Here is a picture that shows how to wire up the switch. Note the manual says to use an SPST switch - this is incorrect, use an SPDT switch, either ON-ON or ON-OFF-ON will do. In the OFF position, both the TX and RX frequencies will be changed by the pot, but they will both stay the same (so it's effectively OFF). Make sure you have removed C3 (0.01uF).
The NC20 TUNE Pot Page shows how to set the TUNE pot links correctly and which 10 turn pot to buy.
A VAST improvement in the NC20's AGC performance can be had with a fairly simple AGC mod.
The center frequency of the audio bandpass filter (U5b) was designed to be 700Hz, which is what I was led to believe most people loved and wanted. Since then, many people have asked for a lower frequency filter to match both their taste and the frequency of the TiCK's sidetone (which is about 600Hz). So, if you want to move the filter to 600Hz, make the following resistor value changes:
R53 = 3.9K (was 3.3K)
R54 = 7.5K (was 6.8K)
R55 = 270K (was 220K)
That's it, simple! I think the radio sounds much nicer with the filter at 600Hz. I changed the Red Hot Radio RH-20 and RH-40 kits to these values.
Larry East, W1HUE, came up with this delightfully simple mod for that cures the small but sometimes annoying amount of thermal drift in the AGC voltage (hence RX gain!). Thanks Larry!
The warm-up drift in the AGC zero-signal voltage can be essentially eliminated by a very simple mod: Insert a 1N914 (or 1N4148) in series with R63 to ground (the diode's cathode goes to ground). You can install the diode on top of the board (vertically) next to R63 - cut the lead from R63 to ground and insert the diode. Readjust VR5 for the proper zero-signal AGC voltage; if you can't get the AGC voltage high enough, replace R63 with a smaller value (680 Ohms worked for me). Before I made this mod, the AGC voltage drifted about 300mV during warm-up. After the mod, the drift was less than 10mV.
Here's another of the Larry East, W1HUE, fine mods that I recommend to NC20 purists everywhere. Thanks again Larry!
The no-signal AGC voltage is basically derived from the +12V buss via a voltage divider. It will therefore track any change in the supply voltage. This is not a problem if the supply voltage is relatively stable. However, it could become a problem if multiple power sources are used and not all supply the same voltage. If the AGC no-signal voltage is set when using a "normal voltage" power source and then a lower voltage supply is used, a noticeable drop in receiver gain can occur. Conversely, if a higher supply voltage is used, the AGC may not respond until input signals become rather large.
It would therefore seem advisable to supply the AGC amplifier (Q10 - Q13) with a regulated supply voltage. Some experimenting showed that a supply voltage of at least 9V is needed for the AGC amplifier to function properly. The 10.2V output of the "pre-regulator" described above is therefore just about right! It is a simple matter to use that to power the AGC amp.
To make the mod, proceed as follows:
Locate the PC trace that supplies +12V to the AGC amp. Looking at the bottom of the board with the tuning pot to your right, this trace is just to the right of Q13. Cut this trace between the point where R65 attaches and where it passes through the board. Solder a jumper wire between this trace (the solder pad for R68 or R61 is a convenient attachment point) and the output of the TK11609 (the solder pad for C12 is a convenient attachment point).
That's it! All that remains to be done is readjust VR5 as per instructions in the NC20 manual (or as described in Dave Fifield's AGC mod - which I highly recommend).
Larry has more NC20 mods on the Minnesota QRP Club's website here.