That's better than the earlier trace. It's pretty good, and could be improved further with an extra stage of filtering. The other exciter at
https://www.ebay.co.uk/itm/165722880409 ... SwGm1gJktZ can be made OK by use of an improved output filter too. I believe that there's a board redesign on the way for that one, that includes an improved filter and a couple of additional components to make the 5V regulator run cooler.
These two designs are reasonably good, simple, reliable circuits and have been widely used. However things that I don't like (with either of these two exciters) are:
- That horrible interstage 4:1 transformer
- The use of a PIC to programme the PLL IC (they're getting expensive)
- Running the VCO at the output frequency (makes RF feedback likely unless you're very careful with screening and positioning of parts)
- No attempt made to make the deviation sensitivity the same at all parts of the band
I find that doubler exciters (like Stephen Moss' Kallitron self-doubling oscillator) make for a cleaner output signal. You need a good highpass filter at the output of the oscillator to prevent
½f breakthrough. I used to use the "Soshin" bandpass filter modules for this purpose - they were cheap and worked well. With good design, the subsequent amplifier stages can be both fully broadband (requiring no adjustment at all) and very clean.
The next iteration of the PLL PRO series was to include one of these Soshin filters before the second RF stage, was going to eliminate the trimmer in the oscillator by using diode-switched capacitors, and add three more varicap diodes to guarantee exactly the same deviation sensitivity at all frequencies. This which would allow the inclusion of a fixed clipper stage to absolutely prevent over-modulation by level clamping!