As I mentioned in a previous post, I did the calculations to change the AF filter in the Willamette from a low-pass filter with a 3.3 kHz cutoff to a peaked low-pass filter with a cutoff frequency nearer to 1 kHz. I finally got around to implementing the mod last night and got a chance to listen to it on the air today and take some measurements of the filter response.
First off, let's take a look at the filter response:
If you compare this to the old response, it probably won't look drastically different, but it does cut off a bit eariler than the original filter. There is a bit of a peak as predicted, although it's a bit wider and shallower than expected.
However, the real proof is in the listening. I found (purely qualitatively) that the response of this filter was much tighter sounding than the original. Much of the high frequency interference is gone, and you can tell by tuning through a signal that it drops off much more quickly at the higher AF frequencies. You do lose some of the "crisp" direct conversion sound, but I feel like this is made up for in the utility of having greater filtering.
Here are the steps that you need to take in order to modify your own rig:
- Replace R50, R51, R54, and R55 with 24 kΩ
- Replace C55, C59 with 100 nF
- Remove C56, C57 then place a 1 nF capacitor from Q12 base to ground (in the place where C56 was located)
- Remove C60, C61 then place a 1 nF capacitor from Q13 base to ground (in the place where C60 was located)
One other small thing that you might want to do is replace C65 with a 1 uF capacitor. I noticed that when the AF gain control was set to maximum, that there would be an annoying popping during keying. This change helps to eliminate this problem.
I hope that you enjoy this modification to the rig. In hindsight, I'm not really sure why I designed such a wide open AF filter, although I suspect it was because I wanted to preserve the "DC" sound of the rig. However, I think that utility trumps a nicer sound in this case and will make the rig more usable overall.