616 SEISMIC METHODS [Chap. 9 



for reflection seismographs are more rigorous than for refraction seismo- 

 graphs because not mereh^ the first but later impulses have to be recorded 

 as faithfully as the particular geologic situation requires and must be 

 balanced in regard to both amplitude and phase. The following section 

 deals chiefly with the caUbration of electrical reflection seismographs, that 

 is, calibration of pick-ups, amplifiers, and galvanometers and determina- 

 tion of over-all response, 



1. Detectors. Both mechanical and electrical characteristics must be 

 determined. The former include: (a) natural frequency, (b) deflection 

 (geometric magnification), (c) friction, and (d) damping. 



The natural frequency of the detector may be determined in a number 

 of ways. A mirror device may be attached to the mass of the seismograph, 

 and its free vibration may be recorded on a laboratory camera provided 

 with a timing mechanism. For this test damping must be eliminated. 

 If electromagnetic damping is applied it should be disconnected. From 

 the record, natural frequency and friction may be determined as previously 

 described. The natural frequency may also be obtained from the reso- 

 nance frequency by (a) driving the seismograph with an electrical oscillator, 

 or (b) by observing its mechanical or undamped electrical response on a 

 shaking table. 



Mechanical deflection tests make it possible to determine the static sensi- 

 tivity and thus the static magnification of detectors. A small mass is 

 placed on the seismograph mass. Its deflection is observed as stated 

 above, or the deflection of the magnification lever is observed by a micro- 

 scope or mirror device. 



For the determination of damping, overshoot records are best suited, 

 since the detectors are usually critically or near-critically damped. Damp- 

 ing may be determined also from the shape of the dynamic response curve 

 of a detector taken on a shaking table. Possibly simpler is the procedure 

 of driving the detector from a beat oscillator with constant input at varying 

 frequencies. This will result in a dynamic response curve of a force-driven 

 device, from which the damping rate may be determined. Damping 

 tests should be made for various oils and various temperatures unless an 

 electromagnetically damped detector is used. 



Virtually the only important electrical characteristic of a detector is its 

 output for a given ground amplitude and frequency. It is best defined in 

 units of open-circuit volts per micron ground motion (see Fig. 9-105). 

 Since it depends on the impedance of generator and load, detectors of 

 different impedances can be compared only by using a matching trans- 

 former whose secondary may be coupled directly to a vacuum tube volt- 



^2 For the details of these and other seismograph calibration methods see Heiland, 

 op. cit., 434-454 (1934). 



