K2 Anti-VOX Schematic



The circuit is a double sided PCB 2" x 1.4" with the top surface as a groundplane. As there are no plated thru holes components with a ground connection MUST be soldered on the top surface of the board. The board is edge glued with Super glue onto the KSB2 (see photos).

The groundplane is connected to the nearest part of the KSB2 groundplane. The +6V supply is derived from U1 Pin 20 (VDD). The track that connects U3 Pin 7 and the net C20, COMP OUT, U4 Pins 2 and 5) is cut just to the right of U3 on the KSB2 top surface. This isolates the VOX comparator and U1 COMP OUT from U3. The PCB signal V_COMP is connected by a short length of wire to U3 Pin 7. The PCB signal COMP_OUT is connected to U4 Pin 5.

I drilled a small hole in the RF board PCB just in front of the loudspeaker connector P5. A small connector pin was inserted and soldered onto the loudspeaker audio side of the connector. A corresponding socket was soldered onto a flying lead that goes to LS_IN on the PCB (Grey wire in the photos).

VERY IMPORTANT: To prevent any possibility of the PCB shorting to the case of Q21 on the RF board a small piece of insulation such as perf board should be glued to the top of Q21 case. The transistor case is at +12V so any short would be very damaging.


Loudspeaker audio is fed into the circuit via an Anti-VOX gain pot. This is the only adjustment on the board. The initial position should be fully anti-clockwise and then it should be turned clockwise until the VOX just stops triggering on strong loudspeaker audio. If it is turned up too much you will have to shout into the microphone to activate the VOX.

The High Pass filter removes signals below about 800 Hz that would cause the Anti-VOX to pump. The Anti VOX Detector IC1C/D1 acts as a precision rectifier outputting a positive voltage in proportion to the AC input waveform. It has a gain of approximately 10. This gain ensures that there is sufficient DC to charge C12. The combination of C12/R14 gives a reasonable discharge time, although you may wish to experiment with this.

The FET Attenuator is based around a FET that I happened to have in my junk box. Any N-Channel FET should do, although you may have to adjust the values of either R11 or R15 to get the FET to just start conducting with +1v DC on it's gate. The attenuator only has to slightly reduce the audio into COMP_OUT on receive to be effective. It should introduce no DC step as the Drain and Gate are at a similar potential to V_COMP.

On selection of VOX operation the DC level of U3 is measured by U1 COMP OUT ADC input and U2 DAC sets the ALC THR above the COMP OUT voltage by either 75mV for SSBA=1&2 or 50mV for SSBA=3. This gives the VOX threshold. The DC conditions for the VOX comparator U4 are maintained through R5.

A half rail supply for IC1B + input is derived from the filtered +6V by R2 and R3.

Monitor test points are provided. TP1 for the rectified loudspeaker audio and TP2 for the attenuated compressor audio. Using a scope TP1 will show a jagged waveform rising and falling in time with the received audio. If a sine wave is fed into the microphone audio it should be possible to see the attenuator action on TP2 as the Anti-VOX input signal is increased.


Ver 7.0 - Unused amplifier inputs tied to prevent RF pick up - 16/09/04 - Jack WA4FIB

Ver 7.1 - C7 must be a Tantalum capacitor - reduced leakage - 02/12/04 - Stewart G3RXQ

Except as noted, this entire site has Copyright 2010 - G3RXQ