The Frequency Test from 2YA

Radio Record, Volume II, Issue 5, 17 August 1928, Page 27

The Frequency Test from 2YA

Some Additional Notes

(By

Meghom

ame. AST week some notes were ewe given regarding the frequency test, which has aroused a considerable amount of interest " amongst listeners generally. Quite a number of sets in Wellington gave audible signals up _ to 12,000, but apart from this fact there is no guarantee that in every case the standard ‘of reproduction. throughout the ordinary musical range would be of a very desirable quality, or that the low notes would be adequately dealt with. Although every effort should be made by constructors to reach the possible limits of reproduction at both ends -of. the audio-frequency scale, it is very evident that the presence of the deep, strumming bass notes are of far more importance to reproduction than are the ultra high frequencies that chiefly serve to fine up the characteristic sound of certain instruments with a tone that is distinctly their own. Certainly, the deep notes will be more readily missed than the high frequencies above, say, 10,000 cycles, that if present at all, must be so faint ‘that on actual test, their presence or absence would probably be indistinguishable to the average ear. LISTENER accustomed to hearing the deep notes reproduced by his own receiver, quickly notices their absence in one that does not deliver them, we the effect appears thin, weak and unsatisfying. Great effort has been made recently to bring about the necessary amplification and better reproduction of bass notes, especially in trans-former-coupled amplifiers, where they were more inclined to be lost, and by making, loudspeakers capable of reproducing such notes from any receiver The editor states that quite a number of readers are not clear as to the purpose of the recent test, so a brief explanation will be given before going further. Object of the Frequency Test. USICAL sounds are measured by the number of vibrations they produce per second, A single vibration is a movement of the air in one direction, and a complete vibration is the movyement away and return to the original position. This complete vibration is more often referred to as a "cycle" or "frequency." The method of causing the air to. vibrate varies in different instruments. Each note has its "fundamental" or strongest frequency, but the characteristic quality of the particular instrument is conveyed by higher though weaker frequencies known as "harmonics" or "‘overtones," set up at the same time, and having a frequency in some kind of agreement with the fundamental frequency, double, three times, and so on. be fundamental of notes on the piato ranges from 26 cycles for the low ext bass note, and 4096 for the otenest treble note. Other instruments come within this scale, but mostly have #% much reduced range, omitting an oc. tave or two of either the high or low notes, But all instruments have the above-mentioned overtones, and for the higher notes:some of these have a frequency above 20,000.

[NS order to obtain good reproduction of music, a receiver should be capable of reproducing frequencies much higher than those of the fundamentals of the highest treble notes of the piano, and at the same time should be capable of reproducing notes well down in the scale, this point being variously fixed at 32, 40 and 60 cyclees. AS a rule, conditions that favour the frequencies at one end of the scale are unfavourable to those at the other end, and tend to suppress them, so that a radio receiver is necessarily a compro: mnise between high and low notes, and is good or poor according to whettier its range of reproduction from high to low is extensive or narrow. The fre-

quency test was given tu enable listen. ers fo find just what range their receiver would actually cover. A diagram: is given showing the actual range of the piano, orchestral instruments, and speech. The Diagram. SCALE of frequencies from 10 per second to 10,000 is shown. giving on the left the lowest note of various instruments, and on the right, the highest notes, 4 soprano singer and several instruments. The range of two classes of loudspeaker is also given, but a few outstanding makes give a much greater range than is indicated in the diagram, as certain makes are capable of reproducing frequencies throughout, and also above the ordinary musical scale. The average human ear perceives frequencies down to about 25 per second, and below this notes are "felt"? rather than heard. This fact is made evident by the vibrations of the lowest pedal

notes of a pipe organ, from pipes 32 feet in length, or their equivalent, made by closing the end of a 16ft. pipe. Reproduction of Speech. HE broadcast receiver has to handle a fair amount of speech, and no better test of good reproduction is to be had. Many listeners complain that they get indistinct and "woolly" reproduction of speech, and in such cases the fault is usually in the receiver, and not in the loudspeaker. Depending upon the type of receiver, there may be reaction distortion, transformer distortion, general cut-off of high frequencies in the amplifier, or excessive condenser capabilities in the amplifier,

or each of these contributing its share to the general distortion. The highest speech frequencies are contained in the sounds th, f, s and z, the highest being about 6400 cycles, so if our outfit reaches this frequency, it is capable of reproducing speech with full clarity, other conditions in the receiver being favourable. The writer’s receiver performs particularly well in this respect. A receiver cutting off above 4000 cycles would fail badly on the above sounds, and the final s would disappear, Sounding more like z, though generally speaking, the speech‘ might be quite intelligible if not too weak in volume, Improving the Receiver. UE chief point for attention in the receiver will be the audio amplifier. It is essential that this, as a whole, should have an amplification characteristic that is independent of frequency. That is te say, that. the

amount of amplification should be equal on all notes, whether deep or high, that is, whatever their fre .quency. When this is done, a big task has been accomplished, for j31any cone structors are heavily handicapped from the commencement. Very often their receivers are built up gradually, a stage of audio added, not always with the most suitable components or value, sometimes because the required makes are not obtainable. One of the causes of distortion is found in the tendency of the audio amplifier to oscillate, when a reaction effect is obtained which undoubtedly. tends to strengthen certain notes more than others. It is in a great measure due to this tendency that not more than two stages of transformer coupling are recommended. In this respect audio. circuits seldom receive the same attention as do the high-frequency, though both are equally important. QGOMETIMES such oscillation is. ine cipient, or hidden, and in addition to affecting note reproduction, as men-« tioned above, may be sufficient to give a "coarse" quality to reproduction. In some amplifiers there is trouble from tendency to "motor-boating." A choke-condenser output-filter arranged so that one side of the loudspeaker is connected to A negative, will very often, prove a cure for this type of trouble. There have been rare eases in which such oscillation has manifested itself as a rhythmic increase and decrease of signals very like fading. In a trans-former-coupled circuit, reversing pris mary connections on one transformer will sometimes effect an improvement. If a B eliminator is used, the size of its smoothing condensers can be increased, High resistance in old B batteries can cause audio oscillation, which is easily cured in such a case by substituting 2! new battery. A 2 mfd. condenser across" the full B battery and another acros¢ the detector voltage is an advantage. — Transformer Coupling. Pus is the most popular form of audio coupling. and it is at the same time the system productive of the most varied results. The poor quality of many transformers on the market is responsible for a great proportion of the poor reproduction from both home« built and factory-made receivers. In order to obtain throughout the musical scale the even amplification mentioned above, it is necessary that the lower or bass notes must be suitably amplified. And these low notes are more difficult to amplify than high ones, because it takes 2500 times the energy to produce a definite sensation at 100 cycles as it does at 1000, and the lower we go in the scale, the more this proe portion rises. This is because the sensi« tivity of the human ear falls off rapidly on the lower notes, Now this amplification can ‘only be obtained’ by having a comparatively high impedance pri+ mary winding on the transformers, or at least on one of them, preferably the second. High impedance is obtained by increasing either the number of turns of wire or the -cross-section of the iron Continued on page 29,

(Continued from page 27.) core, or both. A transformer with high primary impedance is of low ratio, 3 or Bi to 1, and there is less chance of dlisMortion from a low ratio than from a high. The high impedance primary of a well-known transformer contains a mile more wire than the windings of the ordinary type. Now, if a transformer gives high amplification of the low and middle frequencies, it should also amplify the high notes equally well. If it is of good design, it will do this, and its peak or highest amplification may be around 4000 cycles, after which amplification will fall off. The small and cheap type of transformer has a high amplification peak on the high frequencies, but falls off rapidly lower in the scale, so that the lower frequencies ‘weaken rapidly, and the lowest are entirely lost. Sometimes this is not always quite a disadvantage, for a poor type of horn speaker would not properly reproduce the low notes if they were there. To obtain good quality amplification at high frequencies, it is necessary that the windings should be divided up into sections in order to pre- vent loss through internal capacity.

Fixed Condensers.. ad PART from the indispensable impedance coupling condenser in the first audio stage, the writer has no tone controlling or other condensers in his amplifier. The only other condensers present are the one built in the Ferranti A.F.3 transformer and one in series with the horn speaker to suppress low notes. Many constructors will find that by shunting condensers of different values across the loudspeakers, they . Will be able to settle upon one that gives . best results and may be incorporated in ythe circuit permanently. Tone control units have been constructed for the purpose, giving facilities for switching in different condenser values. The effect of such shunting condensers is to soften or subdue the high notes. Ifa resistance is placed in series with the speaker, it reduces the low tones more than the high. HE fixed condensers across the primary windings of the first and second audio transformers have a considerable influence upon the range of notes passed. Increasing their capacity reduces the high notes, so that a "sereechy" or "tinny" effect may be subdued by their increase. A shunting condenser may not always be necessary across the second primary, and is not always absolutely essential across the first, especially in the Reinartz circuit and its modifications, but it will pay to conduct a few experiments in this direction. An attempt should also be made to remove the conditions that necessitate such modification by adding capacity. Valve Influence. HE last valve plays a highly important part in reproduction, and only a power valve of suitable charactoristies can satisfactorily handle the output of an efficient circuit employing four or more valves. To run such a properly, there must be no lack of'B current. The plate current is the 4 equivalent of power, not because of the sieady plate current, but because the fluctuations of that current, caused by the impulses on the grid, are what actuate the speaker. Signals on the grid affecting the flow of plate current cannot reduce it to more than zero, so that the greater the current, the more. can

the variation or fluctuation be. There is a wide range of volume in most music, which cannot be _ properly handled by a set that has been designed chiefly with a view to B_ battery economy. Where dry batteries are used, better output may often be obtained from the last valve by increasing the plate voltage, provided that the speaker will stand the extra direct current. An output filter provides against any possible trouble in that direction. Actual tests show that lowering the resistance of the output valve in any way increases the lower notes in strength, but may not, seriously increase the high notes if the output circuit is adjusted to have a high impedance, in which case the full, increase of power may be obtained with the tone unimpaired. Filter Control of Tone. WHERE a choke-condenser output filter is employed, if the lower notes are too predominant, the coupling condenser between choke and speaker may be reduced. If the high frequencies are too prominent, some of them

ean be shunted away by means of a condenser of small value placed across the choke coil, thus allowing these frequencies to pass through the condenser instead of setting up voltages across the choke. OTHING has been mentioned about grid-bias, for although it has a great influence upon quality it does not come within the scope of this article. The whole movement is towards better reception in every way, and without correct grid-bias no two-stage amplifier can function satisfactorily for loudspeaker volume. Some listeners would no doubt think they were hearing the lowest notes of the frequency test when they were only hearing the overtone an octave higher, The writer found the crystal a handy check in this connection, switching off the B eliminator and listening to the crystal with headphones. One experienced constructor writes to say that whereas he was only able to hear frequencies up to 4000, his wife could hear up to 8000. In last week’s article it was stated that the frequency of high harmonics would reach as high as 30,000 cycles. This should have been 16,000 cycles.