We use Optical Character Recognition (OCR) during our scanning and processing workflow to make the content of each page searchable. You can view the automatically generated text below as well as copy and paste individual pieces of text to quote in your own work.
Text recognition is never 100% accurate. Many parts of the scanned page may not be reflected in the OCR text output, including: images, page layout, certain fonts or handwriting.
loss which cannot be corrected at the print time.
is to be noted that surface induction is easured by this method at discrete freiencies only; i.e., the frequencies wr*e the maxima occur. Also the lovjst frequency measured is that of the lo\st maximum. Therefore, the longmethod is concerned with middle
an
high frequency measurement only.
he long-gap method can be divided three steps: measuring frequency :s, measuring the output of the tape to 3e evaluated, and computing the su:ice induction from the data.
the experiments, a long-gap head made and connected open circuit to relatively flat amplifier. The head da a gap length of 1 8 mils which caused thj maxima to lie approximately 1 kc »rt at 15 in. /sec. A considerable eflrt was made so that the edges of the »a were as nearly parallel as possible. Aually three heads were tried, the data sh ving the results of the best one in this
1 1 the first step, the frequency losses & be measured by any of the methods dcribed for the short-gap method. [r(:he experiments the change-of-speed mhod using tape was used exclusively.
igure 10 shows the octave losses as insured and the accumulated freqincy loss as computed from the meas111 i losses. In measuring the octave io with speed change of two, a gliding pp was recorded in the region of each trkimum so that suitable output could txbbtained.
In the next step, the calibrated system Mi used to measure the output of a dling tone on a tape whose surface ljuction was to be determined. The Ijface induction was known to be the Hie as that which was measured with tl short-gap method so that a compariW could be made between the methods. Ifis was insured by using the same system and head, the same re
ing current characteristic and bias nt, and the same piece of tape.
DB
OCTAVE LOSSES
-2 -4
DB 0
-2
4 6
*»s^
ACCl
— ^— — ^—
JMUL
i—
ATE
^ •
:D
— „
LO
'--*.
ss
^v,^^
x^
Fig. 10. Long-gap frequency losses.
The curves in Fig. 1 1 show the output at 15 in. /sec corrected for the frequency losses. The heavy line drawn through the maxima is therefore the surface induction as found by this method. The locus of minima is shown indicating modulation noise below 3 kc and signal above 3 kc. The latter is due to incomplete cancellation because of lack of parallelism between the edges of the gap. This may cause a slight error in maximum readings. With an ideal head, the maxima would not fall off as rapidly at the high end.
Figure 12 shows similar curves taken at 7.5 in./sec of the surface induction at that speed. Again the output curve is corrected for frequency losses. Here the difference between maxima and minima is not more than 5 db at the shortest wavelength. The output of the system at the minima consists partly of noise but mostly of signal as determined by listening and by filtering. The accuracy of surface induction at the short wavelengths is obviously not very great with this head due to an imperfect gap.
The effect of parallelism between the edges of the gap can be illustrated in this way. The point where the maxima and minima converge represents an error in parallelism such that n wavelengths appear across one side of the gap and n _|_ £ wavelengths appear across the other side of the gap. With the experimental head, it appears that convergence would
J. D. Bick: Measuring Surface Induction
523