Empathy List Archives

J

johnhigginsgis@gmail.com

Wed, Nov 20, 2024 7:45 AM

I have few questions regarding B210.

Ettus recommends -15 dBm for the max input power. Is this power at the RF ports (i.e. RX-A/RX-B)?
Max input power for AD9361 is given as +2.5dBm. Is the difference due to some additional circuitry preceding the 9361 or is -15 dBm just a value Ettus recommends to be on the safe side?
If my signal is guaranteed to be below -15dBm, can I set the B210 gain to its maximum, i.e. 76dB safely (except maybe ADC clipping) without damaging anything?
Is the 76dB gain solely due to the AD9361 or is there a separate LNA on B210?
AD9361 is 12 bits, thus theoretically has 72dB dynamic range though in practice apparently it is around 60 dB. Although B210 is not calibrated after manufacturing what is the usual ballpark dBm level that corresponds to 0 dBFS level?

Thanks!

I have few questions regarding B210. 1) Ettus recommends -15 dBm for the max input power. Is this power at the RF ports (i.e. RX-A/RX-B)? 2) Max input power for AD9361 is given as +2.5dBm. Is the difference due to some additional circuitry preceding the 9361 or is -15 dBm just a value Ettus recommends to be on the safe side? 3) If my signal is guaranteed to be below -15dBm, can I set the B210 gain to its maximum, i.e. 76dB safely (except maybe ADC clipping) without damaging anything? 4) Is the 76dB gain solely due to the AD9361 or is there a separate LNA on B210? 5) AD9361 is 12 bits, thus theoretically has 72dB dynamic range though in practice apparently it is around 60 dB. Although B210 is not calibrated after manufacturing what is the usual ballpark dBm level that corresponds to 0 dBFS level?\ \ Thanks!

MM

Marcus Müller

Wed, Nov 20, 2024 10:02 AM

Hi John,

On 20.11.24 08:45, johnhigginsgis@gmail.com wrote:

I have few questions regarding B210.

Ettus recommends -15 dBm for the max input power. Is this power at the RF ports (i.e.
RX-A/RX-B)?

yes, note though that this is "numbingly loud" to the receivers, so you might not be able
to receive overly useful, but it's "still guaranteed to be damage free loud".

If my signal is guaranteed to be below -15dBm, can I set the B210 gain to its
maximum, i.e. 76dB safely (except maybe ADC clipping) without damaging anything?

yes, but you will definitely be clipping all the time if you get anywhere close to -15dBm.

Is the 76dB gain solely due to the AD9361 or is there a separate LNA on B210?

See the schematic on http://files.ettus.com: You'll find a set of switches (including one
to select the right band), a set of baluns (or rather, unbals) and then the AD9361

AD9361 is 12 bits, thus theoretically has 72dB dynamic range though in practice
apparently it is around 60 dB. Although B210 is not calibrated after manufacturing what
is the usual ballpark dBm level that corresponds to 0 dBFS level?

Depends on the gain you set, and the frequency you're working on. I wish I could give you
a simple number or even just a graph over gain, but it's necessarily a two-parameter
thing. You will have to measure.

At max gain, you'd expect full scale output to be achieved deeeeep in the negative dBm.

Best regards,
Marcus Müller

Hi John, On 20.11.24 08:45, johnhigginsgis@gmail.com wrote: > > I have few questions regarding B210. > > 1) Ettus recommends -15 dBm for the max input power. Is this power at the RF ports (i.e. > RX-A/RX-B)? > yes, note though that this is "numbingly loud" to the receivers, so you might not be able to receive overly useful, but it's "still guaranteed to be damage free loud". > > 3) If my signal is guaranteed to be below -15dBm, can I set the B210 gain to its > maximum, i.e. 76dB safely (except maybe ADC clipping) without damaging anything? > yes, but you will definitely be clipping all the time if you get anywhere close to -15dBm. > > 4) Is the 76dB gain solely due to the AD9361 or is there a separate LNA on B210? > See the schematic on http://files.ettus.com: You'll find a set of switches (including one to select the right band), a set of baluns (or rather, unbals) and then the AD9361 > > 5) AD9361 is 12 bits, thus theoretically has 72dB dynamic range though in practice > apparently it is around 60 dB. Although B210 is not calibrated after manufacturing what > is the usual ballpark dBm level that corresponds to 0 dBFS level? > Depends on the gain you set, and the frequency you're working on. I wish I could give you a simple number or even just a graph over gain, but it's necessarily a two-parameter thing. You will have to measure. At max gain, you'd expect full scale output to be achieved deeeeep in the negative dBm. Best regards, Marcus Müller

MD

Marcus D. Leech

Wed, Nov 20, 2024 3:53 PM

On 20/11/2024 05:02, Marcus Müller wrote:

Depends on the gain you set, and the frequency you're working on. I
wish I could give you a simple number or even just a graph over gain,
but it's necessarily a two-parameter thing. You will have to measure.

At max gain, you'd expect full scale output to be achieved deeeeep in
the negative dBm.

Best regards,
Marcus Müller

I'll augment what my colleague has said.

The amount of gain in the path will vary over frequency--perhaps by a
few dB. But also from device to device,
by 1-2dB.

The "gain" setting doesn't actually tell you anything about total RF
gain between antenna and ADC. It's a
"gain control" setting, which in RF paths is nearly-always
implemented with distributed attenuators. The
total gain ahead of the ADC may be more or less than 72dB. If the
total path is gain at "MAX" is 50dB,
then that 72dB gain-control range takes you down to a spot where you
have -22dB of gain.

If your goal is to estimate the power at the antenna port using a
strictly "numerical" approach based on the
gain setting, you are in a state of sin. You MUST calibrate over
your entire expected operating space,
including center frequency and sample rate, and at various gain
settings (although the gain setting should
be largely linear).

On 20/11/2024 05:02, Marcus Müller wrote: > Depends on the gain you set, and the frequency you're working on. I > wish I could give you a simple number or even just a graph over gain, > but it's necessarily a two-parameter thing. You will have to measure. > > At max gain, you'd expect full scale output to be achieved deeeeep in > the negative dBm. > > Best regards, > Marcus Müller I'll augment what my colleague has said. The amount of gain in the path will vary over frequency--perhaps by a few dB. But also from device to device, by 1-2dB. The "gain" setting doesn't actually tell you anything about total RF gain between antenna and ADC. It's a "gain control" setting, which in RF paths is nearly-always implemented with distributed attenuators. The total gain ahead of the ADC may be more or less than 72dB. If the total path is gain at "MAX" is 50dB, then that 72dB gain-control range takes you down to a spot where you have -22dB of gain. If your goal is to estimate the power at the antenna port using a strictly "numerical" approach based on the gain setting, you are in a state of sin. You MUST calibrate over your entire expected operating space, including center frequency and sample rate, and at various gain settings (although the gain setting should be largely linear).

usrp-users@lists.ettus.com

B210 power and gain levels