On 07/13/2017 07:57 PM, Eugene Grayver via USRP-users wrote:
Hello,
I am trying to understand the ‘actual’ gain in TwinRX.
Experimentally, I see that a 0 dBm input requires gain of +46 dB to
give a full-scale (-1/1) output. Interestingly, the two channels have
about 3 dB of difference in gain (I would have expected a lot less).
Since the A/D expects about 0 dBm (based on the datasheet), it appears
we only have 50 dB of actual gain available. Is that correct? What
does the 95 dB of gain mean?
Thanks.
Eugene Grayver, Ph.D.
Aerospace Corp., Sr. Eng. Spec.
Tel: 310.336.1274
That's the gain-control range, distributed among (AFAIR) a couple of
different variable-gain stages.
The common-mode reference voltage on the X310 ADC (I assume you're
talking about an X310 motherboard?) is, AFAIR, set to 1.5V, which amounts to
about +7dBm. The ADC is 14 bits, so the smallest signal it can
resolve is about -73dBm. This is a very rough estimate.
In front of this, there are fixed and variable gain stages, including a
variable gain stage in the ADC itself. The mixer stage(s) may also
have a small
amount of conversion gain, I don't know off the top of my head.
But the 95dB figure is the total gain-control range, which is the sum of
the gain controls as distributed through the conversion and gain chain
on the card.
Since RF components will typically have actual delivered gain that may
vary from "nominal" by anywhere from 0.5 to 2dB in the same part, you cannot
reliably predict how much actual gain any given gain setting will
produce, even at the same frequency (RF parts also vary in gain with
frequency,
so for a widely-tunable card like TwinRX, you cannot assume that the
actual gain will be the same across all frequencies for the same gain
setting).
On 07/13/2017 07:57 PM, Eugene Grayver via USRP-users wrote:
>
> Hello,
>
> I am trying to understand the ‘actual’ gain in TwinRX.
> Experimentally, I see that a 0 dBm input requires gain of +46 dB to
> give a full-scale (-1/1) output. Interestingly, the two channels have
> about 3 dB of difference in gain (I would have expected a lot less).
> Since the A/D expects about 0 dBm (based on the datasheet), it appears
> we only have 50 dB of actual gain available. Is that correct? What
> does the 95 dB of gain mean?
>
> Thanks.
>
> _______________________
> Eugene Grayver, Ph.D.
> Aerospace Corp., Sr. Eng. Spec.
> Tel: 310.336.1274
> ________________________
>
>
That's the gain-control range, distributed among (AFAIR) a couple of
different variable-gain stages.
The common-mode reference voltage on the X310 ADC (I assume you're
talking about an X310 motherboard?) is, AFAIR, set to 1.5V, which amounts to
about +7dBm. The ADC is 14 bits, so the smallest signal it can
resolve is about -73dBm. This is a very rough estimate.
In front of this, there are fixed and variable gain stages, including a
variable gain stage in the ADC itself. The mixer stage(s) may also
have a small
amount of conversion gain, I don't know off the top of my head.
But the 95dB figure is the total gain-control range, which is the sum of
the gain controls as distributed through the conversion and gain chain
on the card.
Since RF components will typically have actual delivered gain that may
vary from "nominal" by anywhere from 0.5 to 2dB in the same part, you cannot
reliably predict how much actual gain any given gain *setting* will
produce, even at the same frequency (RF parts also vary in gain with
frequency,
so for a widely-tunable card like TwinRX, you cannot assume that the
*actual* gain will be the same across all frequencies for the same gain
setting).
