By Mark Mandelkern, K5AM
Published by the ARRL in NCJ, Jul/Aug, 1998, 12-13. Minor updates; Feb 2008, Aug 2015, Apr 2017.
year project was not all planned out from the beginning! In 1948,
as a typical high-school freshman in Wisconsin with a new ham license,
I was trying to set up as W9ECV on a 25 cent weekly allowance.
Snow-shoveling jobs helped out in winter, and lawn-mowing in summer.
Even so, there was no hope of buying a factory rig. Thus my first
one-tube transmitter and regenerative receiver were homebrew, built
using salvaged parts from large 1930s-style AM broadcast receivers. I
also used some military surplus gear, which was such a bargain at that
time. Converting this gear to ham use provided a good deal of technical
experience, mostly of the trial-and-error variety. But 10 meters was in
excellent condition, and the inexpensive war surplus receivers did not
tune that high. Necessity dictated the need for building a receiving
converter. That's when the homebrew bug first bit me for good.
|The result 67 years later is a
totally homebrew contest station. At
some point necessity ceased to be the driving force, but home-brewing
continued because it was so much fun! Transverters, amplifiers and
numerous gadgets followed for decades. By 1990 practically everything
in the shack was homebrew, and I suddenly noticed that only the
transceiver stood in the way of a 100% homebrew station. That's when a
homebrew transceiver was planned.
The transceiver took three years to design and build; 1990-92. It has dual VFOs, dual-receive, digital read-out, IF shift, RIT, RF speech clipping and filtering, panel-adjustable CW offset, a 200 Hz CW filter, a sensitive integrating squelch for 6 meter DX, a high-performance, no-pop, no-click AGC circuit, a non-crunching noise blanker, and QSK with no dit shortening, no lag, and 50 wpm break-in ability. It uses 142 transistors, 189 integrated circuits, 236 diodes and one 200 watt tube. No microprocessor, no synthesizer, and no phase-lock loops; no phase noise, no spurs, and no birdies.
The K5AM contest station operates at 1500 watts on everything from 160 meter DX to 2 meter moonbounce. The oldest piece of gear in daily use was built while I was still in high-school in 1951: a small 6 meter amp, now used as a driver. My first 3 years of clumsy bread-boarding are sadly lost and not available for on-the-air use; this loss no doubt benefits nearby TV watchers.
It is typical for old-timers that old homebrew gear was not saved. This reflects the fact that the gear was usually built simply to get on the air; home-brewing was done mainly out of necessity because the coveted factory equipment was often not affordable. That old 1951 amp was given away in 1962 when it was replaced by a larger amp. Only 25 years later, when the homebrew and boat-anchor nostalgia set in, was it relocated thanks to the kind efforts of several friends. It had changed hands five times, and was found in a garage under heaps of junk. Now the gear in constant use at K5AM spans almost the full 67 year home-brewing effort.
All the gear was newly designed and built from scratch. But I was not about to reinvent the superheterodyne. Each project begins with a thorough study of the handbooks and relevant magazine articles. I glean ideas from all previous builders, to whom many thanks are due. Design means selecting and choosing the best ideas which will help produce the intended results. Individual circuits are described in the handbooks; the real design work is to combine them into a complete functioning unit. Getting a whole station built in a finite interval of time meant using mostly tried-and-true methods, and setting to work without trying to invent a new circuit for each stage. Only a few circuits are truly new: a self-adjusting, splatter-stopping ALC circuit for tetrodes, an integrating squelch for 6 meter F2 monitoring, a non-crunching noise blanker, and a high-performance AGC system. I've written up these new circuits for the ham magazines, along with a few of the other designs.
This station is used in the most demanding situations: from 160 meter DX with weak signals and high noise levels, to Sweepstakes on 20 meters with heavy QRM, to VHF contesting with rock-crunching local signals, to 2 meter moon-bounce work with infinitesimal signals. Each piece of gear is subject to continual testing, update and improvement. Building simple gear for casual operating and cross-town rag-chews is a fine way to get started in home-brewing and to gain design experience. I soon discovered, however, that achieving state-of-the-art, contest-grade performance is a totally different ball game.
The gear is built on 19 inch black rack panels. The large heavy amplifiers are stacked in the 6 foot rack. The transceiver and smaller items are installed on adjustable pine shelving, to allow easy access and rearrangement. No attempt is made at miniaturization; ample space is allowed in each item for thorough shielding and filtering, and for uncountably many future modifications and improvements. With hinged boards and other devices, provision is made for instant accessibility to allow easy servicing and experimentation. I did receive some essential help in building this gear: one difficult problem was in laying out the panels to achieve a pleasing appearance. Luckily, my XYL Lisa is a professional artist, and most of the panels were designed with her assistance.
There is also another, more rigorous, conception of home-brewing. Some truly amazing hams also design and build their mikes, keys, headphones, CW memory keyers, paddles, voice recorders, towers, antennas, rotators, RF power meters, DSP filters - even test gear and vacuum tubes! To leave ample time for my job, and for working DX and contests, I had to compromise. All those accessory items I buy ready-made. I like building the major items on the rack panels, and find the RF and circuit work intriguing.
Lack of test gear is often thought to be an insurmountable hurdle to home-brewing, but it needn't be. I built and aligned the entire station using inexpensive WWII surplus and Heathkit test gear. Some of the Heath gear I built and some I bought for a song at flea markets. Inexpensive WWII surplus and Heath test gear can still be found at hamfests. Even top-of-the-line, although older, HP test gear is available for reasonable prices at on-line autions; the test gear in my shop is now mostly this sort of bargain HP gear. Formal training is not required, either. I've never taken any courses in electronics, but the handbooks have all the required information. Nor are fancy computer programs needed. Home-brewing for serious DX and contest work is not like advanced development work. All the required formulas are found in the handbooks; they involve no more than the simplest high-school algebra. No one calculates by hand nowadays; the formulas are easily entered into a small programmable calculator, conveniently situated next to the soldering iron. (In earlier days, we used slide rules, and a few wonderful sliding cardboard nomographic rules.)
Does this station come out on top in every contest? No, hardly ever. Devoting so much effort to building the rigs leaves little time for planting a "big gun" antenna farm. I am quite content to be one of the "little pistols" and just have fun. A contest is a merciless test of the equipment; I feel like a winner if none of the homebrew gear breaks down during a contest, and in 67 years it never has. The June VHF contest is my favorite; I hold the section and division top-score records, put up the top nationwide 6 meter score several times, and several times blasted through the eastern monopoly to win a place in the top ten. On HF, my best effort was a division sticker for the ARRL SSB DX contest, way back when it was a 96 hour two-weekend marathon. In the moon-bounce contest, with only a single 2 meter yagi aimed at the horizon for moonrise and moonset, being in the top ten would be an absurd dream, but recently I finally achieved my long-standing goal and got out of the bottom ten.
Ham radio and home-brewing greatly influenced my career choice. At a very early age I built a battery-operated one-tube receiver; it was beautiful, but it didn't receive anything! I looked for information on receivers at the local library, but found the formulas in the electronics books bewildering. That kid sure wanted that receiver to work, so he had to teach himself algebra. Thus ham radio, contesting and home-brewing were crucial steps to college on a Navy scholarship, three years at sea as a deck officer, graduate school, a PhD, and a university research career as a professor in pure mathematics.
Ham radio provided a serious challenge to high-school kids in the 1940s. Young people need serious challenges to capture their enthusiasm and steer them in worthwhile directions. The ham radio challenge in the 1940s was technical; no easy paths were obligingly laid down. The exams were tough and required schematic diagram drawing; entry level was 13 words per minute. A ham ticket required a serious effort, and was considered a highly-valued accomplishment.
Are there more projects planned for the K5AM workbench? A few possibilities, perhaps. But for now, planned improvements to the existing gear will keep me busy for several years. Also, I've recently come down with a bad case of boatanchor fever. My BC-348, ART-13, HRO-5, HQ-129-X, 75A-4, 51J-4, 51S-1, KWM-2, DX-100, Harvey-Wells, Ranger, 6N2, Gonset G-50, and five 6 meter Clegg rigs all need time and TLC.
I've also been spending much time setting up a new VHF contest station at 7900 feet on the north slope of Horse Mountain, in DM54. The heavy, bulky homebrew gear stays home; it would take a week to haul it up the mountain for each contest, and a week to move it back home. Thus the mountain station uses modern, factory-built, store-bought gear. Yes, all the bells and whistles are great fun. Then why am I always so happy to get back home and operate the old-fashioned homebrew gear? It works better!
contest station block diagram. To obtain very high dynamic range
and to minimize spurious responses, the rig uses an overall total of
only two mixing conversions on each band from 1.8 to 144 MHz. This
configuration differs from common practice, so there are some semantic
problems in describing it. The "transceiver" proper consists only of
the IF panel, tuning 40 to 39 MHz, with a 9 MHz second IF. The
front-end section for MF/HF is on a separate panel. The separate
front-end panels for 6 meters and 2 meters, also with 40-39 MHz output,
were formerly called "transverters." Each front-end section employs
high-side local oscillator injection, virtually eliminating images and
To obtain the cleanest signal possible, the station uses class A transistor stages up to 2 watts, and then only tubes - class AB1 tetrodes. The diagram indicates the power available from each unit and also the lesser power required by the next. The resulting headroom yields the best IMD performance. Overall gain is controlled at the milliwatt level in each front-end. To prevent splatter, ALC runs from each driver and each high-power amplifier back to the corresponding front-end panel, with ALC metering at the transceiver. There are no diodes in the signal path at any point in the station.
The 8072 is a conduction-cooled tetrode, identical to the well-known 8122 except that the 8072, with no air-cooling fins, clamps onto a heat sink. The neutralized 8072 driver amp for 2 meters has 26 dB gain.
The Eimac 4-400A bottles used in push-pull on six meters are the originals, 54 years old and still running at full output!
Mark Mandelkern, K5AM, was first licensed in 1948 as W9ECV in Wisconsin. Mark's main ham operating activity is 6 meter DX, with 146 countries worked. He also enjoys HF DXing and contesting, with DXCC confirmed on 10 bands and a number of top-ten contest awards. Mandelkern is professor of mathematics emeritus at New Mexico State University.