Second build sold out. Message will appear here when store is ready for third build ordering.

Feature request - I/Q spectrum reversal option

Hi John,

I'm currently helping some friends who are keen uWave operators, to help develop a remote hilltop radio site.

One of the things I'm contemplating is to use the KiWi in conjunction with a uWave down-convertor, but the only problem is the choice of local oscillator means that the RF spectrum is 'reversed'.

I wondered if, in addition to the "Frequency scale offset (kHz, 1 Hz resolution)" option in the Admin / Config tab, if it would be possible to have the option to 'reverse' the spectrum, possibly by swapping around the I/Q signals ?

This would also facilitate reception of 33.3 to 66.6MHz (spectrum reversed) by a hardware modification of the input filter, or by completely by-passing it and using an external filter, which is probably the easiest to implement (and also reverse if required).

With propagation conditions on the HF bands improving dramatically, I'm now hearing quite a lot of activity on the low VHF bands, and it would be great to be able to add a further KiWi to extend the frequency coverage at an existing site, without the need for an external converter.

Regards,

Martin

Comments

  • jksjks
    edited August 2022

    I think this can be done fairly easily (it has been requested before).

    Rather than reversing the I/Q of the DDS going into the mixer I think I would instead simply reverse the output bins of the waterfall FFT. Otherwise reversing the DDS I/Q makes USB/LSB in the audio channels reverse and probably other problems.

    I like this page to remind me how high-side/low-side injection works: https://www.rfcafe.com/references/electrical/spectral-inv.htm

  • jksjks
    edited August 2022

    Well, the above is not quite right. The waterfall and audio channels each have their own DDC, hence separate DDS(NCO) for each. So I suppose you could reverse the I&Q for the waterfall.

    But it turns out the waterfall code on the Beagle has an FFT-to-pixel-bin mapping function since the size of those two things are never the same. So a adding a little bit of code there gives a spectrum inversion at basically no cost.

    I have this working now and it will be in the next release.

  • Hi John,

    Thanks very much for this, it will lend itself to all sorts of experiments.

    On point I should have made clearer, as the whole spectrum is reversed, the signals also become 'mirror' images, so USB becomes LSB etc. Ideally the demodulation modes would need to be reversed too, which is why I thought that the I/Q would have to be flipped somehow.

    Regards,

    Martin

  • Quite right Martin. And I have tested a reversal to the audio path as well.

  • Superb :-)

    Thanks John.

    Regards,

    Martin

Sign In or Register to comment.