Stability in GW Detector Units
(8-10-91)
It had come to the attention of the writer that many would-be researchers in GW detection techniques may have been discouraged by the tendency of some circuits to develop instabilities or oscillations at the high gain levels. This was particularly true when recently made IC devices were used. These instabilities are believed to be due to scalar-type signal feedback from the output circuitry to the input capacitor detection device. Such feedback could be circumvented to some extent by the use of a small capacitance, say less than 500 pF, across the detector section feedback resistance --- but such "degeneration" also results in significant loss of conversion gain. However, a simple "fix" was reported to me recently by Bill Ramsay. Bill has found out that a small resistance in the order of 47 ohms, in series with the input detector capacitance, appears to stabilize many IC devices! The writer has evaluated this and verified it to be very effective in "squelching" the unwanted feedback! To make the fix more versatile (for many devices), the writer has made the resistor variable, by using a subminiature trimpot of 1k ohms as shown in Figure 1. This was desirable, since some preliminary tests indicated that anywhere from 27 ohms to over 500 ohms may be needed for certain IC devices for an adequate measure of stability.
Figure 1: Detector section
RF: Typically 1-2 megaohm
Cin: Typically 2000uF
Rin: 1 k ohm subminiature trimpot
Use of a small resistance for the input detection device is also effective as a GW detection unit as was shown in Figure 6 of the R-E Electronic Experimenter Handbook article of January 1989. However, signals generated by a capacitative and a resistive detection unit are 180° out-of-phase in the output. Thus, any scalar-type signal developed in the output will also be 180° out-of-phase with the strong GW signal developed by the main input capacitance and thus the system will tend to "degenerate" rather than "regenerate" at the high input levels and thus not go into instabilities or oscillations. There would not be any loss of conversion gain in this process.
Preliminary tests by Bill Ramsay seem to indicate that the units can be driven to higher sensitivity and gain levels without noticeable (?) distortions present. A few tests by the writer appear to confirm this. Perhaps some of you more active researchers may want to try to confirm this also? The GW detectors may now be even more versatile and useful now with this simple revision --- thanks to Bill Ramsay.
Take care, and good experimenting. I remain,
With best regards,
Greg Hodowanec