CARDIOVASCULAR SOUND 



689 



FIG. 7. In this oscillogram the instrumental frequency 

 dominates the picture. Since it is almost purely the natural 

 frequency of the recording membrane which is activated by the 

 sound vibrations, the recording suggests transient pure tones. 

 (From Orias and Braun-Menendez, Heart Sounds in Normal and 

 Pathological Conditions, 1939.) 



difficulty but these are not entirely satisfactory solu- 

 tions. The use of an electronic switch to provide 

 multiple recordings from a single electron beam^ 

 further detracts from the resolution at higher fre- 

 quencies. For these reasons, the opinion has been 

 expressed (19) that "the moving-coil galvanometer 

 (or its modifications) is today, pending the develop- 

 ment of a compact multiple-gun cathode-ray tube, 

 more suitable for phonocardiographic practice than 

 any other recording device." 



The moving-coil, or D'Arsonval, galvanometer 

 has a moving element consisting of a coil or loop of 

 fine wire .suspended in a strong magnetic field. C'ln-- 

 rent is conducted to the coil, causing it to turn in the 

 magnetic field. A mirror or writing arm is mounted 

 on the coil for recording cither photographically or 

 by direct writing. 



Some (8) have considered it necessary to use 

 galvanometers of different frequency characteristics 

 for recording information from different areas of the 

 frequency spectrum. In the opinion of the writers, 

 such a practice introduces further complications in 

 calibration. A 2-kc galvanometer can, and we think 



' As in the recorders of Electronics for Medicine, Inc., White 

 Plains, N. Y. 



should, be used for recordings at all frequency levels, 

 thus eliminating one variable in the interpretation of 

 findings in different channels. 



Flexibility in manipulation of the time scale is a 

 \ aluable feature of oscillographic recording. "Stretch- 

 ing out" the sounds is often desirable. Clearly a 

 greater paper speed is required for routine phono- 

 cardiography than for routine electrocardiography. 

 Whereas 25 mm per sec has been satisfactory for the 

 latter, a paper speed of 50 mm per sec is more useful 

 in phonocardiography. 



Figures 7, 8, g are examples of oscillographic 

 recordings. 



The write-out of the spectral phonocardiogram can be 

 made in several ways: /) The output of the filter can 

 be used to activate an electric spark which marks 

 electrosensitive paper. Electrosensitive paper has a 

 rather poor intensity depth. (In the spectral phono- 

 cardiogram time is represented by the horizontal 

 axis, and frequency scale by the vertical axis and the 

 loudness, amplitude, or intensity of a component at a 

 given frequency and time is indicated by the intensity 

 of the mark on the record.) Only about 10 to 20 db, 

 or at the most 30 db, from faintest to loudest can be 

 displayed using such a method. However, this inten- 

 sity span seems to be adequate for most purposes, 

 probably inasmuch as the sounds are "spread out" 

 in the frequency dimension. (Empirically, 6 db 

 initial equalization seems to give satisfactory results. 

 More or less is less satisfactory.) 2) Or, the output 

 can modulate a tiny light which is made to play on 

 photographic paper (fig. 10). The photographic 

 record provides improved intensity resolution but 

 again it is not certain that this refinement is necessary 

 in most applications. 3) Finally, the output can be 

 put into a cathode-ray oscilloscope with a long- 

 persistence face and the image photographed. The 

 last method has the advantage of flexibility in manipu- 

 lation of the time .scale without influencing the fre- 

 quency scale. 



One of the open questions in the final design of a 

 spectral phonocardiograph is the best method for 

 repetitively playing back the recorded information 

 for the seriatim frequency analysis. The magnetic 

 drum has the limitation of fixed playing time and 

 difficulties with "wow" and "flutter" (artifactual 

 variations in frequency in the final display due to 

 lack of uniform speed of rotation of the disk). Alterna- 

 tives to the disk include the use of tape loops with 

 either moving head or moving loop. 



Display of the frequency spectrum is responsible for 

 the three advantages of spectral phonocardiography : 



