Good engineering dictates never running a square rise and fall waveform into a crystal filter.
The square rise-and-fall waveform will cause filter ringing.
The FT-1000(D) runs an essentially unshaped CW waveform into the filters, generating unnecessary keyclicks that are strongest from almost 1 kHz below to 2kHz above the transmit frequency.
Locate the seven pin connector J3024 at the middle front of the AF board. The AF board has audio, PTT, ALC, and key jacks as well as other connectors.
Cut the wire going to pin 2 (T CNTL). This is the second terminal from the right (front view). This is NOT the lead with 9 volts (nine volts is pin 6)!
Connect the collector of a 2N3904 to the flying harness lead, and the emitter towards the jack. Connect a 1.5 k resistor between the base and collector, and a 10 uF 25 capacitor from base to ground.
I forgot to say the resistor goes between the collector and the base of the transistor.
The transistor turns on slowly because the 10uF capacitor charges through the 1.5k resistor. The stage turns off slowly because the capacitor discharges through the base to the emitter.
This slows the waveform rise and fall times at Q3030 and D3009. The result is a normally sloped CW waveshape hitting the SSB filter.
This reduces clicks beyond 1 kHz above and below the transmitting frequency into the composite noise of the transmitter, about 50 dB reduction in clicks on my radio above my TX frequency.
Clicks 500 Hz away are reduced about 35 dB.
There is almost no change in "keying sound" on the transmitting frequency. Very little change in envelope shape, yet bandwidth changes are profound.