In general, the transceiver switches the 4-element 1500 ohm xtal BPF ends between the inputs and outputs of the two SA602s to reverse the signal flow for R/T operation. Since no IF amplifier is used in the design, 20 dB of additional receiver gain is produced by the 2N2222 receiver RF amplifier, while automatic gain control (AGC) is produced by the peak DC swing of the LM386 output passed through a rectifier and filtered by a capacitor and fed to the gate of a BS170 enhancement mode FET acting as a variable resistor across the input of the LM386.
Both receive and transmit band pass filtering are done by the same half-pi BPF. The diode pair in the mic circuit reduce the "chirp" that occurs during the R/T transition. Additional BS170s could easily be used to mute both the mic and audio instead of the R/T switch directly. These BS170s would be controlled by the +R and +T voltages on their gates while their drains would be tied to 1) the mic circuit between the two coupling capacitors and 2) pin number 1 (audio in) of the LM386 (BS170 sources to ground). Additional power output (perhaps 60 mW) could also be attained by connecting the RF output transistor's collector choke (10 uH) to a 9 V supply instead of the 5 V. Additional biasing current might also be required for this change.
The R/T switches are from Radio Shack and are of the slider-type with silver plated brass connectors. I removed the detent springs from two of them and weakened the third. I then drilled holes through the switches' handles to connect them with a piece of tempered copper-plated steel wire. The switches are marked on their sides with the word "ALPHA." Low current-drain type DPDT relays (10 to 12 mA @ 12 V) could also be used in place of the switches. I used switches to reduce transmit current, and they work very smoothly.
The intent of this design was to minimize the parts count (about 85) and size, reduce power consumption (17 mA receive and 40 mA transmit), run on a 9 V battery, be self...