INTEREST in measuring the surface impedance of metals at high frequencies and low temperatures arises in connection with the study of the phenomena of superconductivity and the anomalous skin effect and, recently, also of cyclotron resonance. A number of methods for measuring the surface impedance of metals are described in the literature H3, most of which were developed for studies of superconductivity. The methods of references 1--4 are based on the measurement of the Joule heat developed by high frequency currents in the specimen, and methods 5-13 on measuring the electrical characteristics of a resonator containing the specimen. The sensitivity to changes of surface impedance achieved in the most successful experiments described in references 1-12 does not exceed ~ 0.5 per cent. The sensitivity of method 13 is apparently considerably higher, but because of insufficient data it is unfortunately impossible to evaluate it more precisely.
All these methods are insufficiently sensitive for accurate quantitative studies of effects, such as cyclotron resonance, which show up as a change in surface impedance of a few per cent or even of a fraction of I per cent. In addition, when evaluating the possibilities of different methods, not only the smallness of the effects studied must be borne in mind, but also the fact that these effects show marked anisotropy. In order to study anisotropic effects in a rational manner, it must be possible to apply the high frequency currents flowing in the specimen and the steady magnetic field in a definite direction relative to the crystallographic directions in the specimen. This can only be realized if the surface of the specimen is plane and the currents in it are rectilinear. This does not hold for any of the work described in references 1-13. In this respect the resonator of reference 4 is constructed better than the others, but even here the currents in the specimen were not rectilinear everywhere.
These considerations indicated the necessity of developing a better method of studying the surface impedance of metals which would, primarily, satisfy the requirement of greater sensitivity and make possible the study of anisotropic effects. This task was undertaken and resulted in a new frequency modulation method which is described here. The first experimental results obtained with this method have been publishedJ 4,15