Re: bug with gain setting in usrp1?

GNU Radio
1

Hmm,

I don’t think so:

What I apply to board A ADC0 shows up on scope0, and what I apply to
board A ADC1 shows up on scope1, so there are definitely two different
ADCs involved.
Looking at the scope traces they are connected as set by the line
self.uhd_usrp_source_0.set_subdev_spec(“A:A A:B”, 0)
i.e:
DDC-channel0, input I from board A, ADC0
DDC-channel0, input I from board A, ADC1

Both Q-inputs to GND

It appears, that the gain settings are somehow skewed up.
I am sure that the gain for both ADCs attached to same LFRX-board got be
set individually.
Do I need a special FPGA code?

Wolfgang

Those are two DDC channels, derived from the same physical, analog, hardware and
ADC.

The LFRX itself has no gain-setting device, but the ADC does, so when you set the
gain on the ADC, you’re setting the gain for both DDC-derived channels.

2

On 05/29/2013 11:53 PM, Wolfgang Buesser wrote:

Both Q-inputs to GND

It appears, that the gain settings are somehow skewed up.
I am sure that the gain for both ADCs attached to same LFRX-board got be set
individually.
Do I need a special FPGA code?

The good news is that there is already software control for this. The
bad news is, there isnt an API to expose this edge case of ADC as two
unrelated streams with individual gain. See:

-josh