Want to achieve the symbol rate to 270.833kb/s

hi all:
I want to make the symbol rate of the data which is transmited
by the USRP at the 270.833kb/s.I know the OpenBTS had already make it
come true. but what I want to know is that can I use the resampler block
to make this happen? (put the resampler block after the GMSK Mod block
)and the value of the interpolation and decimation of the resampler
block can be the same as them in the OpenBTS which are 65 and 96.and the
interpolation of the USRP is 320 which is make the sample rate of the
usrp is 400k.I have not do it now ,just want to ask here if it is
possible.

                                                                                                                                                                                                                 Thank 

you in advence

2010/9/15 intermilan [email protected]:

hi all:
I want to make the symbol rate of the data which is transmited by
the USRP at the 270.833kb/s.I know the OpenBTS had already make it come
true. but what I want to know is that can I use the resampler block to make
this happen? (put the resampler block after the GMSK Mod block )and the
value of the interpolation and decimation of the resampler block can be the
same as them in the OpenBTS which are 65 and 96.and the interpolation of the
USRP is 320 which is make the sample rate of the usrp is 400k.I have not do
it now ,just want to ask here if it is possible.

You should be able to use the gr.pfb_arb_resampler_ccf block to
accomplish this. It’s an arbitrary resampler and so doesn’t rely on
integer relationships. It can be a bit tricky to use, so look at the
examples in gnuradio-examples/src/python/pfb for help in designing the
filter for it.

… One of these days, I need to make a convenience wrapper for this
block that designs the filter for the half sample rate
automatically…

Tom

hi Tom:

  I can not find the examples what you asked me to look for.so what 

the version of your Gnuradio? Mine is 3.2.2.
Besides, I still wonder why I can not use the resampler block in
the /gnuradio-core/src/python/gnuradio/blks2impl in which I can set the
value of the interpolation and decimation.and the program will automatic
to set the parameters of the filter I need.I think it is capable. So can
you tell some specific reasons?

                                                                                                                                                                                      thank 

you

hi Tom:
thank you for your answer,I get it .Besides, how do we know
the symbol rate if the data in the air?what did you do to measure the
symbol rate?

                                                                                                                                                                                            Thank 

you in advance

The communication protocol that you are looking into should have a
predefined symbol rate. At the very least in the pilot sequence of a
frame.

–Colby

2010/9/18 intermilan [email protected]

This is part of the reason we used a 52 MHz clock for the public release
of the OpenBTS GSM stack. It makes 13/48 MHz rational. We would have
used 13 MHz, but the USRP-1 FPGA code fails to transmit for clock rates
below about 48 MHz.

If you use the stock 64 MHz clock you will need a polyphase resampler
for most telecom applications. You will also have wicked temperature
drift problems.

Sent from my Verizon Wireless BlackBerry

2010/9/17 intermilan [email protected]d:

thank you

You’ll need either GNU Radio 3.3.0 or the latest checkout from git.

Then yes, you can use the rational resampler. I misunderstood what you
were saying in your original email. I thought you were having problems
using the rational resampler to make the values work out properly. The
nice thing about the arbitrary resampler is that you don’t have to
find a valid set of integers for the interp/decim to get your sample
rate; instead, you pass it a real number to get out the sample rate
that you want.

But running through the numbers, it looks to me like you’ll get the
desired sample rate. So yes, that should work.

Tom

On Sep 25, 2010, at 10:41 AM, [email protected] wrote:

We would have used 13 MHz, but the USRP-1 FPGA code fails to transmit for clock rates below about 48 MHz.

Does anybody know why this is?


Mark J. Blair, NF6X [email protected]
Web page: http://www.nf6x.net/
GnuPG public key available from my web page.

On 10/08/2010 07:10 PM, Mark J. Blair wrote:

On Sep 25, 2010, at 10:41 AM, [email protected] wrote:

We would have used 13 MHz, but the USRP-1 FPGA code fails to transmit for clock rates below about 48 MHz.

Does anybody know why this is?

That is not correct. The inband signaling code that OpenBTS uses may
not work with those clock rates, but the standard code we developed
works just fine with any clock frequency. I have used it all the way
down to 10 MHz.

Matt

On Sat, Oct 9, 2010 at 11:25, Matt E. [email protected] wrote:

That is not correct. The inband signaling code that OpenBTS uses may not
work with those clock rates, but the standard code we developed works just
fine with any clock frequency. I have used it all the way down to 10 MHz.

I believe that it’s still doesn’t work as expected at low clock rates.
While FPGA code may support this, ADC or DAC (don’t recall exactly)
need a different configuration of capacitors/resistors - USRP use
default configuration from the datasheet, which is suitable for high
clock rates. That’s from my memories, so I may be wrong somewhere.


Regards,
Alexander C…
http://www.fairwaves.ru

This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.

| Privacy Policy | Terms of Service | Remote Ruby Jobs