Radio broadcast .. (1922-30)

RATING RADIO RECEIVERS By ALFRED H. GREBE President, A. H. Grebe & Company, Inc. A Method Offered by A. H. Grebe & Company, Inc., as a Solution to the Problem of Describing to the Non-Technical Public the Merits of a Particular Broadcast Receiver. The author in the laboratory checking the work of Grebe engineers. Newspaper and magazine advertising of radio receivers consists mostly of high- sounding adjectives. It was our feeling that performance claims backed by defin- ite quantitative measurements would form a better basis of advertising than vague claims of superiority. This desire to find a reasonable basis upon which to write • SELECTIVITY c™ SENSITIVITY D AUDIO QUALITY i I Cynic turns fan TABLE Ihc superlatives when he conies in; don't hear Jtrnn with sales talk; this ii..rijeni'd cynic wants eviJfttce, And on this chart he finds the facts that show hitn exactly what to expect of ihe Grebe—clear-cut compurisons that shutter his shell. Now let hint listen to the set that is newtr than screen grid and watch him turn (an. He is critical but hi to the Grebe. This s« them with plenty to year ahtad of the field. Show him Ix.wshjrp- ly the Grebe i demands when put another. Thrill him by reach- ing out for weak, distant broad- casts. Close the sale wiih the' lifelike tone that enables him to identify every instrument, every voice. Then, after he has signed on the Jotted line, be -sure you Jeliveron time. You know how it is with this type of buyer; once you have sold him, you cannot get the set to his home quick.}- enough! There is <\ Grebe franchi separates one slut inn from Grebe radio A. H. GREBE & COMPANY, I Western Branch. H.5So. Sun PcJr Grebe copy has been responsible for the use of a new type of advertisement by our company. Our present series of advertisements have been based on a series of measure- ments on a number of standard makes of radio receivers. The technical data which forms the background of these tests are described in the fol- lowing paragraphs. In the rating of receivers there are three factors of ma- j o r importance—• sensitivity, selec- tivity, and fidelity. Stating these factors in engineering terms such as so many mi- crovolts per meter, or a loss of so many DB at 5000 cycles is understandable to engineers but means little to the layman. We have, therefore, endeavored to work out formulas for these various factors which will permit their expression in percentages so that a set has a fidelity of say 50 per cent., a selectivity of a certain percentage, and so forth. :rti prolit in the c. In addition to Setting normal business, i! si-lls tlmse who won !tl n«l otherwise he rcLidy for an- other >ejr. Allre.1 II Grebe — "In Ihe nrn [older hentx ,ji,lMhulrd 10 the pttbliv hy our deoler* we prove the nnilorn.tr hub .|uiilitv .il riii, new set in every i.nimr- uml eharactcHttlc at radio ie.'c->'ii,i. Uefupport The itarerocnl th.l Grebe flricel will nut he reduced with Itle pledge that Grehe **.J..lr will m.t be reduced. To I rtnfhi ic-huldert ilii, meant that pre/tM will nut he reduced - that even tale «>1. 10., Richmond Hill, New York Street, I ,os.A nijeles, California The advertisement above is indicative of the way in which Grebe explains the merits of its product to the non-technical public. Selectivity The ability of a set to receive the de- sired station with- out interference from another signal on an adjacent or nearby channel is indicated by the se- lectivity curves. The selectivity of all sets is roughly the same at the lower broadcast fre- quencies, but on higher frequencies the selectivity be- comes much worse and there is a great variation between sets. We are therefore using the selec- tivity at 1400 kc. as a basis for com- parison of sets, and are taking the width of the selectivity curve at 100 times the input at resonance. Therefore, if a set has a band width of 40 kc. it means that if our signal were 20 kc. off reso- nance on either side, the input to the set would have to be increased 100 times to give the same loud speaker output that we would have obtained with the set tuned to the signal. Since the sides of the selec- tivity curve are fairly straight, a set with a band width of 40 kc. at 100 times input would have a band width of about 20 kc. at 10 times input. This means that a bta- tion of equal strength on an adjacent chan- nel (10 kc. off) would only produce one tenth the signal in the detector, and this is not enough to interfere with the station to which the set is tuned. Therefore, a set with a band width of 40 kc. can be said to have "10-kc. selec- tivity " over the entire broadcast range. As a matter of fact, a set having 40-kc. band width will give 10-kc. channel selectivity even though the interfering signal on the adjacent channel is twice as strong as the desired signal. A set having 50- to 55-kc. band width is still capable of giving 10- kc. channel selectivity provided the inter- fering signal is no stronger than the desired signal. We have given a set with a band width of 40 kc. at 1400 kc. a rating of 100 per cent, in selectivity. The curve of per cent, rating against band width has been made steeper in the neighborhood of 40 kc. because a 1 kc. improvement in band width here is more difficult to obtain, and is also worth much more than the same improvement at a band width of, for ex- ample, 100 kc. Sensitivity The sensitivity of a set is expressed in terms of the voltage input to antenna re- quired to give a fixed output to the loud speaker. The figures used are in microvolts per meter. The actual input to the an- tenna of the set is four times this figure, since a standard antenna is assumed, hav- ing an effective height of 4 meters. The lower the figure for sensitivity the 276 RADIO BROADCAST FOR MARCH •