On 26/03/2023 01:37, sor...@ayecka.com wrote:
I am working with Ettus B205 mini i. We operate it on Ubuntu + C++.
I have a few problems.
A. It seems the DC cancellation mechanism is quite aggressive. It
creates a notch at DC. I can overcome it by using lo_offset parameter,
but as I am trying to sample at the lowest symbol rate possible, it’s
probably not a very good option.
YOu can always just move the LO to the edge of your passband, where
there'll be roll-off anyway. Not sure I understand this
comment. The LO_OFFSET can be configured in the HARDWARE, so it
will be nearly-transparent to the rest of your processing.
B.There is very high LO leakage - even when the signal is at high power.
To demonstrate I use the Ettus command tx_waveforms with a waveform
set to SINE at 1MHz.
Rate is 5.76e6
Gain is 50 db
freq is 1e9.
You can see the LO is -10dbc and also image at -40 dbc. Both are much
too high.
A gain of 50dB on the B205 is about 32dB below the maximum RF gain.
Try two things:
Lower the baseband amplitude a bit when using tx_waveforms--perhaps 0.85
Crank the RF gain up a bit, that should improve the
carrier-to-LO-leakage ratio.
But, also, consider LO offset in the TX path as well.
The image performance may well improve with a not-saturated baseband.
Any suggestions please?
Thank you,
Sorin
_______________________________________________
USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-le...@lists.ettus.com
_______________________________________________
USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-le...@lists.ettus.com
