Re: [Discuss-gnuradio] High pass frequency response with LFRX boards

[Marcus D. Leech]

On 11/04/2010 07:24 PM, Drew Read wrote:
> Hi All,
>
> We're seeing pretty much the same thing on the USRP1 as this:
> http://www.ruby-forum.com/topic/206912#900641
>
> It seems that there is a very significant roll-off at lower
> frequencies on the receive side which pretty much makes it unusable
> for our baseband signals.
> The aformentioned forum post suggests that it is not caused by the
> LFRX board and so is likely to be something in the fpga...
>
> Just to reiterate the questions from this forum post.
> 1) What on the FPGA is causing this frequency response? (and what can
> we do about it?)
>
> 2) Why does this happen only in the RX path?
>
> And the original poster also wanted to know:
> 3) According to
>      http://gnuradio.org/redmine/wiki/1/USRP2GenFAQ   reference:
>      "How can I use both A/D converters on the USRP2?"
>    feeding the signal to RX_B does not give ANY output on the ADC
>
> If anyone can offer any help at all with this it'd be much appreciated.
>
> Thanks,
> Drew
>
>   
I can't speak for your flowgraph, but the referenced flow-graph in your
post is mis-constructed
  in a couple of difference ways.

The USRP2 input decimation is set to '101' -- odd decimations in USRP2
are *not recommended*.
  Plus, neither the SCOPE sink nor the FFT sink have been told the
correct sample rate, so lord
  knows what they're actually showing in terms of frequency response.

In USRP2, even decimation, preferrably a multiple of 4.

I've noticed no low-frequency rolloff in my applications with either the
USRP1 or USRP2.


-- 
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org