For those like me who love and still use long-in-the-tooth Rocky, here is a note of caution. The spectral purity of the the audio output seems to be less than optimal in some cases. I noted while attempting to use Rocky as a two tone signal source for IMD measurements that I was seeing what looked like inter-modulation products on a single frequency tone source. Since that cannot be, it led me to look at the audio output of the laptop on an audio spectrum analyzer and I noted what seems to be a problem.
On my laptop, if I am outputting a 5KHz tone, as I would be if I were transmitting CW 5 KHz above the center frequency, and I view the audio signal on the analyzer I see the third harmonic at 15 KHz only 13 dB down from the fundamental. So this means, for example, if I am transmitting CW at one watt on 7.005 then I am also transmitting at 50 mW on 7.015! Not good and not acceptable. This was using the internal Realtek HD audio card, so I changed to an iMic USB card and got exactly the same results. This indicates to me that the problem lies with the Rocky tone generation algorithm rather than the hardware.
For further confirmation I tried both the WSPR and HDSDR applications on the identical hardware configuration. WSPR produced audio harmonics that were 32 dB down from the main carrier which should be acceptable as modulation products. HDSDR did the best with harmonics that were down better than 60 dB.
One further note on Rocky.... a work around may be to operate close to the upper and lower edges of the passband so that the harmonics fall outside the sampling range and are presumably attenuated by the audio filters in the DAC. When I looked at operating 10 KHz from the center the harmonics fell of to 30 dB down.
Wondering if anyone else has looked at this issue?
Warren Allgyer - 9V1TD