On 09/29/2020 02:02 PM, Mark Koenig wrote:
Yes, I understand gain varies with temperature and frequency, I just
wasnt sure if there was any receive chain analysis performed with the
daughtecard to give the developer an idea of what type of gain is
provided over the attenuation range at various frequencies. I am not
too concerned about tenths of dBs....I was just interested in what the
actual gain range provided by the card is.
Thanks
Mark
------------------------------------------------------------------------
If you look at the first page of this:
https://files.ettus.com/schematics/twinrx/TwinRX%20RF%20Board%20Rev%20D.pdf
You can see the overall block diagram. You can also see several PE43503
attenuators, sprinkled among several different MMIC amplifiers, and
various different RF pathways through switches and filters depending
on band. It would be hard for me to unwind all of that and give you
a definitive answer.
Even for the IF processing, there are two different IFs, depending on
the frequency band--again with various distributions of gain and
attenuation (either explicit attenuation, or attenuation via
filtering)--all of which have considerable uncertainty--due to
batch-to-batch
variability and temperature effects. I'm fairly sure that even the
designer of the board couldn't tell you, for any given board configuration
what the actual gain measured between the antenna input an the ADC
input actually was, with better than 5dB confidence. Which is where
calibration comes in.
https://files.ettus.com/schematics/twinrx/TwinRX%20IF%20Board%20Rev%20C.pdf
In a laboratory instrument, like a spectrum analyser, all of this is
painstakingly calibrated at the factory, usually using lookup tables (or the
analog-era equivalent), based on well-characterized calibration
sources. So when you set the gain level on the front-panel of the device
to some dB value, you'll actually get that value at the measurement
point and when you look at the measurement on the display and it
says -70dBm, it's actually -70dBm at the input terminal. SDRs aren't
that, typically. Although one could build a fairly nice lab instrument
*around* an SDR, using all the aforementioned calibration exercises, etc.
Now, this all, I admit, sounds a tad "lecturey". I know you probably
know all of this, but many on the list don't, or perhaps haven't thought
about
it much. So, I'm prompted to deliver this, or a very similar
"lecture" a few times a year due to similar queries to yours.
_______________________________________________
USRP-users mailing list
[email protected]
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
