The interesting discussions on this site provoked me to try the experiment with two, fairly identical, Gunn K-band oscillators operating at 24.125 GHz. Individual power outputs were +18.5 and +19.5 dBm or 0.071W and 0.089W, respectively. By adding E/H tuners between the oscillator and precision attenuator used to measure output power,output powers could be increased to 0.118W and 0.148W respectively, an increase of 2.2 dB in each case, and total power of 0.265W.
Next the oscillators wee combined with a matched magic-T. A a Gunn oscillator, followed by E/H-tuner was connected to each end port. A termination was placed in the top (out of phase) port, The in-phase side port was connected to another E/H-tuner, followed by an HP K-382A precision calibrated attenuator, HP K-486A thermistor mount, and HP 432A power meter. Adjusting the three E/H tuners for maximum, a power output power of +24.25 or 0.266W was attained. Thus no power gain was observed by combining with the magic-T. In summary, the sum of the Gunn oscillator powers, operated individually without E/H tuners was 0.160W. Adding E/H-tuners increased their total output to 0.265W. When the oscillators were combined with a magic-T and three E/H-tuners, total power was 0.266W, virtually the same as the sum of individual power outputs. Adjustment of the E/H-tuners between oscillators and the magic-T was critical as the oscillators tended to drop out of oscillation near the maximum power adjustment settings. It appeared the tuners were resonating the output irises of the oscillators so as to circumvent their isolation function and drain more power from the cavities. This was discussed in an earlier message from John C Roos, K6IQL. Conclusion: Gunn oscillators are necessarily very inefficient devices. Their stable operation depends on an output iris to limit output power so as to achieve a balance that provides for stable operation. The designer's intentions can be overcome by resonating the output iris so as to overcome the limitation and increase power and efficiency at the expense of stability. Bruce, KG6OJI _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
