scateu / kalibrate-hackrf

kalibrate for hackrf
BSD 2-Clause "Simplified" License
267 stars 78 forks source link

Wrong ARFCN and frequencies -- is `-E` needed? #24

Open pbraun9 opened 5 years ago

pbraun9 commented 5 years ago

Hi

As far as I experienced, the results provided by kalibrate-hackrf -s are always wrong (at least on recent host/firmware/CPLD). I get wrong ARFCN numbers and their advertised center frequencies correction is way to high (and on top of that, are contradictory, sometimes positive and sometimes negative), for example

    chan:   20 (939.0MHz + 17.102kHz)   power:  579119.98
    chan:   22 (939.4MHz - 32.928kHz)   power:  617854.92
    chan:   45 (944.0MHz + 17.110kHz)   power:  796347.98

Similar behaviour can be seen on various online tutorials here and there. The frequency correction does not make any sense on those reports either.

This issue is very much related to #16 and #23. As for the latter, I had indeed to go up to gains 32/32 to actually get some output, which should not be necessary as the BTS signals are supposedly strong enough.

Maybe this is all about HackRF One's DC spike? Is this what -E manual frequency offset in hz is made for? If so, how far away am I supposed to peak around? In what direction, -200Khz +200Khz? In short, how to use -E?

Thanks update: and this is a duplicate of #6.