Have been testing a KrakenSDR with software on the Pi4 - which was likely more than 2 years old. After raising an Arrow 5 element Kraken antenna above my rooftop, I did a range of measurements.
All, except one troublesome direction was either spot on or +/- 1 degree.
Upgraded to the latest image - JAN 2025 I believe - and now I see results for the same sites ranging between -5 and +6 degrees.
I belive I had f/b averaging ON using the older image - but if I turn it on with the current image - results are way off - like 90 degrees …
Still have to old image on another memory card, so could roll back, but prefer a solution with the current image.
Did I just have a lucky day or two when doing the initial testing - or could there be something I am missing after the upgrade ?
Regards
Mike
Nothing has changed with the core DSP or DAQ algorithms in the updates. Are you still using the same MUSIC algorithm? F/B averaging is also only for a linear array, is that what you are using? It will produce poor results on a UCA array.
I think things have improved - even if I only play around a bit with array offset and gain.
I am using the MUSIC algorithm - and vary between 1 or 5 VFO’s for output - currently 1.
Trying not to use more gain that required, as i think overloading the receiver may lead to bad results.
Looks like I am mainly 1-4 degrees off at the moment, many within 1-2 degrees.
Sounds odd, but feels like system is self calibrating.
One big wish: A memory manger, allowing me to load pre-defined frequencies, mainly for testing accuracy against known transmitters.
And for Arrow - will contact them - make the Kraken antenna with telescopic arms - smaller diameter square rod sliding in a bigger one - so you can configure between VHF and UHF on the fly.
Regards
Mike