January 2002 43

By Al Klase, N3FRQ

Most of us grew up equatingshort-wave communicationsreceivers with the superhetero-

dyne circuit. Yes, we knew there wereregenerative receivers, but most of the oneswe encountered were clearly toys, and wereeffective only on strong short-wave broad-cast stations. However, an examination ofhistory reveals that there was a 20-yearperiod when the vast majority of short-wavereceivers were regens. During this period,amateurs went from transmitting tens ofmiles to regularly contacting the far endsof the Earth, and they did it with radios thatwere the second cousin of the lowly Knight-Kit “Ocean Hopper.”

Edwin Howard Armstrong, an electri-cal engineering student at New YorkCity’s Columbia University and long-time radio experimenter, invented theregenerative circuit in 1912. The triodevacuum tube, in the form of the De For-est Audion, had been in use as a radio de-tector since 1906, but its operation waspoorly understood. Its performance wasbetter than that of crystal detectors, butits cost and unreliability kept it from dis-placing them from general use.

The original Audion circuit was whatwe would recognize today as a grid-leakdetector. The grid-leak resistor was usu-ally omitted, as the internal leakage ofthese primitive electron devices made itsuperfluous. Armstrong observed thatadding a small bypass capacitor acrossthe headset affected the output signallevel. This led him to believe, contraryto Lee DeForest’s assertions, that RF sig-nals were present in the plate circuit.Being an experienced RF hand, he soughtto peak these signals by adding a tunedcircuit to the plate lead.

The effect was immediate and astound-ing. As Armstrong wrote, “Great amplifi-cation obtained at once!” By adding thetuned circuit to the plate, Armstrong hadintroduced positive feedback between theplate and grid circuits. At low levels, thisfeedback caused the signals to be ampli-fied over and over achieving much greatergain that the original circuit. At higher

The Age of the AutodyneIn the early days of Amateur Radio, the regenerative receiverwas king of the world.

Armstrong’smodification to thebasic Audionreceiver adds anadditional tunedcircuit to the platelead.

levels of feedback, the circuit would breakinto oscillation and generate a radio-frequency signal of its own. This secondmode of operation is important for tworeasons: It marks the invention of thevacuum tube RF oscillator, the basis for anew generation of radio transmitters.Equally important, this oscillating grid-leak circuit is the autodyne detector thatmade continuous-wave (CW) telegraphya practical reality. 1,2

Prior to this time, “radio” almost uni-versally meant spark-gap transmitters andreceivers of limited sensitivity. Spark wasinherently wasteful of spectrum space,emitting what was essentially broadbandnoise, and attaining sufficient transmit-ter power to overcome the shortcomingsof the receivers was a considerable chal-lenge. Experiments had already beenconducted with CW transmitters usinghigh-frequency alternators (rotating ma-chinery) and arc (dc spark) transmitters,but the necessary heterodyne receivers1Notes appear on page 45.

44 January 2002

needed a second alternator or arc to serveas a beat-frequency oscillator (BFO). Inone fell swoop, Armstrong solved thebasic problems at both ends of the trans-mission path.

Armstrong’s regenerative circuit wasprotected by a patent, which constrainedcommercial applications and resulted ina long period of unpleasant litigation.

However, this had no impact on the av-erage amateur. All he had to do was addone more variable inductor, known as avariometer, to his Audion setup.

The Pace of Change AcceleratesAmateur Radio was banned in April

1917 for the duration of World War I.When operations resumed in September

1919, changes came rapidly. Much im-proved vacuum tubes for both transmit-ters and receivers were now available.Hams soon recognized that a few wattsof CW “got out” better than a kilowatt ofspark. Widespread CW operation un-covered serious weaknesses in theArmstrong-inspired receivers then in use.When one wanted to change frequency

A 1930s TRF-Autodyne ReceiverI purchased the homebrew radio shown in Figure A at an

Antique Wireless Association auction some years ago. I wasattracted by its excellent construction, and by the fact that itcame with five sets of plug-in coils. The small size of theband-spread capacitor strongly suggested it was meant forham use.

Not knowing exactly what it was, I took it to GerryMathis, W3GM, who was the local Grand Old Man of hamradio. Gerry lifted the lid and said two words, “rationalizedautodyne.” Beyond that, he didn’t add much that I hadn’talready figured out, but it was an important clue. Later, Idiscovered George Grammer’s “Rationalizing the Autodyne”article in QST. In this landmark article, he summed up thecurrent state of the art in regenerative receivers, and de-tailed the construction of a highly developed receiver thatwas widely built and elaborated upon by others.

My radio was constructed on a salvaged broadcast-receiver chassis (see Figure B), and was carefully mountedin a handsome mahogany case that formerly held an RCARadiola 16. While the physical layout is somewhat differentfrom the set in Grammer’s article, the circuit is almost iden-tical. Fit and finish are excellent for a home-brew rig.

The RF-amp and regenerative detector stages bothutilize type 58 RF pentodes. Band selection is by way ofplug-in coils. The coils for the 20 and 10-meter bands arewound on salvaged tube bases. Each tube, and its associ-ated tuned circuit, is surrounded by a separate metal shieldto prevent oscillation in the RF stage and to limit interactionbetween the controls. Audio-frequency amplification, suffi-cient to drive a headset, is provided by a type 56 triode. Thetubes all have 2.5-V ac heaters.

Tuning is accomplished by the parallel capacitor method.The center tuning knob, below the lighted dial, is the “band-spread” control. It actuates two small capacitors that tunethe RF amp and detector simultaneously. The knobs at theupper right and upper left are the “band-set” controls. The

Figure A—Rationalized Autodyne, power supply andspare coils.

Figure B—Inside the Rationalized Autodyne.

detector band-set is used to select the desired portion of acoil set’s tuning range, and then the RF band set is tweakedfor maximum signal. If you’ve logged band-set positions, thisneed be done just once when one changes coils.

The coils are designed so that a fair amount of band-setcapacitance is always in the circuit when tuned to a hamband. This “high-C” arrangement greatly improves stabilityby making the unavoidable variations in tube capacitance,which occur as supply voltages and circuit settings change,small compared to the total circuit capacity.

Grammer pointed out that many operators used batterypower supplies to avoid problems of voltage variations onthe ac line. However, this particular receiver, while it did nothave a built-in power supply, was clearly intended for acoperation. It even has a switch brought out on the powerconnector to control the power transformer primary. The1934 Radio Amateur’s Handbook shows plans for a sophisti-cated voltage-regulated power supply for just such applica-tions. I constructed one, using vintage parts, that I use withthis radio.

On the air, the “rationalized autodyne” gives a good ac-count of itself. Tuning is precise and repeatable. The detec-tor slides smoothly in and out of oscillation making it easy toset the best operating points for both AM and CW/SSB.Selectivity, at least on 160 through 40 meters, is adequatefor reasonable CW work. Sensitivity is very good, and the RFgain control keeps things nicely under control when thesignals are strong. The set’s stability is impressive, andinteraction between the controls is minimal. It’s easy to copySSB on 20 meters.

The overall performance of this radio, at least for CWuse, is superior to most of the inexpensive non-crystal-filtersuperheterodynes of the era. After using one of these radiosfor a while, it’s easy to understand how the simple home-brew TRF-autodynes held off the superhets for as long asthey did.

January 2002 45

“Improved Reinartz Tuner” from March 1922 QST.

A vintage Reinartz receiver from the JohnDilks, K2TQN, collection.

“Rationalized Autodyne” schematic from January 1933 QST.

when operating in the autodyne mode, allthe adjustments—primary tuning, sec-ondary tuning, plate tuning and filamentcurrent—had to be readjusted to obtainpeak performance.

In a series of QST articles in 1921 and1922, John L. Reinartz, 1QP, introduceda simplified regenerative circuit thatproved to be a much better solution foramateur operation. The Reinartz “tuner”uses a single parallel LC tank in the gridcircuit. A tapped nonresonant tickler coilprovides feedback from the plate circuitthat is throttled by a variable capacitor.A third tapped winding allows variableantenna coupling. The three-winding coilwas an easily reproducible “spider-web”affair. Hams around the world built thou-sands of these receivers, and, fortunately,a few have survived. 3,4

Receiver development continued toadvance as new and more effectivevacuum tubes were introduced through-out the broadcast boom of the ’20s and’30s. For instance, late 1927 saw the in-troduction of the “shield-grid” tube ortetrode. A second grid was added betweenthe control grid and plate of the triode.

This added element acts as a Faraday(electrostatic) shield between the grid andplate. This reduces unwanted feedbackbetween the output and input of an am-plifier, and allows significant RF gain tobe achieved.

Shield-grid RF amps were quicklyadded in front of the traditional triodedetector to improve performance on theemerging 20 and 10-meter bands wherereceiver sensitivity was at a premium. RFamps have some additional advantages.They provide isolation between the an-tenna and detector. This greatly reducesradiation from an oscillating detector, andprevents i ts frequency from being“pulled” by the antenna swaying in thewind. A tuned circuit on the grid of theRF amp can improve selectivity, but greatcare must be taken to prevent the tuned-RF stage from oscillating due to straycoupling between the input and outputtuned circuits. An RF stage can also ex-acerbate detector overload on strong sig-nals by providing too much gain.

Other vacuum-tube innovations in-cluded “heater” type amplifier tubes andhigh-voltage rectifier tubes to allow the

radio to be powered from the ac line ratherthan batteries, low-filament-current typesto allow portable operation from dry bat-teries, and the introduction of the“remote-cutoff” pentode to allow effec-tive RF gain control.

The End of the Autodyne EraIn January 1933, QST assistant tech-

nical editor George Grammer publishedan article called “Rationalizing the Auto-dyne, A Three-Tube Regenerative Re-ceiver of Unusual Performance” in whichhe enumerated the common problems inmost amateur receivers, and detailed theconstruction of a highly effective radio,based on the latest tube types and circuitinnovations, that carefully avoids theseproblems. 5

The year 1932, however, had seen theintroduction of the quartz-crystal IF fil-ter. This circuit allowed superheterodynereceivers to deliver “single-signal” selec-tivity for CW signals. This developmentmade superhets, another EdwinArmstrong invention dating back to 1918,sufficiently better than the autodynes tobreak into mainstream amateur usage. Bythe mid 1930s, many of the leading hamoperators had abandoned their homemadeTRF-Autodynes in favor of commerciallymade receivers from the likes of National,RME, Hammarlund, and, of course,Hallicrafters.

You can contact the author at 22Cherryville Stanton Rd, Flemington, NJ08822; skywaves@bw.webex.net.

Notes1Edwin H. Armstrong, “Some Recent Develop-

ments in the Audion Receiver,” Proceedingsof the IRE, Vol 3, no. 3, September 1915,pp 215-248.

2Edwin H Armstrong, “The Regenerative Cir-cuit,” The Electric Journal, Vol. XVIII, no. 4,April 1921.

3“A Receiving Tuner for C.W.,” QST, June 1921,pp 5-7.

4“The Improved Reinartz Tuner,” QST, March1922, pp 8-10, 26.

5George Grammer, “Rationalizing the Autodyne,”QST, January 1933, pp 11-16, 23.