Here is a post from the forum:
http://www.antiqueradios.com/archives/ElementsOfRadioServicingCh1-4.pdf
on the choice of 455 kHz:
The Climb To 455 kHz
By Richard T. Ammon
The whole intermediate frequency (IF) thing in early superheterodynes was more than just a competitive numbers game. From the first superhets, the frequencies used in intermediate amplifiers were, by necessity, very low by today’s standards, near audio thresholds.
In France during World War One, Edwin H. Armstrong developed a circuit to circumvent the vacuum tube’s inability to effectively amplify signals above 600 kilocycles (KC, now Kilohertz). These were considered extremely high frequencies, but are today at the low end of the AM radio dial.
His idea for the superhet meant an incoming, high frequency radio signal, that couldn’t be easily amplified in early triode tubes, had to be brought down to a lower manageable frequency. This change in frequency still had to retain the audio information impressed on the RF signal in the transmitter. Armstrong was forced to use a very low frequency barely above audibility.
By heterodyning or beating the output frequency from a “local” oscillator inside the receiver against the incoming radio frequency, a third and fourth frequency would also result from where the original two were mixed together. By using only one resultant, the much-lower third frequency, called the “intermediate”, amplification was easier to accomplish. At the end of several cascaded intermediate frequency stages, the audio, still riding the IF, would be pulled away in the second detector and amplified on its own. This entire process makes up the guts of a typical superhet.
After the war, Armstrong openly discussed his patented circuit. Enthusiasts, using his detailed plans and schematics featured in magazines and, later, from kits bought through mail-order houses, could build his receiver, similar to his 1919 model. Radio clubs and school classes could provide expertise in construction of these early sets. With varying results, amateurs hand-wound IF transformers that operated at these very low frequencies, most around 30 kilocycles. Because of its overall success, the popularity of the superheterodyne had no equal from its infancy.
In January 1922, General Electric through RCA introduced the UV-1714 and the UV-1716 RF transformers. This meant builders of superhets could get a ready-made IF transformer for less than nine bucks. RCA claimed the UV-1716 reached its peak amplification at 42.83 kilocycles while others claimed it was most efficient at 47.5 KC. By testing them today, collector Duane Bylund found they ran closer to 30 KC.
Charles Leutz bought the 1716 for his famous, early custom-build supers and his kits until RCA shut off his supply in 1923 in preparation for the introduction of their own superhets. Then, for his Model C, Leutz used Acme and General Radio transformers, operating at 42 and 33 KC, respectfully, as substitutes until he produced his own. The Leutz E.I.S. transformers peaked at 47 KC.
As other manufacturers cashed in on the phenomenon, they were restricted by the same limitations of circuit components. Early kits were sold with IF transformers and filters peaking only from 30 to 60 KC. Remler, Baldwin-Pacific, and Scott promoted 45 KC and the early Silver-Marshall IF module peaked at 50 KC. Western Electric built commercial receivers around their 45 KC IFs. With a different approach, the Daven resistance-coupled Super Amplifier in the 1925 Popular Mechanics super worked best at 45 KC.
Among the lowest successful IFs was from Madison Moore of Denver. For a short time in 1925, they used 35 KC transformers shielded in individual nickeled cases. Even lower at 30 KC were the All American and the Jefferson. The lowest this author found was a shortwave super designed in late 1926. Its low hand-wound 22 KC transformers were necessary since the receiver was used to listen to CW on 100 meters and did not emphasize voice reception.
Low IFs caused a multitude of negative side effects. Most obvious and annoying to the user was that some radio stations appeared twice on the tuning dial, due to the heterodyne process. Re-radiation through the antenna occurred until frequencies got high enough that the antenna was effectively detuned and those frequencies wouldn’t pass back through. “Wave trapping” was a problem at nearly all frequencies, but particularly in the lower range. Again, due to the nature of heterodyning in the first detector, “holes” were left in the incoming radio spectrum when it mixed with the local oscillator. Some stations wouldn’t come through at all. And, an advantage of higher IFs meant greater sensitivity.
By the mid-1920s, most manufacturers ran up their IFs as amplification in triodes improved. Lacault’s Ultradyne worked at 115 KC while the Victoreen successfully used 88 for over four years. Pinkerton’s Pink-A-Tone chose 150 KC in late 1924. Haynes-Griffin and Radio Receptor claimed 100 KC. But, H-G moved up to 150 the next year, 1925, as did Remler by late 1927. At that same time, Silver Marshall jumped to 112 KC with their new Jeweler’s Module, but by 1929 S-M peaked at 175 KC. The multi-tube Melo Heald Models 11 and 14 were at 125 KC. Using Western Electric 215 Peanut tubes, the Canadian-built Mercury pumped it up to 192 KC in the fall of 1925.
Outside the mold were several oddballs. The St. James kit in 1924 had glass-enclosed, dehydrated IF coils operating at 240 KC. One could build the Benson reflexed superhet in early 1925, using the second-generation Acme RF transformers running at 500 KC! But, topping them all in August 1926 was the Infadyne, distributed by Remler. Developer E. M. Sargent, using an IF module with extremely close tolerances and 199 tubes, created his first model powering through at 3150 KC! A month later, because of wave trapping issues, the IF was pushed up to 3490 KC, near 86 meters. This set was extremely popular because of its great performance.
Major changes were rapidly coming. For 1926, Madison Moore moved to 92 KC, but a year later successfully climbed to 490. Scott reached that frequency by mid-1929. HFL’s Isotone pushed up to 450 a year earlier, which was a far departure from 37 KC of their earlier superhets. Lincoln with its user-tunable IFs swept between 350 to 550 KC by late 1928.
The dramatic leaps to higher frequencies were due to the introduction of the screen grid vacuum tube. In September 1927, Ernst Tyrman brought out the first shielded, screen grid radio, the "Amplimax 70", running at 340 KC with impedance coupling rather than transformers. Many consider it the first true, modern superhet. Tyrman, who had designed early HFL receivers, broke the ice so others could follow the quick climb to much higher IFs. The next year, Tyrman chose 475 KC for his superhets.
In the early 1930s during the depression, because of the need for cheap, interchangeable parts between radios, the intermediate frequency of 455 kilocycles was adopted as a standard in the industry.
Bob S
