This is a speech processor circuit that can be used in communication radio circuit. This circuit consist of two sections, they are amplification and compression or dipping. The combination of these processing will improve the  SSB communication intelligibility and provide higher average level while avoiding clipping. The 100K pot is used to adjust the compression degree, as well as adjusting the input. Here is the schematic diagram of the circuit:

Speech Processor circuit schematic

The compression principle is based on the diode’s equivalent resistance characteristic of CR2, which is inversely proportional to the forward current passing the junction. CR2 and 33K resistor acts as a voltage divider that attenuate the signal as he diode current increases because the CR2 resistance decreases. CR1 is used to rectify the signal to get DC voltage level that is further filtered by a 50uF capacitor, this voltage is then control the CR2 attenuation. Since there will be more attenuation for the bigger signal, the effect is the stabilization of the output. You can use any type of small signal high gain transistor for this circuit, and small signal silicon diode for CR1 and CR2 (such as 1N4148).