Compared to the original circuit, C2 and C3 have been divided by about eight, and C1 has been divided by about two, give or take values found in my junk box.
U1 – LMC567 PLL
R1a – 2.2 k
R1b – 5 k pot
R2 – 22 k
R3 – 220 ohm
R4 – 220 ohm
R5 – 390 k
C1 – 0.15 µF
C2 – 1 µF
C3 – 3.3 µF
C4 – 0.1 µF
C5 – 100 µF
C6 – 470 µF
In addition R3 and C6 have been added. They add a time constant of 0.1 sec or so, but unlike just increasing the output filter capacitor C3, this gives a fade-out when the LED extinguishes, and results in a smoothness to the display much like the indicator on the FT-1000. The value for R3 has been found from the maximum output current of 20 mA of U1, and is obtained if C6 has been charged to 5 V and is then discharged through U1’s output.
R5 has also been added for increased sensitivity. I found that my circuit required a little over comfortable speaker level to trigger before R5 was added. Values as low as 270 k can be used and give even higher sensitivity, but also more triggering on noise.
The output is connected to the right-most LED of the K2, an idea that I first heard from W7DZN. The output goes to pin 20 of the DS2 LED array on the Control Board. In the ‘Nite’ setting, the LED is fed from 2.7 V through a 120 ohm resistor, while in the ‘Day’ mode it is fed from 4 V. R4 results in less than 50% of the current in the LED than in the usual S-meter, and makes it easier to see that this segment has a different function.
Vcc is connected to 5V from the output of U8 on the RF board (across C90 on the underside of the board), and the audio input is connected to the output of the AF power amplifier. The circuit is mounted on the original N0SS/KR5L PCB.
The frequency can be adjusted to match the sidetone in two ways: By measuring the output of pin 5 with a frequency counter and divide the frequency by two, or by increasing the sidetone level in Test mode until it turns on the LED, and then tune R1b carefully back and forth until one has found the center between the low and high turn-on frequencies.
Reference: KR5L’s circuit/PCB: http://www.n0ss.net/kr5l_led_cw_tuning_ind.pdf
This article was printed in Sprat, Summer 2002 under the title "QRP Zero-beat Indicator for the Elecraft K2 (and other transceivers)" and in a German translation as "QRP-Zero-Beat-Indikator für den Elecraft K2" in QRP-Report, 3-2002 (Note that R3 was incorrectly labeled 220k).
Recently, I moved the zero-beat indicator to a board mounted on the back of the K2's front panel board. I also added an audio buffer for an AF output through the mic plug that is independent of the AF pot setting. The zero-beat indicator uses the same signal, and its indication is therefore independent of audio pot setting. See K2 Audio Output for details and new pdf-file with schematic.