Radio Broadcast (May 1928-Apr 1929)

Part I of a Series GRID-LEAK GRID-CONDENSER DETECTION By FREDERICK EMMONS TERMAN Stanford University DETECTION is a subject upon which little real information is available to the radio experimenter. While amplifier circuits are designed with full knowledge of the results that will be obtained under different conditions, the detector is left to chance. Recent investigations have shown that grid-leak grid-condenser detection of weak or moderate-strength signals is determined by a single new tube constant, and that the exact behavior of the detector can be obtained simply by using this new constant. The sensitiveness and distortion resulting with grid-leak detection when strong signals are being received can also be readily analyzed, and it will be shown that a properly operated grid-leak " power " detector is more sensitive and gives less distortion than the usual plate-rectification detector. A few of the worth-while articles which have considered this subject are as follows : "A Theoretical and Experimental Investigation of Detection of Small Signals," by E. L. Chaffee and G. H. Browning, Proc. I. R. E., 15, 113; February, 1927. "The Rectification of Small Radio-Frequency Potential Differences by Means of Triode Valves," by F. M. Colebrook, Experimental Wireless, 2, 865; 1925. "Detection by Grid Rectification with the High-Vacuum Triode," by Stuart Ballantine, Proc. I. R. E., 16, 593; May. 1928. 'some Principles of Grid-Leak Grid-Condenser Detection," by Frederick Emmons Terman, Proc. I. R. E., Vol. 16, p. 1384, Oct. 1928. "Detection Characteristics of Three-Element Vacuum Tubes," by Frederick Emmons Terman and Thomas M. Googin, Proc. I. R. E. Vol. 17, Jan. 1929. Process of Detection DETECTION is the name given to the rectification of high-frequency alternating-current voltages in radio receivers. In the grid-leak method of detection, the circuit for which is shown in Fig. 2, the rectification takes place in the grid circuit by making use of the curvature of the grid-current grid-voltage characteristic. The case of weak signals will be considered first. The grid-leak "power" detector acts very differently, and will be taken up in another article. The relation between grid voltage and grid current in a typical vacuum tube is given in Fig. 1. It will be noted that there is a small grid current even when the grid is negative with respect to the negative side of the filament. This is the result of the velocity which the electrons have as they leave the filament. In the absence of a radio signal voltage, the grid assumes a voltage which is the potential of the grid return lead (the lead which completes the circuit from the grid back to the filament) minus the voltage drop due to the grid current flowing through the grid leak. This can be readily seen by examining Fig. 2. This actual grid voltage is the operating grid potential, and gives the point on the grid-voltage grid-current characteristic of Fig. 1 at which the detector operates. The operating grid potential is usually within a fraction of a volt of the negative filament voltage when the grid return lead is connected to the positive side of the filament. A higher Fig. 1 -0.2 0 +0.2 GRID VOLTAGE, Eg -Grid-current grid-voltage +0.6 characteristic of a 201A-type tube resistance grid leak makes the operating grid potential more negative (or less positive), but the grid voltage changes only a volt or so when the grid-leak resistance is varied from | to 10 megohms. The principal function of the grid leak is to fix the operating grid potential at a point on the grid voltage-current characteristic suitable for rectification in the grid circuit. When a radio-frequency signal, such as developed by the tuned circuit LC of Fig. 2 is applied to the detector grid this voltage is superimposed on the operating grid potential, making the actual grid voltage alter Whal goes on in a detector circuit is not the easiest thing in the world to understand, but we believe Professor Terman has made the operation of grid-leak grid-condenser detectors as clear as possible. In this article he points out that there are two detector constants that tell the whole story about what a tube will do as a detector, and advocates that tube manufacturers put the values of these constants on tube cartons. We agree. Several other articles on the long-neglected subject of detection are awaiting publication. Some are from Professor Terman and others are from Roger Wise and his former associates at the Cunningham laboratory The Editor. nately more and less than the operating grid potential. This is illustrated in Fig. 1 in which Eo is the operating grid potential, Es is the amplitude of signal voltage, and the curve SSS is the variation in actual grid potential when the signal voltage is present. Principle of Detection THE signal makes the instantaneous grid voltage swing alternately from Eo + Es to E0 — E5, as indicated in Fig. 1. This fluctuation in grid voltage causes the grid current to vary, but, due to the curvature of the grid voltagecurrent characteristic, the grid current increases more during the half cycles when the signal voltage is positive than it decreases during the half cycles when the signal voltage is negative. The net result is a rectified current flowing in the grid circuit produced by the application of the radiofrequency signal voltage to the grid. Reference to Fig. 1 will make clear how the rectification is accomplished. When the signal is present the instantaneous grid potential varies as indicated by the sine wave SSS. This variation in grid potential causes the grid current, Iga.c, to vary according to the curve to the right in Fig. 1. The middle dot-dash horizontal line shows the grid current that flows when the grid potential is E„ (no signal present). The average grid current that flows when the signal is present is indicated by the light dash line. The difference between these two horizontal lines represents the rectified current flowing in the grid circuit as a result of the application of the signal voltage to the grid. The amplitude of the rectified grid current depends upon the amplitude of the signal voltage. When the signal is a modulated alternating-current voltage, the rectified grid current varies in amplitude at the frequency of modulation. Thus, when the signal is modulated at 1000 cycles, the rectified grid current pulsates in amplitude at a 1000-cycle rate. In Fig. 3 there is shown the rectified grid current resulting when an unmodulated, a simply modulated, and a complexly modulated wave is rectified in the grid circuit of a detector. The rectified grid current produced in the manner that has been described by the application of a signal voltage must flow through the impedance offered by the grid-leak gridcondenser combination, and will produce a voltage drop in this impedance. This drop causes the grid potential to become more negative by the amount of the drop, and the change of grid potential thus produced affects the plate circuit by ordinary amplifier action. It is the change of grid potential caused by the rectified grid current flowing through the grid-leak grid-condenser impedance that gives the detection of the signal. The explanation that has been given of grid-leak detection with weak signals differs considerably from the familiar one in which the function of the grid leak is to let the gridcondenser charge leak off and return to the filament. march, 1929 page 303 •