US1596916A - Measuring transmission delay - Google Patents

Description

Aug. 24 1926.

S. B. WRIGHT ET AL MEASURING TRANSMISSION DELAY Filed Sept. 1, 1925 HHIHII INVENTQRS 61B. 14% I06 1 BY [KI 1. 1 er ATTORNEY Patented Aug. 24, 1926.

safes PAET @FFHQE.

SUMNER B. WRIGHT, OF EAST ORANGE, KENNETH W. PFLEGER, 0F ARLINGTON, NEW JERSEY, ASSIGNORS TO AMERICAN TELEPHONE AND TELEGRAPH COMPANY,

I A CORPORATION OF NEW YORK.

MEASURING TRANSMISSION DELAY.

Application filed September 1, 1925. Serial No. 53,939.

An object of our invention is to provide a new and improved method and suitable apparatus for measuring delays in transmission lines and other transducers. Another object of our invention is to provide for a wide range of cases, including delays of several periods and delays of only a small frac-' tion of a single period. In one form in which our invention may be embodied, two oscillators are employed and the frequency is varied from one to the other corresponding to an easily determined change in the phase shift through the transducer under investigation. The difference of frequency between the two oscillators may be determined by the beat method and from these data the delay through the transducer may be computed. In the accompanying drawing we have illustrated one example of practice according to our invention but it will be understood that the invention is defined 1n the appended claims and that the following specification relates to this particular embodiment thereof.

Referring to the drawing, this is 2. diagram showing circuits and associated apparatus by which. our invention may be practiced.

Two oscillators O and 0 are provided, each adjustable as tofrequency. Either or both of them may be connected through two parallel paths to the receiver R... Also, either or both oscillators may be connected through certain elements .of apparatus 'to the polar relay PR. If. the transducer Q gives a gain instead of a loss, then enoughloss is interposed at L to make the overall efiect a loss. Further details of structure will be pointed out in connection with the following description of .the procedure to measure the delay through the; transducer Q.

Switch S is closed downvand switch S, is closed on the point 1. Current from the oscillator 0 will reach the receiver R over two parallel paths, and in general there will be a tone heard in the receiver corresponding to the frequency of the oscillator 0 The frequency of the oscillator 0 is now varied slowly and the resistance R is adthe delayover the other path by 180 degrees'or an odd multiple of 180 degrees.

Now reverse the switch S and a loud tone will be heard in the receiver R Then throw the switch S down to connect both oscillators O and O alike in the circuits leading to the receiver R... The beats between the two frequencies for these two oscillators will be heard in the receiver R0 and the oscillator 0 should then be adjusted in frequency to no beat, thus giving-assurance that it has the same frequency as 7 0 Now disconnect oscillator 0 and throw S Assugne first that the transducer Q has a delay of several periods, as may readily be the case for a loaded cable circuit, for example. The frequency of the oscillator 0 will be gradually increased and presently a frequency f, will be reached at which the tone disappears from the receiver R This will mean that the frequency has been changed from f,. to f over a range that changes the delay through the transducer Q by a half-period or 180 degrees.

The formula for this delay is TILZB/dw where T is the delay in seconds of time, ,Bis the phase shift which is 1rin the present instance, and o) is 2w times the frequency. The frequency increment is f f,.

Throwing the switches S vand S up and closing the switch S, causes currents of these two frequencies to be transmitted through the potentiometer P, and their electromotive forces are superposed in the detector D. Ac-' cordingly, the armature of the polar relay PR in the detector output circuit vibrates to the beat frequency and by means of a suitable calibration curve, the reading of the Substituting in the foregoing formula and this reduces to But if the delay through the transducer Q is not as large as several times a halfperiod, then the change of a half-period between the two frequencies will be too great a change to get a precise value at a particular mean frequency, and in such a case we pursue a somewhat different method.

After having proceeded as before to the 'point of adjusting the frequency of the oscillator O to the condition of zero tone in the receiver B with the switch S, on the point 1, the switch S is shifted to one of the other points, 2 or 3, so as to interpose a small adjustable reactance such as that due to coil L or condenser C. As in the first case, designate the first frequency adjustment for oscillator O, as f,, and also as in that case readjust it to another frequency f at which there is again a null effect, this time with thesai-d small reactance in series. The artificial lines L, and L, are made up of resistances and are of considerable equivalent length and each is of 600 ohms value looking into it from between them. Accordingly, the intermediate circuits through the reactance G or L and the two artificial lines L and L has an impedance value of 2 times 600 ohms plus'the reactance value referred to. In short, the impedance of this circuit expressed numerically is 1200+i2'n'fL, as-

. suming that the switch S is on the point 3.

It follows at once that by shifting the switch S from the point 1 to the point 3, there has been introduced a phase shift.

AB:tan" (27rfL/1200) We know the frequency increment, and accordingly A (f2rf1) which consists in changing the frequency to annul the delay due to a small increment of reactance in the circuit of the transducer, and from the data of these changes computing the delay.

3. In combination, an oscillator, a receiver connected therewith by two paths, one comprising a transducer whose delay is to be measured, a reactance and a switch by means of which it may be cut in or out of one of the paths, and means to adjust the frequency of said oscillator.

4. In combination, an oscillator, a receiver connected therewith by two paths, one com: prising a transducer whose delay is to be measured, a reactance and a switch by means of which it may .be cut in or out of one of the paths, means to adjust the frequency of said oscillator, and artificiallines connected in on both sides of said reactance.

5. The method of measuring the delay of transmission through a given transducer at a certain frequency, which consists 'in receiving the input and output currents in a common receiver and adjusting the frequency to zero tone, then causing a delay in one path by a small known angular magnitude, and adjusting the frequency again to get zero tone and computing the delay from the data of these changes.

6. In combination, an oscillator adjustable for frequency, a, receiver connected therewith by two paths, one comprising a trans ducer whose delay is to be measured, means to reverse the current in one path if the said delay is large and make a smaller phase shift in one path if the said delay is small,

and means to determine the frequency dif ference between two adjustments of the oscillator. v

' In testimony whereof, we have signed our names to this specification this 27th day of August 1925.

SUMNER B. \VRIGHT. KENNETH W. PFLEGER.

Images