QST Magazine August 1922
On June 28th Mr. E.H. Armstrong delivered another address on his latest invention before a meeting of the Radio Club of America which packed and overflowed from the lecture hall at Columbia University, New York City. A crowd hungrier for information we never saw. Every square foot of space was occupied by a radio ham of some sort endowed with a worried look, flapping ears, a craning neck and a pencil and paper. They simply ate up every word Mr. Armstrong uttered and watched in rapt admiration as he demonstrated his latest development.
His talk was based mainly on his 
original
paper before the I.R.E. which was reviewed in the last QST, but it was simplified
and some purely theoretical discussions omitted and some actual wiring diagrams
and circuit constants supplied for the practical amateur. We will not consume
space in reporting that part of his. lecture which duplicated his I.R.E. paper,
and the reader is referred to the article on "Super-Regeneration" in our issue
for July.
We will confine ourselves in the present article to the presentation of what new information we gleaned from the second address.
The recommended "variation frequency" is now a little clearer: for phone and I.C.W. it should be practically above audibility, say 12,000 to 15,000 cycles; for spark, the same, if the natural note is to be preserved, or an audio frequency may be used with increase in strength if one does not mind having the tone changed into a peculiar hard twang; and for C.W. telegraphy it may be either a superaudible frequency used in connection with a separate heterodyne for beat reception or the variation frequency itself may be audible, say 500 to 1000 cycles, thereby modulating the incoming C.W. but giving every signal exactly the same tone, which is hardly feasible where normal QRM prevails. Major Armstrong presented the diagram shown in Fig. 1 of our last article and discussed it briefly, recommending big aircore inductances of about 2 henries as preferable to iron-core windings for the audio oscillator. Now a large part of the attendance at this meeting was made up of broadcast listeners and it seemed that the greatest interest of the evening attached to the arrangement that worked best for telephony.
Accordingly the diagram of Fig. 2 of our last article was shown as a circuit more suited for phone reception. This is the scheme wherein the first tube is both the regenerative amplifier and the detector, with the phones in its output circuit, while coupled to it is the super-audible oscillator which produces a variation of the positive resistance of the first tube. The speaker pointed out, however, that better results could be had by performing the detection in the second tube or oscillator, which led him to our Fig. 4 of last month, wherein the phones are in the plate circuit of the oscillator. This is the circuit of the. future, Mr. Armstrong said, altho this diagram shows only the fundamental parts and some auxiliary apparatus is necessary to make it a practical device. The practical circuit shown in the diagram herewith is the result, and is an outgrowth of Fig. 4 of our July article.
Colonel Edwin Armstrong invented the regenerative receiver, the super-regenerative,
the superheterodyne, and perfected FM. In other words, Armstrong was the
inventor of radio as we know it.
Now if no audio-frequency amplifier is to be used, the phones may be connected across the fixed .005-mfd condenser in the output circuit of the second tube. The amount of energy developed by the oscillator at the variation frequency, however, is sufficient to completely paralyze an audio amplifier if impressed on its grid, and a low-pass filter is necessary between the detector-oscillator and the amplifier to keep out the 12,000-cycle component and yet pass the signal frequency. Any form of such filter is satisfactory, the arrangement shown in the diagram consisting of two 12,000-ohm Lavite resistances, an air variable of .005-mfd maximum and an iron cored inductance of 0.1 henry being merely a make-shift built up of pieces of equipment Mr. Armstrong had at hand. The variable condenser is tuned to eliminate the variation frequency in the amplifier by providing a shunt path across the primary of the amplifying transformer of low impedance at the frequency to which it is tuned. The amplifier valve was a Western Electric telephone repeater with 200 volts on its plate. An additional 200 volts is not essential; it might as well be supplied from an additional 110 volts and then connected to the same 90-volt battery that supplies the other tubes.
Hard tubes were used exclusively, it being impossible to maintain a stable adjustment with soft tubes. The grid of the first tube had a negative bias of 3 volts; this was merely to make the operation occur on the proper part of the characteristic curve and might as well have been attained by tapping off a different plate voltage for that tube, if any more convenient.
On test, this set picked up vocal selections from WJZ on its small loop and delivered them very QSA thru a loud-speaker connected in place of the phones. We had no idea of the amount of energy present, however, until Mr. Armstrong plugged in the phones instead of the horn. It almost tore up the phones; they clicked and rattled and the diaframs hammered on the magnets so hard that it sounded like a Western Union sounder instead of a lady doing her best to sing.
We found ourselves regretting that Mr. Armstrong was confining himself chiefly to a circuit best adapted to phone reception. He had another set-up ready for demonstration, embodying his "Case 3" where both the positive and negative resistances are varied simultaneously, and this was the arrangement which he had previously recommended as giving the best results in the hands of a patient amateur. This is the scheme of Fig. 3 of our July article, wherein the first tube not only is the regenerative amplifier but has a second feed-back coupling whereby it generates its own variation frequency, and the second tube is the detector. This set was demonstrated, and proved a whizz. It seems to us that it's the set for us amateurs. It is simpler to arrange as far as apparatus is concerned, and the mind of a ham will immediately turn to the ease with which an old honeycomb coil mounting can be worked in to provide for the variation-frequency coupling, using coils for the super audible frequencies and plugging in small iron-cored transformers or big condensers across big honeycomb coils when the audible frequencies are to be used a most versatile circuit, one which can be changed easily for any frequency, involves both kinds of variation, and is designated by Mr. Armstrong as the one which gives by far the best results. In fact, if the variation frequency is above audibility the separate detector is unnecessary the phones can be placed in the circuit of the first tube. The arrangement is then reduced to a single-tube affair but equal, Mr. Armstrong stated, to a 6-tube superheterodyne.
First it was demonstrated with two tubes, the variation being at audio frequency. The signals were estimated as representing an energy between 100,000 and 1,000,000 times that evident from the same signal in an ordinary regenerative detector and two step audio amplifier. Then the detector was disconnected and all the functions performed on one tube, a phonograph record from WJZ being audible all over the hall. A weak, high-pitched squeal was audible, presumably being the variation frequency, well within the limits of audibility. A second tube was then added, as an audio frequency amplifier of the detected signal, and the racket was very QSA. The phones were connected instead of the loud speaker, to give an idea of "what was there," and honestly it was a crime to put that much stuff into a pair of cans - it was enough to ruin them and to us seemed quite as loud as the other set had been with its three tubes. This we took as evidence of the correctness of the dope that the lower the frequency of the variation the greater the regenerative amplification and that audibility is sacrificed for quality in the super-audible variation employed for phone reception in the set illustrated in our drawings this month.
We regret that we have no constructional data respecting this two-tube set of July's Fig. 3 which worked so well for Mr. Armstrong, yet there seems but little needed. It appears to us as being the best yet for us telegraphing amateurs. And for C.W. telegraphy it seems to us that instead of using a super-audible frequency and a separate heterodyne we might use some moderately high-pitched frequency, say around 5,000 cycles, giving the immensely greater amplification that the lower frequency provides, and still heterodyne it separately with the selectivity provided thereby and the ability to read different C.W. signals with different notes.
Mr. Armstrong does not recommend the use of the super-regenerator on an aerial. He says that signals will be as loud on a loop as on an antenna, as in any event they build up towards infinity and are limited only by the tube characteristics. Use on an antenna furthermore is inadvisable because when improperly adjusted the set radiates very strongly.
We mentioned last month that we thought the super-regenerator was a first-power amplifier. It seems it is even better than that. Ordinary radio amplification falls off rapidly as the wave length shortens, and the big advantage of the super-heterodyne is that it maintains constant amplification regardless of the shortness of wave. But the super-regenerator amplifies inversely as the square of the wave length and works better the shorter the wave length, for example giving 16 times the response on 50 meters that it will on 200 meters, and is essentially a short wave device. In fact its upper limit of efficient action is about a thousand meters.
Now at first glance the diagram looks complicated but the operation is much simpler than might be supposed. The oscillator after once being adjusted requires no changes over a considerable band of wave lengths so the variable condensers in the middle portion of the diagram play no part in tuning in the DX station. In fact, after their correct capacity has been determined, equivalent fixed condensers may be substituted. The secret of best operation is to learn the meaning of certain sounds which indicate a poor adjustment in a part of the circuit. These indications are indescribable but, according to Mr. Armstrong, when "gotten onto," make the operation of a super-regenerative set as easy as the ordinary regenerative circuit with which we are all familiar.
 | ||
 | | |
