Tracing the Development of the
AM Broadcast Transmitter

By John F. Schneider W9FGH, 2018

www.theradiohistorian.org
Copyright 2011 - John F. Schneider & Associates, LLC

 [Return to Home Page]

(Click on photos to enlarge)

San Francisco, 1902:  Thirteen-year-old Francis McCarty is shown speaking into his invention, the McCarty Wireless Telephone.  Although he did successfully transmit the human voice, the intelligibility was poor because of his reliance on spark transmitter technology.  Successful voice transmission would have to wait for the development of continuous wave (C.W.) technology a few years in the future.  (California Historical Radio Society photo)

Front and side views of a typical spark transmitter.

A Poulsen arc converter transmitter, seen in this photo, created a continuous wave through the creation of a sustained high-frequency oscillating arc enclosed in a magnetic field.  

This 1 megawatt Poulsen arc converter was built for the US Navy by Federal Telegraph Company in 1919.  It was installed in Croix d'Hins, France..

---

Prof. Reginald Fessenden's alternator, installed at Brant Rock, Massachusetts in 1906.

Ernst Alexanderson at General Electric developed the Alexanderson Alternator, another early transmission system that was capable of transmitting a continuous wave radio signal before the development of power vacuum tubes. These monstrous machines were manufactured between 1910 and the early 1920’s, and several of them remained in operation through the 1940’s.  The last remaining Alexanderson Alternator, at Grimeton, Sweden, is still operated occasionally for special events.  This photo shows a pair of 200 kW Alexanderson alternators at RCA Radio Central, Rocky Point, Long Island.

Most broadcast stations in the early 1920’s assembled their own transmitters.  This photo shows the entire facilities of WRM at the University of Illinois in Champaign.  The transmitter (center) received its DC power from a motor-generator (lower left).  The student announcer (right) is speaking into a converted telephone microphone.  Batteries on the floor supply the power for the amplifier on the desk.  The two tubes in this transmitter were the only ones owned by the university, and they were occasionally borrowed during off-air hours by a professor doing sound-on-film research.  WRM remains in operation today, now using the call sign WILL. 

Here is another home-built transmitter at WFBE in Cincinnati, 1927.  We can see the power supply with its light-switch controls in the left-hand cabinet, and the RF section on the right.  Lethal voltages are exposed, with the warning sign and the operator’s good sense being the only protection.  When a home-built transmitter was ready to be put into operation, the station would call the local government radio inspector, who would come to the station, make some measurements and certify the transmitter. Even so, the mechanical and electrical quality of these early rigs varied greatly.  WFBE operated with 250 watts from the Garfield Place Hotel from 1927 to 1935, when it was purchased by the Cincinnati Post and became WCPO.  

The General Electric Company was always at the vanguard of radio technology, and numerous innovations in the radio art came out of G.E.’s Schenectady laboratories.   This photo shows the experimental breadboard 1,000 watt transmitter that broadcast G.E.’s station WGY in March of 1922.   The power supply was housed in the frame at right, and the RF tubes were mounted in the open frame at left.  The two pancake coils sitting on the table at center appear to have served as the output coupling transformer.  WGY claimed this transmitter was the most powerful in America in 1922, and its signals were heard as far away as Cuba.  Later, more tubes were added, increasing the power first to 1,500 watts, and then finally to 5,000 watts with fourteen RF tubes.

This photo shows the transmitter room of WMAQ in Chicago in December, 1925.  The Western Electric 1,000 watt model 6-B transmitter (center) has just replaced the four-year-old and obsolete 500-watt Western Electric 1-B (left).  These two models were the most popular factory-built transmitters in the 1920’s, and saw use at many of the country’s biggest and best-financed stations.  WMAQ’s transmitter was located on the fourth floor of the LaSalle Hotel in Chicago, and the antenna was suspended between two towers on the hotel’s roof.  

Here is another view of the Western Electric 6-B transmitter.  This one was installed in 1925 at KPO in San Francisco, located in the Hale Bros. Department Store on Market Street.  One detail to notice is the marble electrical panel in the background.

Front and rear views of the Western Electric 5B, the first factory-built 5,000 watt transmitter.

This Western Electric diagram shows the typical  installation of a 5B in a  radio station..

This image shows a 5B transmitter installation at WMAQ in Chicago.

General Electric put the country's first 50 kW transmitter on the air at WGY in Schenectady in 1925.   Here is a view of that first transmitter at GE's experimental radio facility in South Schenectady. 

This is an aerial view of the General Electric experimental radio facility in South Schenectady, NY, 1926.  WGY's first 50 kW experimental transmitter operated from this location.

By the mid 1920’s, the quest for higher power by the country’s clear-channel stations led to the development of several experimental 50 kW transmitters.  This is the RF section of the RT-150A, built by General Electric for WEAF in New York.  The ten water-cooled 10 kW tubes mounted on this open frame panel were operated in parallel to achieve the target power.   Only two RT-150A’s were ever built -- soon afterwards, G.E. developed the 100 kW UV-862 power tube, eliminating the need to parallel many smaller tubes.   With this new tube, RCA, GE and Westinghouse collaborated on a new transmitter to be called the 50-B. 

In the beginning, there was King Spark

 The first transmitters grew out of the observation that, if a continuous chain of sparks was discharged into a tuned antenna circuit, a signal would be radiated that could be received at a distance and could carry telegraphically-coded communication.  Guglielmo Marconi was the first to develop a practical communications system using spark transmitters.  

There were a few early attempts at using spark equipment to transmit the human voice.  Francis McCarty in San Francisco developed a crude system between 1902 and 1906, but the speech quality was poor.  This was because a spark signal consists of a continuous sequence of decaying waves, called “damped waves”.  The signal faded in intensity as the energy of each spark dissipated, until it was replaced by a new signal from the next spark.  These “holes” in the signal prevented the transmission of clear intelligent speech.  What was needed was a “Continuous Wave”, or C.W., signal.  In the early 1900’s, there were only two devices that were capable of generating a continuous wave – an arc transmitter and a high-frequency alternator. 

The arc transmitter was conceived by the Danish inventor Vlademar Poulsen in 1903.  It functioned by generating a continuously-oscillating arc between carbon and copper electrodes inside a magnetic field, which converted the arc’s high voltage DC to a continuous wave RF signal in the VLF frequency range.  A number of high power arc transmitters were built by the Federal Telegraph Company in Palo Alto, California, for the U.S. Navy before and during World War I. 

At about the same time, in nearby San Jose, Charles D. Herrold was able to broadcast intelligent speech from an arc transmitter by inserting a microphone between the transmitter and antenna.  This crude system of modulation operated on the principal that sound waves caused the resistance of a carbon microphone element to vary, producing a corresponding change in antenna current.  The complication was the great amount of heat dissipated in the microphone; Herrold solved this by using an array of six water-cooled mics in parallel.  This method only created a modulation level on the order of ten percent.   Even so, using this crude system Herrold was able to maintain a schedule of weekly music broadcasts to local ham radio operators between 1912 and 17.

The second device able to generate a continuous wave signal was the high frequency alternator, first developed by Ernst Alexanderson of General Electric for Dr. Reginald Fessenden.  This was a completely mechanical system –a high speed motor was used to drive a specially-constructed alternator, producing an A.C. current that oscillated at very low radio frequencies (20 to 100 kHz).  Fessenden is famously said to have used an early version of his alternator to broadcast speech and music to vessels in the Atlantic on Christmas Eve, 1906, utilizing the same microphone absorption method as Herrold.  Over the next several years, G.E. developed alternators up to 200 kW that were used by the Navy, RCA, and other major communications actors for high speed CW communication well into the 1940’s.

The Vacuum Tube Changes Everything!

 The invention of the “Audion” triode vacuum tube by Lee de Forest in 1906 created a revolution in radio communications.  Its ability to function as both an oscillator and amplifier opened doors to the creation of a practical all-electronic speech transmitter.  In 1913, de Forest sold the commercial rights to his tube to AT&T, where the erratic device was further developed into a practical and stable product.  World War I saw additional development, and the vacuum tube was quickly adapted for use in both transmitters and receivers. 

Lee de Forest was also one of several pioneers in early experimental broadcasting, using his vacuum tube transmitter to broadcast recorded music from his station 2XG in the Bronx starting in 1916, and later transmitting live opera music from 6XC in San Francisco in 1920. 

Many amateur radio operators, prohibited from transmitting from 1917 to 1919 due to wartime security measures, entered the armed forces as radio operators, and they considerably enhanced their knowledge of tube electronics in those years. When the wartime moratorium was lifted, dozens of these hams experimented with audio transmission utilizing war surplus tubes.  A number of these hams joined the ranks of the first commercial broadcasters as the radio boom swept the country in 1920-22.  Several of the country’s first broadcasting stations - including WWJ, WHA, KDKA and KJR - grew out of amateur stations.

In 1922, radio broadcasting became an overnight sensation, as millions of average Americans were bit by the radio “bug”.  Between March and June, the number of radio stations swelled from 67 to nearly 400.  Most of these stations operated with home-brewed transmitters of varying power and quality, utilizing a variety of circuit designs.   Available transmitting tubes ranged in power from ten to 250 watts input, and so it was common practice to operate several tubes in parallel to achieve higher powers.  The results were often unstable and unreliable, as most of these early rigs were nothing more than high-power free-running oscillators.  Modulation was accomplished with a high power Class A modulator stage using the Heising “Constant Current” method:  the plate current for both the RF and modulator stages was delivered through an inductor that resisted changes of current with varying modulation, so that modulation peaks would cause a corresponding drop in the PA voltage.  These transmitters were typically capable of modulation peaks of only about 50%.

The first factory-built broadcast transmitter was the Model 1-A, released by the AT&T subsidiary Western Electric in 1921.  Like many of those rudimentary home brew rigs, it was a 500 watt free-running oscillator with Heising modulation.  (500 watts was considered “high power” in 1921.)  Four 250-watt 212-A vacuum tubes provided the carrier power and modulation.  High power rectifier tubes did not yet exist, and so the filament and plate voltages were supplied from DC motor-generators.  The 1-A’s first users were AT&T’s WEAF in New York and WWJ in Detroit.  These stations -- like most early broadcasters – transmitted from “flat top” horizontal wire antennas, which were an outgrowth of the old maritime spark antennas.  Because the antenna capacitance was part of transmitter’s tuned circuit, they would drift off frequency whenever the antennas blew in the wind.  Western Electric resolved this problem by adding an output tuning network, and the resulting transmitter, now called the model 1-B, was soon installed at more than thirty of the country’s most important radio stations.

Nonetheless, the majority of the country’s broadcasters were still using homemade transmitters, and they were legally in violation of patents that AT&T controlled on a number of critical transmitter circuits.  AT&T attempted to enforce its patent rights by demanding these stations pay royalties.  It also alleged that it had the exclusive right to broadcast on-air advertising, and demanded that all other stations cease the broadcasting of advertising messages.  The proposed license agreement was so onerous that most broadcasters refused to sign it.  As a test case, AT&T sued the New York broadcaster WHN, and although it ultimately won its lawsuit, the negative publicity created by these heavy-handed methods finally caused AT&T to drop its patent enforcement efforts.  Nonetheless, its hold on a number of key patents kept other companies out of the transmitter business.  (An exception was made for its RCA patent pool partners Westinghouse and General Electric, but they could only make transmitters for their own stations.)  It wasn’t until the patents expired at the end of the 1920’s that RCA, de Forest and a few other smaller manufacturers could enter the field and supply factory-built transmitters.

When first organized in 1919, RCA was simply a pool of the radio patents controlled by General Electric, Westinghouse, AT&T, and a few smaller players.  Subsequently, most all the development of the more modern commercial transmitter technologies grew out of laboratory research conducted at Western Electric, G.E. and Westinghouse in the 1920’s and early 1930’s.  Each of these companies operated their own broadcasting stations and they used them as research test beds, exchanging innovations among themselves.   Particularly, G.E.’s broadcast station WGY in Schenectady was a key test bed for the development of high-power transmitter tubes and more stable circuits.  Thanks to that company’s work, a second generation of transmitters emerged in the late 1920’s.

The Transmitter Come of Age

In 1931 the Federal Radio Commission issued two new regulations governing broadcast transmitters.  General Order 111 required stations to modulate a minimum of 75%, and General Order 116 required stations to maintain their carrier frequencies within +50 Hz to eliminate heterodyne whistles on the broadcast band.  The old free-running oscillator rigs became obsolete overnight, particularly due to the frequency stability requirement.  Their usual method of frequency control was for the operator to adjust the transmitter’s frequency from a front-panel knob while zero-beating the transmitter’s signal against a reference crystal oscillator, but they would usually quickly drift off frequency again.   The stations that were measured off-frequency were fined, and several station licenses were even revoked.  Almost overnight, the nation’s installed base of broadcast transmitters was replaced with new transmitter designs using crystal-controlled RF oscillators, a technology recently developed by the G.E. labs in Schenectady.  Many small stations couldn’t afford the investment, and they either disappeared or were merged into larger operations.

 About 1928, Western Electric introduced its revolutionary model 6-B one kilowatt transmitter.  It utilized a Master Oscillator-Power Amplifier (MOPA) design, driven by an oven-controlled low power crystal oscillator stage.  Heising modulation was applied at a mid-level stage, followed by a Class “A” final amplifier using a single 228-A water-cooled tube.  It was capable of modulation peaks approaching 100%. Hundreds of these transmitters were installed at broadcast stations around the country, and many of them continued in operation through World War II.  That same year, Western Electric introduced its model 5-C, a 5,000 watt transmitter housed in an imposing row of six cabinets.  This was the first mass-produced factory-made transmitter to produce its PA voltage with rectifier tubes instead of motor-generators.

By the mid 1920’s, several clear-channel stations were experimenting with the 50 kW power level, enjoying nearly-nationwide coverage on their privileged frequencies.  WGY was the first station to achieve this power level experimentally in July, 1925, using the call sign 2XAG.  The transmitter was built by GE's Radio Engineering Department at its experimental facility at South Schenectady.  Three shortwave stations also operated from this location.  In subsequent years, GE conducted further tests from this site at 100 kW (1927), 150 kW and 200 kW (1930).

For its part, RCA contracted with both Westinghouse and G.E., with each designing and building one high-power transmitter for its flagship New York stations, WJZ and WEAF.  The Westinghouse unit went on the air at WJZ in Bound Brook, NJ, in November, 1925.  It was a conventional 50 kW self-power oscillator in an open-frame design utilizing twenty water-cooled tubes.  (See the Spectrum Monitor article, July 2016)  For its part, G.E. delivered its more innovative RT-150A to WEAF at Bellmore, Long Island.  It also combined 20 water cooled tubes in an open frame construction, but the resemblance ended there.  A dedicated high-power crystal-controlled transmitter excited the final amplifier, and the modulation was accomplished at the final RF stage using a high-powered modulator and Heising modulation.  It also used mercury-vapor rectifier tubes instead of motor-generators for its PA power supply.  It was clearly superior to the Westinghouse design, and RCA soon ordered a second RT-150 for WENR in Chicago.  (Spectrum Monitor article, December 2015) 

In 1928, Western Electric joined the high power club with its model 7-A 50 kW AM broadcast transmitter.  It consisted of a 5 kW modulated driver followed by a final amplifier.  Its ten cabinets held 25 tubes, including fourteen that were water-cooled.  Its frequency was crystal-controlled, and it was said to be the first transmitter capable of 100% modulation.  The transmitter was developed at Western Electric's radio test facility in Whippany, NJ, and operated sporadically during its development under the call sign 3XN in late 1927.   On January 9, 1928, an open house tour of facility was held for the members of the Institute of Radio Engineers.  The first commercial installation of the 7-A transmitter was made at WLW in Cincinnati in August of 1928.  

All the design innovations created for these early custom transmitters were quickly rolled into the first 50 kW factory-built design - a joint effort of General Electric, Westinghouse and RCA.  The 50-B – first branded as a General Electric product, but later marketed under the RCA label, was fabricated at both the G.E. and Westinghouse factories.  This massive rig required an RCA-designed two story building to house it.  The main unit, on the upper floor, consisted of four groups of operating panels:  the first was a complete RCA 5-B five kilowatt transmitter, which featured dual crystal oscillators and mid-level Heising modulation.  It was followed by a 50 kW Class A linear amplifier with two water-cooled UV-862 tubes, each rated at 50 kW.  The third set of panels contained a row of six mercury vapor rectifier tubes for the plate voltage, and the fourth panel was for overall power control.  On the lower floor, motor-generators provided DC power for the tube filaments.  Distilled water cooled the tubes, with water pumps and a heat exchanger feeding an outdoor spray pond.  The first 50B went to WTIC in Hartford in 1929 - it thereafter became an industry standard product, installed at most of the major clear channel stations in the country, including all of the NBC-owned 50 kW stations.  A number of these systems remained in use until 1960s.

WLW’s Monster Transmitter

 In 1932, this trio of American electronics manufacturers took on their biggest broadcast assignment yet –a massive 500 kW transmitter custom-built for Powel Crosley’s WLW in Cincinnati.  That June, the Federal Radio Commission granted WLW’s request for experimental “super power” operation, and Crosley tendered his order for the country’s largest AM broadcast transmitter.  RCA, Westinghouse and G.E. collaborated on the design, each building sections of the system.  WLW’s existing Western Electric 7-A 50 kW unit became the driver for a huge modulated final amplifier containing twelve water-cooled 100 kW PA tubes, and with another eight serving as modulator tubes. To reduce the massive power consumption of such a huge system, high-level Class “B” modulation was employed.  This technique, invented by Loy Barton with his patent assigned to RCA in 1932, substituted a modulation transformer in place of the customary Heising reactor at the final amplifier stage.  This allowed the use of a Class “B” modulator and Class “C” power amplifier, resulting in considerable power savings. 

WLW operated at 500 kW from 1934 to 1939 under an experimental license that was terminated when the FCC decided to establish 50 kW as the ceiling for all United States AM radio stations.  No other AM broadcast band station in the United States has operated with as much power, either before or since.  Although it has been unused since 1939, this transmitter still sits in the WLW transmitter building in Cincinnati.

The Search for Power Efficiency

We can consider the WLW transmitter to be a third-generation design due to its use of High-level Class “B” modulation.  The previous generation of transmitters generally delivered good quality, stable signals with reliable operation and clean audio quality.  The implementation of Class “B” modulation represented the first step towards improved efficiency and reduced power consumption.  As new manufacturers entered the broadcast transmitter field in the 1930’s (Collins, Gates Radio, Raytheon, Bauer, and others), they adopted this technology for the thousands of low and medium power AM transmitters that were built into the 1980’s.  But at 50 kW, the physical size and cost of the huge modulation transformers was a disadvantage, and their high electric power cost was still an issue for the country’s hundred-plus 50 kW AM stations.  The search continued for even more efficient and cost-effective transmission systems.

RCA took a step forward with the introduction of its high-efficiency air-cooled tubes, which eliminated the elaborate and troublesome water cooling systems of earlier designs.  In 1947, RCA introduced its model BTA-50F, which utilized its 5671 thoriated tungsten filament tube.  The transmitter was quickly adopted by a number of important stations in the U.S. and around the world.  Westinghouse and G.E. also introduced similar designs.

In an effort to eliminate the modulation transformer and further reduce power consumption, Western Electric introduced its Doherty power amplifier in 1938.  Invented by William H. Doherty of Bell Telephone Labs, it utilized two Class “B” final amplifier tubes – one generated the signal up to the carrier level, and the other added the extra power needed for modulation peaks.  The first Doherty transmitter was installed at WHAS in Louisville, and it was exclusively utilized by Western Electric until 1953, continued afterwards at Continental Electronics when that company purchased Western’s transmitter division.  Continental built its updated versions of the Doherty amplifier through the 1990’s.

Another efficiency improvement was outphasing modulation, based on the 1935 design of Frenchman H. Chireix, and first developed by McClatchy engineers in 1948 at the company’s station KFBK in Sacramento.  This method completely eliminated the high level modulation section.  Instead, the outputs of two Class “C” tube amplifiers were combined 135 degrees out of phase.  Phase modulation was applied to each amplifier at a lower power stage, so that the amplifiers were in phase on positive peaks (adding), and 180° out of phase on negative peaks (canceling). Adopted by RCA, the technology was marketed under the Ampliphase brand name and sold in various models between 1956 and 1978. 

Today, the most commonly used AM technology is Pulse Width Modulation (PWM), first introduced in 1978 by the Broadcast Division of Harris Corporation (formerly Gates Radio Co., now known as GatesAir).  This design utilizes high frequency pulse switching of tube’s plate voltage, with the duty cycle (width) of each pulse corresponding to the modulation percentage.  This pulse train then passes through a low pass filter that removes the pulses and delivers smooth modulated DC to the final amplifier.  First implemented in the Harris MW-50 tube transmitter, it has since been adapted by most manufacturers to today’s solid state MOSFET power amplifiers.  (GatesAir, Nautel, Broadcast Electronics, several others).   In 1991, Harris also developed an innovative digital modulation method which it applied to its solid state DX-10 and DX-50 transmitters.  In this technology, analog audio is converted to digitized data which turns on and off a series of low power solid state amplifier modules that are added to create the modulated waveform. 

These evolutionary developments in transmitter design – new modulation methods, better cooling systems, and solid state power amplifiers – have seen the overall transmitter efficiency (AC in to RF out) increase from under 25% in the early 1930’s to nearly 90% today.  Along the way, many of the technologies developed for AM broadcasting also found their way into products designed for the communications, aircraft, and amateur markets.  A few technologies became obsolete and disappeared, only to reappear later in a new form – as witness the modern liquid-cooled FM and TV transmitters.  While the future of the legacy AM band is uncertain as it approaches its 100^th birthday, it’s certain that many of the technologies developed for that industry will continue to live on in other applications.

RESOURCES:

• Big Business and Radio by Gleason L. Archer, 1939
•  Commercial Broadcasting Pioneer: The WEAF Experiment, 1922-26 by William Peck Banning, 1946
•  “United States Early Radio History”, by Thomas H. White:
•  “Expanded Audion and Vacuum-tube Development (1917-1930)”, https://earlyradiohistory.us/sec014.htm
•  Broadcasting Becomes Widespread (1922-1923), https://earlyradiohistory.us/sec018.htm
•  Wikipedia:
• Lee de Forest     https://en.wikipedia.org/wiki/Lee_de_Forest
•  Doherty Amplifier:  https://en.wikipedia.org/wiki/Doherty_amplifier
•  “Federal Radio Commission, Fifth Annual Report to Congress”, 1931
•  “Radio Telephone Broadcast Equipments (1-A and 2-A)”, a brochure published by Western Electric Company, May, 1922.
• “Radio Telephone Broadcast Equipments (106-B)”, and “Radio Telephone Broadcast Equipments (105-C)”, brochures published by Western Electric Company, 1928.
• "3XN", booklet prepared for the Institute of Radio Engineers for their tour of the Western Electric facility at Whippany, NJ, January 9, 1928.
• "Spanning a Continent" booklet, General Electric, July, 1926.
  
• “Radio News” Magazine, October, 1927 - “New Experimental 100 kW Transmitter at Schenectady”
• “Radio Digest” Magazine, 8-29-1925 – “General Electric Company Builds Big Laboratory for Transmission Study”
•  “Radio Digest” Magazine, 9-12-1925. - “High Power Downs Static, Tests Show”
•  “Powerful Station Reaches Far in Day – WJZ Uses 50,000 Watts, Hits Texas at Noon” – “Radio Digest” Magazine, 12-5-1925.
• “QST” Magazine, October, 1929 – “WTIC, America’s Most Modern Broadcasting Station”.
• "Radio World" Magazine, July 17, 2007 - "Loy Barton, a Forgotten Radio Pioneer", by James E. O'Neal
• “Proceedings of the Institute of Radio Engineers”, Vol. 22, No. 10, October 1934, pg. 1151 – “The WLW 500 Kilowatt Broadcast Transmitter”.
• RCA Broadcast News No. 119, Feb. 1964, “Ampliphase  … For Economical Super-Power AM Transmitters” by D.R. Musson.

 NOTE:  This article originally appeared in the October, 2018 issue of "The Spectrum Monitor" magazine (Vol. 5, No. 10)