Managed to complete the above project for my K3. I really must recommend this panadapter, now after switching on the K3 the first thing I do is start Rocky and have a 'look' around on 10 and 6m for openings. On several occasions whilst tuning the K3 and hearing a flat band, I've had to reverse tune on seeing a spike jump up on the screen, and as a result completed a contact on what at first seemed like a flat band.
It all started when I came across the following website:
This page gives the details on how to purchase a K3 SoftRock V6.2 lite. In my e-mail to Tony Park's I mentioned that it was for use with the K3 IF and also the offset crystal frequency of 8.191MHz. He will also give you postage costs to your home location but the price of the Softrock is currently around $12. I paid using PayPal which was very convenient.
I also bought the Z10000U (general purpose) buffer amp board from Clifton Laboratories to help with reverse isolation between the SoftRock and the K3. I didn't try the softrock without the Z10000U in-line, but with it there is no evidence on the K3 rx of the panadapter being there. I used the 150 Ohm resistor for R907 to set the gain of the buffer amp to 6.7dB. I haven't experimented with different values as this seemed to work sufficiently for my needs. To this date I haven't bothered with the optional bandpass filter.
The build presented no problems except you have to be capable of working with surface mount components. All in all it took me 5-6 hours to get it all up and working with the K3. I then boxed both boards up in an Altoids tin.
With the tin open you can follow the signal path from the LHS BNC connector (connected to KXV3 IF out) through the Z10000U buffer amp (LHS card) then linked with coax to the SoftRock (RHS card) then finally out on a screened jack3.5mm stereo socket (top right) to the sound card.
Here we can see me writing this article with Rocky software running at the bottom of the screen. I'm listening on the K3 to the small spike shown at the centre of the display. I find it really useful to see what's happening round about where I'm listening to. I've seen huge spikes right next to me and not been able to tell they were there on the K3 :-).
My K3 has the following roofing filter options: 2.7/2.1/1/0.4 KHz.
When using the panadapter it is interesting to listen to the K3 audio output in cw mode and look at the IF output when switching between filters. My 2.7KHz filter has a specified shift of -0.93KHz which is corrected in firmware to correctly line up with the IF. On the panadapter it is clear that the IF actually shifts to correct this error, as all visible signals bounce to the right by approx 1.5KHz in comparison with the 400Hz roofing filter. My Inrad 2.1KHz (0KHz offset) roofing filter also causes signals to bounce to the right but this time by only 500Hz. There is no shift between the 1KHz and 400Hz roofing filters. Whilst switching between these roofing filters, there is no difference in pitch from the audio output. It would seem from these very rough observations that the IF is shifted up by approx 500Hz (plus any offset) when using the 2.7/2.1 filters in comparison to the 1/0.4 IF. I imagine this extra shift for wide filters brings the 700Hz pitch cw signal more to the centre of the filter passband i.e. 1.2KHz. Any thoughts?
The same shift happens on transmit and means that you cannot align your tx and rx on screen. I had thought about buying the INRAD 2.8KHz filter to lose the 0.93KHz offset, but I would still get the 500Hz shift so probably not worth it on cw.
The panadapter is very useful for the identification of noise sources. Without one it's very easy to connect up a new SMPS or Router in the shack and and after a quick tune around believe it to be noise free. With the panadapter it is instantly obvious that noise has appeared, so far I've had to assign a laptop power supply to the bin and can also see the squirmy warble associated with my 2 SMPSs (Alinco DM-330MW and MFJ 4125). Anyone who believes these to be noise free could benefit from a panadapter :-).