Station of the Moment

Transmitter: homemade 6BW6 beam-power-tube crystal oscillator with 2.2 W output at 40 meters and 3.6 W output at 80 meters; receiver: modified Kenwood R-599A; antenna: 100-foot doublet, 25 feet up, fed with ladder line via a homemade balanced tuner; straight key: Mesco.

The First W9VES

The amateur radio call sign W9VES has been assigned to—held by, as ham radio operators like to put it—only two people, of whom I am the second. My maternal uncle, Phil Simmons, was the first W9VES, and I enjoy my amateur radio in celebration of him. By mid-2008, the expanding sphere of Effect created by his first transmission in 1935 spanned a diameter of 146 light-years.

Following a Long-Abandoned Thread

Years ago, while building up the gumption to take the Novice amateur radio license examination as a teenager in the summer of 1969, I bought a little yellow Editors and Engineers paperback book, Easy to Build Ham Radio Projects by Charles Caringella, W6NJV. Its simplest construction project, "Novice Transmitter for 80 or 40 meters," consisted of a grid-plate crystal oscillator that used a 5763 transmitting beam power vacuum tube. By mail, from Lafayette Radio on Long Island, I bought an small aluminum chassis to house it. The rest of its components came from the parts collection—junkbox in ham-radio speak—of my father, who has operated amateur radio station W9BRD since the late 1930s.

Although I completed the transmitter to the point of dimly lighting an incandescent lamp with its radiofrequency (RF) power output, I never went on the air with it. I was persuaded by my ham-radio mentor—my Elmer, by ham-radio tradition—that I would be disappointed with its low power. He was accustomed to using an outdoor, tower-mounted antenna and running as much power as the law allowed, and he knew that I would begin my ham career with my father's quasi-indoor antennas. And so I put the little 5763 transmitter aside, later going on the air for the first time, as WN9CJS, Norridge, Illinois (near Chicago), with a more-powerful Lettine Model 240 transmitter, once owned by the first W9VES. The Lettine used a 6L6GB crystal oscillator driving an 807 power amplifier, and likely put out perhaps six times the RF power of the 5763 oscillator.

Later, per an article in QST (Donald Mix, W1TS, "A Simple Transmitter for the Beginner," Beginner and Novice, QST, September 1968, page 22), I would add a 6C4 triode crystal oscillator to the 5763 circuit and enjoy many solid contacts and several radio adventures with the modified transmitter, including working KH6ALD, Hawaii, on 40 meters during a middle-of-the-night thunderstorm. (I would also discover that the transmitter still acted as a crystal oscillator with the 6C4 stage disabled, a common characteristic of the Boosted Pierce.) And later still I would discover, using transmitters with output powers from 6.3 milliwatts to 20 watts or so, some using transistors but most (including the 6.3-mW job) using tubes, that I could have limitless fun with low transmitting power as long as my antenna and receiver allowed me to hear signals down to the band noise.

Thinking this all over, I decided that I would begin my amateur radio activities as AB2WH with the transmitter that I had forsaken: the 5763 grid-plate crystal oscillator from Easy-to-Build Ham Radio Projects. Having had 38 years of ham radio experience in the meantime, I modified the circuit somewhat. For example, I used voltage-regulator (VR) tubes rather than a series voltage-dropping resistor to supply screen-grid voltage to the 5763, as I knew that this would stabilize my operating frequency by generally reducing, and reducing changes in, RF heating of the crystal.

Amateur Radio Station AB2WH was therefore initially configured as follows. The 5763 grid-plate oscillator served as the transmitter. An ongoingly modified Kenwood R-599A served as the receiver. A relay-and-toggle-switch-based transmit-receive (TR) switch, built into the steel box from a defunct computer power supply, switched the antenna between receiver and transmitter, and also switched the transmitter between dummy antenna and real antenna. The dummy antenna, built into a butter-cookies tin, was a 50-ohm resistance consisting of five 10-ohm, 10-watt Radio Shack power resistors connected in series, with the inductive reactance of this resistance "tuned out" with a squeaky old two-section broadcast-receiver tuning capacitor, its sections paralleled, wired in series with the resistors. The real antenna was an outdoor, low random wire, its electrical length somewhere between a quarter wavelength and a half wavelength at 3.5 MHz. I tuned out its reactance with an L impedance-matching network made from a 280-pF capacitor and a homemade inductor made from #14 black-jacketed THHN house wire wound on a sparklingly transparent polycarbonate drinking glass and tapped at intervals so I could adjust its inductance by short-circuiting part of it with a wire equipped with an alligator clip. The matching indicator was a Monimatch Mark IV (Lewis G. McCoy, W1ICP, "The Monimatch Mark III and Mark IV," Beginner and Novice, QST, September 1964, page 20), built into two International Coffees tins. I keyed the transmitter with the straight-key portion of the Brown Brothers CTL-B combination paddle/straight key I've used since my Seattle days as AK7M.

At 3.5 MHz, the transmitter put out 4 watts, a value based on measurements made with a digital multimeter and a homemade RF peak voltmeter as described in Wes Hayward, Rick Campbell, and Bob Larkin, Experimental Methods in RF Design. On my first-ever call with it, a 3.550-MHz CQ in the wee hours of March 11, 2007, I worked W8TY, Jon, in Elida (near Lima), Ohio.

Flux

I hold an amateur radio license for the purpose of having occasional radio fun by transmitting signals in frequency bands allocated to the Amateur Radio Service. Because I am technically inclined and enjoy amateur radio's history, including its history in my family, as much as amateur radio's present, that occasional radio fun often involves the construction, use, and exploration of relatively simple transmission and reception equipment for radiotelegraph (Morse code) communication. My radio activities and configurations are therefore in ongoing flux and have included:

• In April 2004, electing to celebrate, meditate, and mourn in radio silence with friend Byron Goodman, W1DX, terminally ill; with 35 years gone since the passing of the first W9VES; and with my having attained Phil's age (also, incidentally, his sister's—my mother's—age) at their deaths.

• In January 2007, passing test elements for the Amateur Extra license at an examination session administered by the Columbia University Amateur Radio Club. Having long before passed General, Advanced, and Amateur Extra tests before official Federal Communications Commission examiners, I had never before experienced an exam proctored under the Volunteer Examiner system.

• Assignment of the call sign AB2WH on February 1, 2007.

• Further evaluation of the W6NJV 5763 transmitter circuit using the two-band output inductor I had used in experiments with the Mighty Midget circuit. The oscillator's 3.5-MHz output was now 5.0 watts, likely due to a more optimal output matching. Doubling to 40 meters with the same 80-meter crystal produced 1.5 watts; using a fundamental crystal at 40 meters, 3.0 watts. Keying was good at 80 meters with an 80-meter FT-171B crystal; excellent at 40 meters when doubling from an 80-meter FT-171B crystal; good at 40 meters with a large World-War-era crystal (7.01 MHz); and yoopy with two modern, HC-49-cased 40-meter crystals. I attribute yoop, a relatively slow frequency change that occurs across longer code elements or across multiple code elements—just holding the key down illustrates it best—to crystal heating and consider it to be a sign of excessive feedback and crystal endangerment.

• Evaluation of a 6GW8 Boosted Pierce transmitter (a Pierce crystal oscillator [6GW8 triode] driving a non-neutralized power amplifier [6GW8 pentode]). Using the Mighty Midget output inductor, this transmitter put out 5 W (80 meters, excellent keying) or 4 W (40 meters, good keying).

• Ongoing evolution and enjoyment of the BG-1 Mix-Goodman receiver, a band-imaging superheterodyne receiver that uses a local oscillator in the 5-MHz range to mix the 80- and 40-meter bands to a intermediate frequency (IF) of 1.7 MHz for reception by means of a high-capacitance (high-C) 12SJ7 pentode regenerative detector.

• Replacement of the initial AB2WH antenna (a random wire fed against ground) with a ladder-line-fed 100-foot-long doublet 27 feet above ground

• Review of the ICOM IC-R9500 wideband receiver in the January 2008 issue of ARRL's QST magazine
  

• Participation in the 2007–2008 ARRL Straight Key Night using the highly modified transmitter portion of a Heathkit HW-16 transceiver (restyled as the HW-15 by Bruce Kampe, K1BRK) cathode-keyed by four 6AS7 dual-triode vacuum tubes to simulate the beautiful, bongy, blown-soda-bottle sound of primary keying

• Diagnosis and repair of failure of the OSC OUT (oscillator output) functionality of K1BRK's Oak Hills Research OHR-100A 7-MHz QRP transceiver to allow use of the radio's companion DD-1 digital display

• Construction in an 8-by-9-inch s'mores cookie tin of the "Winter 80," an 11-watt-output 3.5-MHz Morse code transmitter/receiver using (in the receiver) an NE602 direct-conversion detector, op-amp audio filtering and LM386 audio power amplifier and (in the transmitter) a 12AL11 pentode crystal oscillator, 12AL11 beam power tube driver, and 1631 beam power tube (12-V metal 6L6) final amplifier

• Power crystal oscillator experiments with the 17JQ6 beam power tube that found the xJQ6 to be a good candidate for amplifiers, multipliers, and power crystal oscillators in the 5763/6417 class

• Regaining the W9VES call sign as of April 1, 2008, April Fool's Day. Phil would have guffawed his distinctive guffaw.

• In May 2008, revamping the oscillator/Boosted Pierce test chassis into the "Summer 40," an 11-watt-output Morse code transmitter consisting of a crystal oscillator (12HL7 pentode), driver (5659 [special 12A6] beam power tube), and power amplifier (1631 [12-V metal 6L6] beam power tube). By July, results with the Summer 40 and 100-foot-long doublet antenna had included contacts as far west as Seattle and as far east as France and the Madeira Islands.

• In late July 2008, adding bandswitched 40-meter capability and variable output loading to the transmitter section of the Winter 80, making it suitable for year-round use as the Summer 40/Winter 80. I measured its maximum output as 12 W at 80 meters and 11 W at 40 meters with 360 V, key down, on the final plate.

• In late September 2008, modifying the TUNE function of the Summer 40/Winter 80's TUNE/OP switch to adjust the rig's output power to 5 W, the traditional maximum power limit for QRP operation.

• In September 2008, equipping the HW-15 transmitter with an inboard two-2SK3563s power-MOSFET cathode keyer/shaper, and replacing the rig's electrolytic high-voltage filter capacitors after its decades-old three-sections-in-a-can capacitor failed and emitted that characteristic brownish chemical fuzz. (Think of the mess you get when a forgotten Coke explodes in a freezer.)

• At the end of November 2008, adding muting and transmit-monitor level control to the modified Kenwood R-599A receiver. A PNP bipolar-junction-transistor switch in the 9-volt line to the receiver's RF GAIN control does the muting; a 1-MΩ potentiometer wired as a rheostat and connected across the switch serves as the MONITOR LEVEL control.

• In early December 2008, constructing a injection-locked-oscillator frequency-halver based on Byron Goodman, W1JPE, and Hal Bub, W1JTD, "Frequency-Halving Oscillators" (QST, September 1941, pages 46, 47, and 98), to get 3560-kHz output from a 7120-kHz crystal. My version of the circuit used a 5963 (special 12AU7 twin triode) tube, of which one section operated as a Pierce crystal oscillator at 7 MHz, and the other section operated as an LC oscillator at half the frequency of the crystal. Loosely coupling both oscillators together causes the LC oscillator to lock to the crystal at its second harmonic. A two-transistor buffer amplifier after Roy Lewallen, W7EL, added isolation and amplified the output of the LC oscillator to 50 milliwatts in 60 Ω. The result is crystal-class stability at spot frequencies not available from my collection of 80-meter crystals.
        With very few circuit changes (reducing the supply voltage from 108 to 13.5, reducing the injection coupling capacitance from 5 pF to 3.3 pF, and lowering the series resistance at the input to the buffer amplifier from 6.2 kΩ to 4.7 kΩ), the circuit works as well with 2N3819 junction field-effect transistors (JFETs) substituted for the vacuum-tube triodes. Solid-stated thus, its output is 25 mW in 60 Ω—still more than sufficient to drive my three-stage tube transmitters (Summer 40/Winter80 and HW-15).
        For 2009, I plan to use this frequency-halving arrangement to get down (or, as the old old-timers would say it, thinking in terms of wavelength, up) to 160 meters from crystals in the 3.6-MHz range.

• In late 2008 and early 2009, wrote a chapter on circuit computer-aided-design (CAD) for, and updated the mixers and modulators chapter of, the 2010 ARRL Handbook for Radio Amateurs.

• In June and July 2009, built a 6Z7G-6AG7 injection-locked frequency halver/amplifier with plug-in coils for output at 3.5 and 1.8 MHz. Practically useful locking can be obtained from crystals oscillating at 2× to 5× the output frequency. When W9VES operates at 3.56 MHz, that signal originates from this halver, dividing from one of several 7.12-MHz crystals; at 1.82 MHz, from a 3.64-MHz crystal from Digi-Key.

• In September 2009, reprised the W6NJV 5763 oscillator transmitter with its original circuit—series screen voltage-dropping resistor and greater-than-optimal pi-network inductance—including use of a lower-voltage plate suppy (circa 260 V key down) on the order of the one I used with the transmitter's subsequent 6C4-5763 Boosted Pierce incarnation in 1969. Output with a 5763 was 2.2 W at 40 meters and 3.3 W at 80 meters with the output TUNING capacitor adjusted to greater-than-that-necessary-for-maximum-output capacitance to obtain yoopless keying; with a 6BW6 (9-pin 6AQ5 with its separate beam-forming plates grounded), 2.2 W at 40 meters and 3.6 W at 80 meters. Across a weekend I made three contacts at 40 meters (Indiana, Michigan, and Missouri) and one contact at 80 meters (New Jersey). In these brief tests, the 6BW6 was arguably the better tube, giving the same or greater output than the 5763 for yoopless keying. Key-down screen voltage was 180 by means of a 22-kilohm dropping resistor from the plate supply; the grid leak was 47 kilohms; the Colpitts divider capacitances, 22 and 220 pF. The output pi-network inductor was an 8.6-µH solenoid (ex-BC-342 receiver) at 40 meters and a 21.2-µH (44 turns on a T130-2 powdered-iron toroidal core) at 80 meters. A 1-µF capacitor in series with 100 ohms across the key jack suppressed clicks on break.