I do not claim ownership of the information which follows. I have had this information since the late 90's. I can't remember where I got it (possibly off packet radio) or who was the original author.
I have made the modifications described below to my 2010 and, at an ODXA DX Camp, to a few other 2010's. The improvement is definitely noticeable.
Some Audio Section Mods for the Sony ICF-2010 This paper describes some very simple modifications to the audio section of the Sony ICF-2010 (2001D). VERY IMPORTANT 1) Disclaimer notice: Do all of this at your own risk. Soldering iron burns, electrical shocks, frying the receiver, or any other surprises are your problem, not mine. These mods will obviously blow the warranty. 2) Make sure you really know what you're doing and can deal with surface-mount components and dense PC board soldering. Follow all of the normal anti-static practices. 3) This work can only be done with the full Sony service manual in hand. It diagrams the circuits being modified and is necessary for physically locating certain components on the PC board. THE PROBLEMS In the WIDE filter mode, the 2010 passes more high frequency audio than is useful. This mode is really only practical for very clear channel listening. At the same time the NARROW filter mode is really too muffled to be useful. It is somewhat better if the receiver is tuned off-carrier by about 1 khz, but that is beyond the tracking range of the sync detector. THE CIRCUIT Q19 and related components form a simple low pass filter labeled "AF LPF" in the diagram. In the NARROW mode, Q19 is biased on and becomes an emitter follower. In WIDE mode, Q19 is biased off. Q20 is simple emitter follower like Q19, but is not a low pass filter. It is turned on in the WIDE mode and off in the NARROW mode. THE SOLUTION The fundamental ideas are: - Use the LPF in the WIDE mode, cutting out some of the excessive high frequencies. - Eliminate the LPF in the NARROW mode, reducing the muffled audio. - Also in the NARROW mode, reduce some of the low frequencies to reduce the boominess. This is all very easy. Q19 and Q20 are switched on and off through 10K resistors R194 and R197 from signals at IC13. We're simply going to swap ends of these resistors. The original parts are surface-mount and are replaced with normal components with leads. Remove and discard R197 and R194. Replace them with 1/4 or 1/8 watt 10k resistors, but swap the switching signals. Now the LPF is used in the WIDE mode and not the NARROW mode. To reduce the bass response in the NARROW mode, remove and discard C171 (0.01uf). Replace it with a small 0.001uf capacitor. C171 couples the signal into Q20, and by reducing it, forms a simple high-pass filter. Other values were tried, but this seems to be the best overall compromise. TONE CONTROL One additional modification is useful. The three-position tone control could use more high-frequency cut in the "NEWS" position. Connect a 0.1uf cap in parallel with the existing C93 (0.068uf). RESULTS These changes made the 2010 MUCH more useful to me. I used the following as a very subjective metric: In school I studied Spanish, but am now very rusty. Occasionally I tune into some Spanish broadcast and try to follow the announcer. Previously in the NARROW mode I couldn't even pickup enough of the sound of a word to look it up in a dictionary. Now I never have that problem. Because I can pick out syllables much more easily, I recognize many more words. English also, is easier. The new WIDE position response reduces adjacent station interference somewhat and is much more useful for shortwave listening. Previously it was really only good for the AM broadcast band. DESIGN TRADEOFFS The only useful thing the LPF did for the NARROW position was to reduce some high-frequency noise. With these modifications, the noise is more noticeable with very weak signals. But the more effective tone control "NEWS" position reduces the noise, but without the original muffled sound. Since the WIDE mode now has a LPF in it, "hi-fi" AM listening is less hi-fi. FM sound is unaffected.