usrp-users@lists.ettus.com

Discussion and technical support related to USRP, UHD, RFNoC

View all threads

Re: [USRP-users] [Discuss-gnuradio] USRP source gain vs osmocom source gain?

MD
Marcus D. Leech
Tue, Jun 11, 2019 10:51 PM

On 06/11/2019 06:33 PM, Eamon Heaney wrote:

Hey all,

I'm trying to turn the wifi receiver flowgraph in this repo
(https://github.com/bastibl/gr-ieee802-11/blob/maint-3.8/examples/wifi_rx.grc)
into one that will work with the HackRF.

The example uses a USRP source, with a normalized gain of .75. I'm not
sure how to translate that into parameters for the osmocom source,
which has 3 gain options: RF gain, IF gain, and BB gain, all in dB.

I tried setting the RF gain = 20log(.75) = -2.49877, but that just
resulted in mangled packets.

Any suggestions?

--

Different hardware has different controls -- the whole "normalized gain"
thing is a USRP concept that isn't replicated in other hardware as far
as I know.  Now, "normalized gain" just means normalized relative to
the maximum gain setting of the underlying hardware, so your
conversion to dB is completely wrong.  In a "normalized gain"
scenario, 1.0 represents the maximum gain setting, and 0 represents the
minimum gain setting delivered by the actual hardware.  Various
USRP hardware types have various gain-control ranges--the
"normalized gain" concept is just UHDs way to providing an
abstraction away from having to know the valid range for the hardware at
hand.

Just like frequency, sample-rate, and analog bandwidth are
hardware-specific, so too is gain.

Usually, the RF gain control will be implemented by most hardware
supported by osmocom, and some will also implement IF and BB gain
controls.

End of the day, you will have to deal with this for each hardware type.

On 06/11/2019 06:33 PM, Eamon Heaney wrote: > Hey all, > > I'm trying to turn the wifi receiver flowgraph in this repo > (https://github.com/bastibl/gr-ieee802-11/blob/maint-3.8/examples/wifi_rx.grc) > into one that will work with the HackRF. > > The example uses a USRP source, with a normalized gain of .75. I'm not > sure how to translate that into parameters for the osmocom source, > which has 3 gain options: RF gain, IF gain, and BB gain, all in dB. > > I tried setting the RF gain = 20log(.75) = -2.49877, but that just > resulted in mangled packets. > > Any suggestions? > > -- Different hardware has different controls -- the whole "normalized gain" thing is a USRP concept that isn't replicated in other hardware as far as I know. Now, "normalized gain" just means normalized relative to the maximum gain setting of the underlying hardware, so your conversion to dB is completely wrong. In a "normalized gain" scenario, 1.0 represents the maximum gain setting, and 0 represents the minimum gain setting delivered by the *actual hardware*. Various USRP hardware types have various gain-control ranges--the "normalized gain" concept is just UHDs way to providing an abstraction away from having to know the valid range for the hardware at hand. Just like frequency, sample-rate, and analog bandwidth are hardware-specific, so too is gain. Usually, the RF gain control will be implemented by most hardware supported by osmocom, and some will also implement IF and BB gain controls. End of the day, you will have to deal with this for each hardware type.