the West from where I expected some
signals to come. Pointing by +/- 30 degrees I trained to achieve without any
mechanism. I succeeded to make several QSOs with it, but chances were rare. In
1972 Germany celebrated 50 years of public radio broadcasting. A special radio
station DF0SAR operated from Saarbrucken at 70 cm band , and I could receive
this stable signal. Later they switched to 23 cm, I heard it too. I was able
to point my antenna quite well, and after I called them, they responded. There
was a lot of excitement, we continued corresponding, and I could improve my
equipment.
My first QSO on 23 cm with HB9RG happened on 10-23-1976, and thanks to
training it was perfect. I could have used only a four Yagis on a pole but
with the dish it was more impressive. Another event happened on 13 cm. The
band was always empty, but I still listened . One day I heard a strong noise.
My first idea was there may be a problem in my receiver, but I observed that
noise peaked where there was the Sun behind the clouds. I wrote a report to
Ondřejov Astronomy Observatory (Academy of Science Institution). Their first
response was that the cause might be some interference, but later Dr.Křivský
admitted a solar disturbance and also their radio telescopes were overwhelmed.
He wondered how somebody was able to record that disturbance, and we became
friends. I was invited to various astronomy activities , to Úpice Observatory
and more, but somehow we failed to meet in person. I, however, met his son
when he worked on his thesis. He received my schematics and some practical
advice. His thesis was about detecting solar noise at 3 cm wave length.
Another „happening“ occurred when there were suddenly propagation openings
toward the North. Over 500 m high mountains we could hear stations from
Sweden. After a QSO on 70 cm, SM7CFE wanted to try also 23 cm. My dish had no
suitable support in that direction, so I used my new Škoda car as a support.
The result was quite comic- the dish pointed upwards into a grown forest, Fig.
58, but we did make a QSO! The following day I had to stop in a drug store, to
buy a spray paint and cover the wounds on car body. The QSL for the QSO of
9-8-1981 is shown in Fig. 59.
Fig.56 When the weather was nice, my family took part in my microwave tests:
Figs. 57, 58 Two winter views of OK1AIY dish, now with a log-periodic feeder:
Fig.59 QSL for the QSO with a Swedish station on 1296 Mhz:
As the time flew on, situation was changing. The nice hotel Panorama director
in Benecko has retired, and new alternating owners were not so welcoming to my
activities. Over the most „productive“ opening a new forest trees grew, and
after 25 years it was time to end my „mission“. I sold my dish for 1 Kč (5
cents). It has waited for another use for 15 years. New mesh should be
installed, with smaller eyes (originally 18 x 18 mm). The original label is
shown in Fig. 60.
During 1970s, most-used antennas were the Yagis. For 70, 23 and 13 cm I used
single up to quadruple Yagis, Figs. 66,67. Frost is an enemy to Yagis: glaze
ice some millimeters thick antenna stops working, sometimes completely
collapses. Removing ice by hands helps. Such icing is shown in Fig. 65, made
on Sněžka by Franta, OK1CA, 1986. A ladder was laying on the roof (not visible)
and it is a proof that operating from there was a tough job, with results
hard earned. Fig.64 is very rare: Frozen Loop Yagi for 23 cm is held by Jindra
OK1VR visiting Sněžka during UHF/SHF Contest in 1983, just 25 years after he
made a record QSO on 144 Mhz with England.
The followers were bringing to Sněžka summit more updated equipment. They were
making wonderful long-range QSOs, many firsts, many not-repeatable ones.
Fig.62 shows František, OK1AIB (OK1CA) at the UHF/SHF Contest in 1983. He was
enchanted by Sněžka as we can see in Fig.63, made in 1951 before the Polish
meteo station. He was also enchanted by microwaves as we can see in Fig.61 of
the Field Day 1963 with OK3KAS. Life on Sněžka is documented in the pages
„Sněžka and radio amateurs“ written by Petr OK1AXH. There he also included a
story of establishment and decay of Česká Bouda , Czech Hut.
Fig.60 Manufacturer's label of my parabolic dish:
Fig.61 Franta OK1AIB as a soldier in service during the Field Day 1963 with
OK3KAS, summit of Mikulčin vrch (archive of OK1EM):
Fig. 62 OK1AIB 20 years later on Sněžka (1983). From left, TRX + PA 23 m,
(HT323) + FT 480 , PA 70 cm (2x HT323) + FT-780 , FT-225 + PA 2m (QQE 06/40):
Fig.63 Frantík Střihavka, later OK1AIB and OK1CA, liked Sněžka since his
boyhood (photo of 1951). In the background is the old Polish meteo station and
a section of the original Polish Silesian Hut:
Fig.64 Jindra OK1VR on Sněžka, October 1983:
Fig.65 Antennas of OK1CA (ex OK1AIB) for 2m, 70 and 23 cm, lightly frosted,
1986:
Figs 66, 67 Antennas of OK1AIY on Benecko, frosted in 1970s:
Peter OK1AXH has introduced to Sněžka summit also 5.7 and 10 Ghz bands. He
wrote:
At the very start, Pepa OK1UWA has run one contest from an outdoor tripod but
this way was not good enough. I had ordered a sheet metalworker to make a
large and strong window dormer which I transported to Sněžka summit and
installed it to the hut roof on the East looking side, under antennas of the
former transponder OK0A.. Two runs of steps were installed next to it, one
from the loft above the „large hall“, the other from the upper loft to the
dormer so the activities and equipment installation were comfortable ever
since. The dormer window was really big, 140 x 80 cm, so we can pass a 120 cm
parabolic dish through it. We had a „door to microwaves“ there! On the short
strong pole left after OK0A dipoles we mounted a small table and a dish holder,
with a room for transvertors. The pole can also be rotated by hand from the
top loft (above the large hall). On the bottom there was an angle gauge
(compass) so the azimuth pointing could be done to one degree. In strong winds
the antenna can be fixed in a preset angle while elevation angle stayed at
zero. All mounting can be done by one man in even a bad weather. On the table
a transvertor could be laid, for 6 or 3 cm operation. We could run it all from
the roof but the more comfortable access was from the loft and pointing
compass was there, too. Cables were run through the hollow pole, on the bottom
the power supplies were located, and a 2-meter station to arrange QSOs. Out of
contests we used a „big“ two-meter station, or the 23 cm rig in the small room.
Our first lightweight parabolic dish, 120 cm, was blown away into Poland one
day. We tried in vain to find it. Then we made a more massive holder and it
survived also with a larger dish. This dish we kept there all the fall, only
the transverters were connected as desired. The installation and roof
modifications were a bit illegal, so we kept silence around it, but today
nobody cares..
Start of 1980s, new antennas, cables and a preamplifier
Some readers may think we keep too long describing the past but back then the
novelties that were coming significantly modernized the technical features of
microwave-band equipment. G3JVL had introduced the new Loop Yagi antenna, for
23 and later for 13 cm. This was a nice challenge to improve our equipment. I
made quadruple Loop Yagis for both bands. At the same time, a foam-dielectric
coaxial cable was introduced which offered a low loss. As the old designation
was, there was one named VFKV 630 and a thicker one, VFKV 640. One could not
find any in stores but shorter sections were distributed among interested hams
at some meeting as they did not meet crosstalk specification in a laid
cabling. They had to be replaced with a „better“ version with a double sheath.
Our requirements were more modest, so we used the new cables at 70 and 23 cm
bands. For 13 cm another new cable appeared with Trolitul dielectric cups
(named Duha-Rainbow). It had even lower loss (Note of the translator :
Trolitul-cupped cables were used by Germans during WW II but were replaced by
polyethylene-dielectric cables in 1943). With 13 mm diameter it resisted
freezing and was still pliable. To modify the diameter of its inner conductor
for use in impedance transformers, we made a reflectometer. It used a
precision directional coupler from RVG958 test unit. Detectors had to be
developed, too, TESLA diodes GA301 were used.
Compared to earlier shorter Yagis, the Loop Yagis performed better but no
measurements were done. Fig.71 shows Loop Yagi setups for 23 and 13 cm bands
plus one long Yagi (F9FT) for 70 cm band.
Fig. 68 Petr Hrabák , OK1AXH, spent whole winter seasons on Sněžka summit:
Fig.69 Jindra OK1VR (left) with Petr OK1AXH in a good mood during UHF/SHF
Contest in 1983:
Fig.70 A Synoptic map was back then the only useful tool in microwave work. We
only needed the data on height, temperature, pressure and dew point. Forms
were available at any airport... :
Fig.71a,b Antennas and workplace by OK1AIY, 1983:
Starting in 1980s it was often possible to communicate over long distances
under improved conditions. Frequently we contacted stations from the
Netherlands and UK. During one such QSO with a PA0 station on 13 cm by SSB,
the opposite operator wondered why I still had used a BFR34 instead of a
better GaAs FET. I told him that such devices were not available even in
Prague. He asked for my mailing address (prohibited by Communist authorities),
and the next mail envelope brought me the new GaAs FET, Mitsubishi MGF1400. A
practical version of a preamplifier is shown in Fig. 72. An important element
is the inductance created by a 12-mm long wire, 2 mm dia., shorting the input
for lower frequencies ( high-pass filter). The PC board followed the design by
DJ6PI. My Duroid board, however, was only a single-clad to which I riveted the
ground foil. The connectors were products of Glass Huette company, an
East-German watchmaker. Its mechanical structure was perfect and included a
tiny clamp. The SMA-type connectors were not yet available. The amplifier was
inserted between antenna relay and receiver input. The improvement was
significant but no measurements could be made.
The end of 1970s was marked by introduction of a new technology in radio
communications. Old Rafena Radeberg equipment, RVG958 (often mentioned above)
was being replaced by new radio links by NEC (frequencies from 4 to 8 Ghz),
completely solid-state. Lot of „bulk iron“ was discarded but for us experts it
was a much valued material. Nothing similar was ever available and now we
could reuse or modify many parts. Good test equipment and skills were needed.
One important part was a 0,5 to 1.0 Ghz multiplier, Fig.73. We could use it as
a 23.cm amplifier by inserting an aluminum ring in the plate resonator, and
adding a M6 screw to tune the cathode cavity, close to the tube. Another even
more important part was the amplifier tuned to around 2 Ghz which could be
retuned for 13 cm, Fig.74. In addition to cathode and plate tuners there were
also coupling tuners at input and output. These were welcome to avoid the
thermal frequency pulling as described earlier. A closer coupling caused a
looser resonance and a lesser thermal pulling. There were two amplifier
versions, one with a directly grounded grid, another with the grid grounded
for RF only, with a mica capacitor and the DC with a feedthrough. To this
point we could connect a SSB signal, some 5W, from a 2m transmitter, through a
pi-network , and the LO signal, 2176 Mhz into the cathode cavity. This mixer
generated several Watts output power. Following stages only needed some care
not to kill the tube, only 160 mA was allowed under 800 V plate voltage.As
this version was efficient and reliable, it was mounted in the first version,
seen in Fig.51 in PE-AR 12/2015. The setup of two cascaded amplifiers was
built in another panel unit with power supplies and a cooling fan, Fig. 75. A
transistor exciter was used to generate LO and SSB signals as described in
Figs. 46, 47, PE-AR 11/2015. Two complete equipments existed, the older one
installed in our club station, OK1KZN in Křižlice. The cable to the roof
antenna was 12 m long, a cable from RVG958 had 30 mm diameter, Fig.76.
Fig.72 A 13 cm Preamplifier with MGF1400:
Fig.73 The 0.5 to 1.0 Ghz doubler (used after modification as 23 cm
amplifier). In the front the aluminum ring can be seen to shorten plate
resonator, after Zdeněk, OK1DEF:
Fig.74 13 cm amplifier:
Fig.75 A mobile table with transverters for 13, 9 , and 6 cm bands (3 cm band
not yet installed) Panel unit with amplifiers for 23 and 13 cm is located
below:
Fig.76 Ok1KZN antennas , the quadruple Yagi for 13 cm can be seen on an arm
under 2 m antenna:
The mentioned surplus was best used in OK1KIR club .They had excellent results
already in the 1960s and we had many nice QSOs. They were a good club of
experts and high-art performance. Tonda Ok1DAI with his mechanical workshop
continued later in EME operations, Fig.77. Their 23 cm and 13 cm equipment on
a mobile table with 180-cm dia. Parabolic dish is shown in Fig.78.
Jirka OK1DCI wrote about that time:
We have ran many contests with that equipment from Klínovec and also from
Praděd mountain. In Fig.78 you can see the HV power supply on left bottom for
GI 7B , 23 cm, next the 23 cm transverter with HT323. On the HV power supply
there is a 23-cm PA with GI7B, next to it the mechanical CQ keyer (made by
Tonda), above that the operator's panel with FT-221, and the LNA for 23 and 13
cm. On its bottom rim there are separate T/R switches and reflectometers for
each band. Behind the dish there was a mechanical switch by which the
radiators and both band systems were switched from 23 to 13 cm.
Fig.78 OK1KIR equipment for 23 and 13 cm, 1980:
Next more improvements were made to allow EME operations. Toník OK1DAI, writes
about it:
In the SSB transvertor , mixer and PA with HT323 were used in 1980, made after
QST. Only later we could use the cavities of the equipment used with the 4
-meter dish of 1964. Those amplifiers we used on 13 cm, one for tropo-scatter
tests, and later in the EME „combine“ in Všenory. To get a good power output,
we used 1400 V plate voltage, and each tube passed 200 mA current. This
amplifier required a careful keying, otherwise after a click the sked was
finished. Four combined PAs driven from one such stage we launched in 1984. In
1987 we built an oil cooler. Distilled water would be better but oil worked
under winter freeze. The monster delivered 100 W max. Today, one PC board with
a good transistor can do it, with a good heat sink. The systém was used till
2002 when it was hit by flooding. Gradually, bronze contacts burned in PA
Lecher wire structure, and silver plating was also damaged. So the flooding
finished the 13-cm EME power plant, and we discarded it all. Also discarded
was that 4 m dish originally adjusted for 2.2 Ghz. It worked well even at 5.7
Ghz but not at 10 Ghz. Its overlaid riveted aluminium sheet segments were no
more good.
Fig.77 From the left: Jirka Ok1DCI, Vláďa OK1DAK, Tonda OK1DAI :
As the time ran and ran, there was always something to improve. In the 1990s a
next-generation transvertor was built completely printed. It was difficult,
however, to achieve the performance of the earlier one with silver-plated
tubes. There was a discussion among amateurs about the future of the 13-cm
band. Some insisted that this band be abandoned, by amateurs and professionals,
as there would never be good results. Then the UHF/SHF contest came in 1986,
with nice conditions. There was a chance to make long-range QSOs, with new
countries, some records survive even today. OK1KKW gives more details, Fig.80.
Many participants had to correct their opinion about microwaves. An average
676 km for a 13-cm QSO was impressive, and our first place in IARU region 1
came as a nice surprise. A similar surprise happened twice more but then the
assessors divided the field of participants into two sections, by which it was
impossible to sum up the results of the lower and higher bands. Actually, with
a good equipment, several tens of Watts an a good dish, using ON4KST to
arrange, one can achieve similar results under good conditions, from a good
and tactical QTH. And not only on 13 cm..
Fig.79 and 80. A maps showing QSOs of OK1KIR and OK1AIY during UHF Contest of 1986 (source:
OK2KKW.com):
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