IG
Isaac Gerg
Fri, Jan 28, 2011 7:11 PM
Just saw this from here:
http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
"It also supports multiple channels per USRP (though not possible with
the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx and Tx
boards will not work until the hardware is modified?
Isaac
Just saw this from here:
http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
"It also supports multiple channels per USRP (though not possible with
the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx and Tx
boards will not work until the hardware is modified?
Isaac
NF
Nick Foster
Fri, Jan 28, 2011 7:46 PM
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a sinusoid with a
frequency high enough to get through the BasicRX/TX transformers? It has
to be at least 100kHz. I'm not sure where you're specifying the
frequency of the complex vector you generate. I'm also not sure what
baseband frequency you're setting the USRP to when you initialize it. I
have a feeling the pulses you're seeing are DC filtered through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
> Just saw this from here:
> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
> "It also supports multiple channels per USRP (though not possible with the hardware yet) and looks much like the
> single-USRP API."
>
> Does this mean a loopback using a single USRP with a BasicRx and Tx boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a sinusoid with a
frequency high enough to get through the BasicRX/TX transformers? It has
to be at least 100kHz. I'm not sure where you're specifying the
frequency of the complex vector you generate. I'm also not sure what
baseband frequency you're setting the USRP to when you initialize it. I
have a feeling the pulses you're seeing are DC filtered through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
> Isaac
> _______________________________________________
> USRP-users mailing list
> USRP-users@lists.ettus.com
> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
IG
Isaac Gerg
Fri, Jan 28, 2011 7:58 PM
Hi Nick,
I am generating 2500 samples which compose one full cycle of a complex
exponential. I am using fc=1e6. So to address your first comment: No, Im
not sure if my signal is making it through. I am under the impression that
whatever samples I send out to the BasicTX will be received at the BasicRX
(with low-pass filtering) and that I dont have modulate anything (the board
does that for me). Is this not true? (I understand there will
be attention, carrier offset, etc)
I am working on a project which is creating a new waveform. I originally
thought about using the GNUradio framework (I have used it in the past
successfully), but decided to use the UHD since it appeared flexible and
simple for the prototype I am building (we have multiple threads and
eventually seek to write a device driver to communicate between nodes and I
dont want the extra overhead of the GNURadio framework). Additionally, I
have written a simple transceiver class which functions very similar to what
the GNURadio framework would be giving me anyways (as far as an entry point
to get the data with) without all the hassle of dealing with python, blocks,
etc especially if we decide to do GPU processing in the future (which is
likely).
Can you recommend a waveform that I can send that you believe I should be
able to clearly observe (fc=?, etc). It sounded like if I generated a
100kHz waveform and sent it, that it should get through... am I correct on
this?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
"It also supports multiple channels per USRP (though not possible with
the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx and Tx
boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a sinusoid with a
frequency high enough to get through the BasicRX/TX transformers? It has
to be at least 100kHz. I'm not sure where you're specifying the
frequency of the complex vector you generate. I'm also not sure what
baseband frequency you're setting the USRP to when you initialize it. I
have a feeling the pulses you're seeing are DC filtered through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
Hi Nick,
I am generating 2500 samples which compose one full cycle of a complex
exponential. I am using fc=1e6. So to address your first comment: No, Im
not sure if my signal is making it through. I am under the impression that
whatever samples I send out to the BasicTX will be received at the BasicRX
(with low-pass filtering) and that I dont have modulate anything (the board
does that for me). Is this not true? (I understand there will
be attention, carrier offset, etc)
I am working on a project which is creating a new waveform. I originally
thought about using the GNUradio framework (I have used it in the past
successfully), but decided to use the UHD since it appeared flexible and
simple for the prototype I am building (we have multiple threads and
eventually seek to write a device driver to communicate between nodes and I
dont want the extra overhead of the GNURadio framework). Additionally, I
have written a simple transceiver class which functions very similar to what
the GNURadio framework would be giving me anyways (as far as an entry point
to get the data with) without all the hassle of dealing with python, blocks,
etc especially if we decide to do GPU processing in the future (which is
likely).
Can you recommend a waveform that I can send that you believe I should be
able to clearly observe (fc=?, etc). It sounded like if I generated a
100kHz waveform and sent it, that it should get through... am I correct on
this?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster <nick@ettus.com> wrote:
> On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
> > Just saw this from here:
> >
> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
> > "It also supports multiple channels per USRP (though not possible with
> the hardware yet) and looks much like the
> > single-USRP API."
> >
> > Does this mean a loopback using a single USRP with a BasicRx and Tx
> boards will not work until the hardware is modified?
>
> Loopback is supported. You should be able to send and receive
> simultaneously just fine. Are you sure you're sending a sinusoid with a
> frequency high enough to get through the BasicRX/TX transformers? It has
> to be at least 100kHz. I'm not sure where you're specifying the
> frequency of the complex vector you generate. I'm also not sure what
> baseband frequency you're setting the USRP to when you initialize it. I
> have a feeling the pulses you're seeing are DC filtered through the
> BasicRX/TX transformers.
>
> Incidentally, you might find it easier to work within Gnuradio instead
> of manually generating the waveforms in C++. You can also use
> gnuradio-companion to put together a loopback test in minutes.
>
> --n
>
> > Isaac
> > _______________________________________________
> > USRP-users mailing list
> > USRP-users@lists.ettus.com
> > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>
>
>
IG
Isaac Gerg
Fri, Jan 28, 2011 8:00 PM
Please correct me if Im wrong, but even despite the DC filtering, I should
at least see contiguous energy thats the length of the burst I am sending.
Right? (Because Im not seeing this per my previous post).
Isaac
On Fri, Jan 28, 2011 at 2:58 PM, Isaac Gerg isaac.gerg@gergltd.com wrote:
Hi Nick,
I am generating 2500 samples which compose one full cycle of a complex
exponential. I am using fc=1e6. So to address your first comment: No, Im
not sure if my signal is making it through. I am under the impression that
whatever samples I send out to the BasicTX will be received at the BasicRX
(with low-pass filtering) and that I dont have modulate anything (the board
does that for me). Is this not true? (I understand there will
be attention, carrier offset, etc)
I am working on a project which is creating a new waveform. I originally
thought about using the GNUradio framework (I have used it in the past
successfully), but decided to use the UHD since it appeared flexible and
simple for the prototype I am building (we have multiple threads and
eventually seek to write a device driver to communicate between nodes and I
dont want the extra overhead of the GNURadio framework). Additionally, I
have written a simple transceiver class which functions very similar to what
the GNURadio framework would be giving me anyways (as far as an entry point
to get the data with) without all the hassle of dealing with python, blocks,
etc especially if we decide to do GPU processing in the future (which is
likely).
Can you recommend a waveform that I can send that you believe I should be
able to clearly observe (fc=?, etc). It sounded like if I generated a
100kHz waveform and sent it, that it should get through... am I correct on
this?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
"It also supports multiple channels per USRP (though not possible with
the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx and Tx
boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a sinusoid with a
frequency high enough to get through the BasicRX/TX transformers? It has
to be at least 100kHz. I'm not sure where you're specifying the
frequency of the complex vector you generate. I'm also not sure what
baseband frequency you're setting the USRP to when you initialize it. I
have a feeling the pulses you're seeing are DC filtered through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
Please correct me if Im wrong, but even despite the DC filtering, I should
at least see contiguous energy thats the length of the burst I am sending.
Right? (Because Im not seeing this per my previous post).
Isaac
On Fri, Jan 28, 2011 at 2:58 PM, Isaac Gerg <isaac.gerg@gergltd.com> wrote:
> Hi Nick,
>
> I am generating 2500 samples which compose one full cycle of a complex
> exponential. I am using fc=1e6. So to address your first comment: No, Im
> not sure if my signal is making it through. I am under the impression that
> whatever samples I send out to the BasicTX will be received at the BasicRX
> (with low-pass filtering) and that I dont have modulate anything (the board
> does that for me). Is this not true? (I understand there will
> be attention, carrier offset, etc)
>
> I am working on a project which is creating a new waveform. I originally
> thought about using the GNUradio framework (I have used it in the past
> successfully), but decided to use the UHD since it appeared flexible and
> simple for the prototype I am building (we have multiple threads and
> eventually seek to write a device driver to communicate between nodes and I
> dont want the extra overhead of the GNURadio framework). Additionally, I
> have written a simple transceiver class which functions very similar to what
> the GNURadio framework would be giving me anyways (as far as an entry point
> to get the data with) without all the hassle of dealing with python, blocks,
> etc especially if we decide to do GPU processing in the future (which is
> likely).
>
> Can you recommend a waveform that I can send that you believe I should be
> able to clearly observe (fc=?, etc). It sounded like if I generated a
> 100kHz waveform and sent it, that it should get through... am I correct on
> this?
>
> Thanks,
> Isaac
>
>
>
> On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster <nick@ettus.com> wrote:
>
>> On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
>> > Just saw this from here:
>> >
>> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
>> > "It also supports multiple channels per USRP (though not possible with
>> the hardware yet) and looks much like the
>> > single-USRP API."
>> >
>> > Does this mean a loopback using a single USRP with a BasicRx and Tx
>> boards will not work until the hardware is modified?
>>
>> Loopback is supported. You should be able to send and receive
>> simultaneously just fine. Are you sure you're sending a sinusoid with a
>> frequency high enough to get through the BasicRX/TX transformers? It has
>> to be at least 100kHz. I'm not sure where you're specifying the
>> frequency of the complex vector you generate. I'm also not sure what
>> baseband frequency you're setting the USRP to when you initialize it. I
>> have a feeling the pulses you're seeing are DC filtered through the
>> BasicRX/TX transformers.
>>
>> Incidentally, you might find it easier to work within Gnuradio instead
>> of manually generating the waveforms in C++. You can also use
>> gnuradio-companion to put together a loopback test in minutes.
>>
>> --n
>>
>> > Isaac
>> > _______________________________________________
>> > USRP-users mailing list
>> > USRP-users@lists.ettus.com
>> > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>
>>
>>
>
IG
Isaac Gerg
Fri, Jan 28, 2011 8:11 PM
Nick,
Just modified the waveform to send 2500 samples of this vector:
Float32 v = 0;
Float32 gain = 1.0;
Float64 fc = 100e3;
Float64 fs = 250e3;
Float64 tInc = 0;
for (Int32 k=0;k<vec.size();++k)
{
v = 2*M_PI*fc*(Float32(k)/fs); // 2 pi fc * t
vec[k] = complex<Float32>(gain*cos(v), gain*sin(v));
}
Still get the same result (10-20 samples of data containing energy, not 10ms
of them as expected).
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
"It also supports multiple channels per USRP (though not possible with
the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx and Tx
boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a sinusoid with a
frequency high enough to get through the BasicRX/TX transformers? It has
to be at least 100kHz. I'm not sure where you're specifying the
frequency of the complex vector you generate. I'm also not sure what
baseband frequency you're setting the USRP to when you initialize it. I
have a feeling the pulses you're seeing are DC filtered through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
Nick,
Just modified the waveform to send 2500 samples of this vector:
Float32 v = 0;
Float32 gain = 1.0;
Float64 fc = 100e3;
Float64 fs = 250e3;
Float64 tInc = 0;
for (Int32 k=0;k<vec.size();++k)
{
v = 2*M_PI*fc*(Float32(k)/fs); // 2 pi fc * t
vec[k] = complex<Float32>(gain*cos(v), gain*sin(v));
}
Still get the same result (10-20 samples of data containing energy, not 10ms
of them as expected).
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster <nick@ettus.com> wrote:
> On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
> > Just saw this from here:
> >
> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
> > "It also supports multiple channels per USRP (though not possible with
> the hardware yet) and looks much like the
> > single-USRP API."
> >
> > Does this mean a loopback using a single USRP with a BasicRx and Tx
> boards will not work until the hardware is modified?
>
> Loopback is supported. You should be able to send and receive
> simultaneously just fine. Are you sure you're sending a sinusoid with a
> frequency high enough to get through the BasicRX/TX transformers? It has
> to be at least 100kHz. I'm not sure where you're specifying the
> frequency of the complex vector you generate. I'm also not sure what
> baseband frequency you're setting the USRP to when you initialize it. I
> have a feeling the pulses you're seeing are DC filtered through the
> BasicRX/TX transformers.
>
> Incidentally, you might find it easier to work within Gnuradio instead
> of manually generating the waveforms in C++. You can also use
> gnuradio-companion to put together a loopback test in minutes.
>
> --n
>
> > Isaac
> > _______________________________________________
> > USRP-users mailing list
> > USRP-users@lists.ettus.com
> > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>
>
>
IG
Isaac Gerg
Fri, Jan 28, 2011 8:26 PM
I just made the following changes (for grins) and got a new result (still
not completely working though).
// Changed
Int32 burstSizeSamps = m_pDev->get_max_send_samps_per_packet(); // 3968
// Added
md.has_time_spec = false;
// kept the same (for reference)
Int32 numSampsSent = m_pDev->send(&vec.front(), vec.size(),
md,uhd::io_type_t::COMPLEX_FLOAT32,
Now, I get about 100 samples that have energy instead of the 10-20 I got
before.
For my system, the send and receive packet sizes are (respectively): 4096
3968
Isaac
On Fri, Jan 28, 2011 at 3:00 PM, Isaac Gerg isaac.gerg@gergltd.com wrote:
Please correct me if Im wrong, but even despite the DC filtering, I should
at least see contiguous energy thats the length of the burst I am sending.
Right? (Because Im not seeing this per my previous post).
Isaac
On Fri, Jan 28, 2011 at 2:58 PM, Isaac Gerg isaac.gerg@gergltd.comwrote:
Hi Nick,
I am generating 2500 samples which compose one full cycle of a complex
exponential. I am using fc=1e6. So to address your first comment: No, Im
not sure if my signal is making it through. I am under the impression that
whatever samples I send out to the BasicTX will be received at the BasicRX
(with low-pass filtering) and that I dont have modulate anything (the board
does that for me). Is this not true? (I understand there will
be attention, carrier offset, etc)
I am working on a project which is creating a new waveform. I originally
thought about using the GNUradio framework (I have used it in the past
successfully), but decided to use the UHD since it appeared flexible and
simple for the prototype I am building (we have multiple threads and
eventually seek to write a device driver to communicate between nodes and I
dont want the extra overhead of the GNURadio framework). Additionally, I
have written a simple transceiver class which functions very similar to what
the GNURadio framework would be giving me anyways (as far as an entry point
to get the data with) without all the hassle of dealing with python, blocks,
etc especially if we decide to do GPU processing in the future (which is
likely).
Can you recommend a waveform that I can send that you believe I should be
able to clearly observe (fc=?, etc). It sounded like if I generated a
100kHz waveform and sent it, that it should get through... am I correct on
this?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
"It also supports multiple channels per USRP (though not possible with
the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx and Tx
boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a sinusoid with a
frequency high enough to get through the BasicRX/TX transformers? It has
to be at least 100kHz. I'm not sure where you're specifying the
frequency of the complex vector you generate. I'm also not sure what
baseband frequency you're setting the USRP to when you initialize it. I
have a feeling the pulses you're seeing are DC filtered through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
I just made the following changes (for grins) and got a new result (still
not completely working though).
// Changed
Int32 burstSizeSamps = m_pDev->get_max_send_samps_per_packet(); // 3968
// Added
md.has_time_spec = false;
// kept the same (for reference)
Int32 numSampsSent = m_pDev->send(&vec.front(), vec.size(),
md,uhd::io_type_t::COMPLEX_FLOAT32,
Now, I get about 100 samples that have energy instead of the 10-20 I got
before.
For my system, the send and receive packet sizes are (respectively): 4096
3968
Isaac
On Fri, Jan 28, 2011 at 3:00 PM, Isaac Gerg <isaac.gerg@gergltd.com> wrote:
> Please correct me if Im wrong, but even despite the DC filtering, I should
> at least see contiguous energy thats the length of the burst I am sending.
> Right? (Because Im not seeing this per my previous post).
>
> Isaac
>
>
> On Fri, Jan 28, 2011 at 2:58 PM, Isaac Gerg <isaac.gerg@gergltd.com>wrote:
>
>> Hi Nick,
>>
>> I am generating 2500 samples which compose one full cycle of a complex
>> exponential. I am using fc=1e6. So to address your first comment: No, Im
>> not sure if my signal is making it through. I am under the impression that
>> whatever samples I send out to the BasicTX will be received at the BasicRX
>> (with low-pass filtering) and that I dont have modulate anything (the board
>> does that for me). Is this not true? (I understand there will
>> be attention, carrier offset, etc)
>>
>> I am working on a project which is creating a new waveform. I originally
>> thought about using the GNUradio framework (I have used it in the past
>> successfully), but decided to use the UHD since it appeared flexible and
>> simple for the prototype I am building (we have multiple threads and
>> eventually seek to write a device driver to communicate between nodes and I
>> dont want the extra overhead of the GNURadio framework). Additionally, I
>> have written a simple transceiver class which functions very similar to what
>> the GNURadio framework would be giving me anyways (as far as an entry point
>> to get the data with) without all the hassle of dealing with python, blocks,
>> etc especially if we decide to do GPU processing in the future (which is
>> likely).
>>
>> Can you recommend a waveform that I can send that you believe I should be
>> able to clearly observe (fc=?, etc). It sounded like if I generated a
>> 100kHz waveform and sent it, that it should get through... am I correct on
>> this?
>>
>> Thanks,
>> Isaac
>>
>>
>>
>> On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster <nick@ettus.com> wrote:
>>
>>> On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
>>> > Just saw this from here:
>>> >
>>> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
>>> > "It also supports multiple channels per USRP (though not possible with
>>> the hardware yet) and looks much like the
>>> > single-USRP API."
>>> >
>>> > Does this mean a loopback using a single USRP with a BasicRx and Tx
>>> boards will not work until the hardware is modified?
>>>
>>> Loopback is supported. You should be able to send and receive
>>> simultaneously just fine. Are you sure you're sending a sinusoid with a
>>> frequency high enough to get through the BasicRX/TX transformers? It has
>>> to be at least 100kHz. I'm not sure where you're specifying the
>>> frequency of the complex vector you generate. I'm also not sure what
>>> baseband frequency you're setting the USRP to when you initialize it. I
>>> have a feeling the pulses you're seeing are DC filtered through the
>>> BasicRX/TX transformers.
>>>
>>> Incidentally, you might find it easier to work within Gnuradio instead
>>> of manually generating the waveforms in C++. You can also use
>>> gnuradio-companion to put together a loopback test in minutes.
>>>
>>> --n
>>>
>>> > Isaac
>>> > _______________________________________________
>>> > USRP-users mailing list
>>> > USRP-users@lists.ettus.com
>>> > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>>
>>>
>>>
>>
>
NF
Nick Foster
Fri, Jan 28, 2011 9:18 PM
On Fri, 2011-01-28 at 14:58 -0500, Isaac Gerg wrote:
Hi Nick,
I am generating 2500 samples which compose one full cycle of a complex
exponential. I am using fc=1e6. So to address your first comment:
No, Im not sure if my signal is making it through. I am under the
impression that whatever samples I send out to the BasicTX will be
received at the BasicRX (with low-pass filtering) and that I dont have
modulate anything (the board does that for me). Is this not true? (I
understand there will be attention, carrier offset, etc)
If you're setting the USRP DDC to a center freq of 1e6 then yes,
whatever you send should be modulated on top of that freq. If you
happened to be sending DC instead, you would see a pulse on receive,
which would quickly return to zero. The transformer is a high-pass
filter.
Plot the data you're creating, not the data you're receiving, and make
sure you're actually sending what you think you are.
--n
I am working on a project which is creating a new waveform. I
originally thought about using the GNUradio framework (I have used it
in the past successfully), but decided to use the UHD since it
appeared flexible and simple for the prototype I am building (we have
multiple threads and eventually seek to write a device driver to
communicate between nodes and I dont want the extra overhead of the
GNURadio framework). Additionally, I have written a
simple transceiver class which functions very similar to what the
GNURadio framework would be giving me anyways (as far as an entry
point to get the data with) without all the hassle of dealing with
python, blocks, etc especially if we decide to do GPU processing in
the future (which is likely).
Can you recommend a waveform that I can send that you believe I should
be able to clearly observe (fc=?, etc). It sounded like if I
generated a 100kHz waveform and sent it, that it should get through...
am I correct on this?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
"It also supports multiple channels per USRP (though not
possible with the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx
and Tx boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a
sinusoid with a
frequency high enough to get through the BasicRX/TX
transformers? It has
to be at least 100kHz. I'm not sure where you're specifying
the
frequency of the complex vector you generate. I'm also not
sure what
baseband frequency you're setting the USRP to when you
initialize it. I
have a feeling the pulses you're seeing are DC filtered
through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio
instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
On Fri, 2011-01-28 at 14:58 -0500, Isaac Gerg wrote:
> Hi Nick,
>
>
> I am generating 2500 samples which compose one full cycle of a complex
> exponential. I am using fc=1e6. So to address your first comment:
> No, Im not sure if my signal is making it through. I am under the
> impression that whatever samples I send out to the BasicTX will be
> received at the BasicRX (with low-pass filtering) and that I dont have
> modulate anything (the board does that for me). Is this not true? (I
> understand there will be attention, carrier offset, etc)
If you're setting the USRP DDC to a center freq of 1e6 then yes,
whatever you send should be modulated on top of that freq. If you
happened to be sending DC instead, you would see a pulse on receive,
which would quickly return to zero. The transformer is a high-pass
filter.
Plot the data you're creating, not the data you're receiving, and make
sure you're actually sending what you think you are.
--n
>
>
> I am working on a project which is creating a new waveform. I
> originally thought about using the GNUradio framework (I have used it
> in the past successfully), but decided to use the UHD since it
> appeared flexible and simple for the prototype I am building (we have
> multiple threads and eventually seek to write a device driver to
> communicate between nodes and I dont want the extra overhead of the
> GNURadio framework). Additionally, I have written a
> simple transceiver class which functions very similar to what the
> GNURadio framework would be giving me anyways (as far as an entry
> point to get the data with) without all the hassle of dealing with
> python, blocks, etc especially if we decide to do GPU processing in
> the future (which is likely).
>
>
> Can you recommend a waveform that I can send that you believe I should
> be able to clearly observe (fc=?, etc). It sounded like if I
> generated a 100kHz waveform and sent it, that it should get through...
> am I correct on this?
>
>
> Thanks,
> Isaac
>
>
>
>
> On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster <nick@ettus.com> wrote:
>
> On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
> > Just saw this from here:
> >
> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
> > "It also supports multiple channels per USRP (though not
> possible with the hardware yet) and looks much like the
> > single-USRP API."
> >
> > Does this mean a loopback using a single USRP with a BasicRx
> and Tx boards will not work until the hardware is modified?
>
>
> Loopback is supported. You should be able to send and receive
> simultaneously just fine. Are you sure you're sending a
> sinusoid with a
> frequency high enough to get through the BasicRX/TX
> transformers? It has
> to be at least 100kHz. I'm not sure where you're specifying
> the
> frequency of the complex vector you generate. I'm also not
> sure what
> baseband frequency you're setting the USRP to when you
> initialize it. I
> have a feeling the pulses you're seeing are DC filtered
> through the
> BasicRX/TX transformers.
>
> Incidentally, you might find it easier to work within Gnuradio
> instead
> of manually generating the waveforms in C++. You can also use
> gnuradio-companion to put together a loopback test in minutes.
>
> --n
>
> > Isaac
> > _______________________________________________
> > USRP-users mailing list
> > USRP-users@lists.ettus.com
> >
> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>
>
>
>
IG
Isaac Gerg
Fri, Jan 28, 2011 9:22 PM
Hi Nick,
I am setting the tx and rx fc's to 1e6.
Thanks for the information on the transformer.
Will write out to disk and plot my samples in just a moment.
Thanks,
Isaac
On Fri, Jan 28, 2011 at 4:18 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:58 -0500, Isaac Gerg wrote:
Hi Nick,
I am generating 2500 samples which compose one full cycle of a complex
exponential. I am using fc=1e6. So to address your first comment:
No, Im not sure if my signal is making it through. I am under the
impression that whatever samples I send out to the BasicTX will be
received at the BasicRX (with low-pass filtering) and that I dont have
modulate anything (the board does that for me). Is this not true? (I
understand there will be attention, carrier offset, etc)
If you're setting the USRP DDC to a center freq of 1e6 then yes,
whatever you send should be modulated on top of that freq. If you
happened to be sending DC instead, you would see a pulse on receive,
which would quickly return to zero. The transformer is a high-pass
filter.
Plot the data you're creating, not the data you're receiving, and make
sure you're actually sending what you think you are.
--n
I am working on a project which is creating a new waveform. I
originally thought about using the GNUradio framework (I have used it
in the past successfully), but decided to use the UHD since it
appeared flexible and simple for the prototype I am building (we have
multiple threads and eventually seek to write a device driver to
communicate between nodes and I dont want the extra overhead of the
GNURadio framework). Additionally, I have written a
simple transceiver class which functions very similar to what the
GNURadio framework would be giving me anyways (as far as an entry
point to get the data with) without all the hassle of dealing with
python, blocks, etc especially if we decide to do GPU processing in
the future (which is likely).
Can you recommend a waveform that I can send that you believe I should
be able to clearly observe (fc=?, etc). It sounded like if I
generated a 100kHz waveform and sent it, that it should get through...
am I correct on this?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
"It also supports multiple channels per USRP (though not
possible with the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx
and Tx boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a
sinusoid with a
frequency high enough to get through the BasicRX/TX
transformers? It has
to be at least 100kHz. I'm not sure where you're specifying
the
frequency of the complex vector you generate. I'm also not
sure what
baseband frequency you're setting the USRP to when you
initialize it. I
have a feeling the pulses you're seeing are DC filtered
through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio
instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
Hi Nick,
I am setting the tx and rx fc's to 1e6.
Thanks for the information on the transformer.
Will write out to disk and plot my samples in just a moment.
Thanks,
Isaac
On Fri, Jan 28, 2011 at 4:18 PM, Nick Foster <nick@ettus.com> wrote:
> On Fri, 2011-01-28 at 14:58 -0500, Isaac Gerg wrote:
> > Hi Nick,
> >
> >
> > I am generating 2500 samples which compose one full cycle of a complex
> > exponential. I am using fc=1e6. So to address your first comment:
> > No, Im not sure if my signal is making it through. I am under the
> > impression that whatever samples I send out to the BasicTX will be
> > received at the BasicRX (with low-pass filtering) and that I dont have
> > modulate anything (the board does that for me). Is this not true? (I
> > understand there will be attention, carrier offset, etc)
>
> If you're setting the USRP DDC to a center freq of 1e6 then yes,
> whatever you send should be modulated on top of that freq. If you
> happened to be sending DC instead, you would see a pulse on receive,
> which would quickly return to zero. The transformer is a high-pass
> filter.
>
> Plot the data you're creating, not the data you're receiving, and make
> sure you're actually sending what you think you are.
>
> --n
>
> >
> >
> > I am working on a project which is creating a new waveform. I
> > originally thought about using the GNUradio framework (I have used it
> > in the past successfully), but decided to use the UHD since it
> > appeared flexible and simple for the prototype I am building (we have
> > multiple threads and eventually seek to write a device driver to
> > communicate between nodes and I dont want the extra overhead of the
> > GNURadio framework). Additionally, I have written a
> > simple transceiver class which functions very similar to what the
> > GNURadio framework would be giving me anyways (as far as an entry
> > point to get the data with) without all the hassle of dealing with
> > python, blocks, etc especially if we decide to do GPU processing in
> > the future (which is likely).
> >
> >
> > Can you recommend a waveform that I can send that you believe I should
> > be able to clearly observe (fc=?, etc). It sounded like if I
> > generated a 100kHz waveform and sent it, that it should get through...
> > am I correct on this?
> >
> >
> > Thanks,
> > Isaac
> >
> >
> >
> >
> > On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster <nick@ettus.com> wrote:
> >
> > On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
> > > Just saw this from here:
> > >
> >
> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
> > > "It also supports multiple channels per USRP (though not
> > possible with the hardware yet) and looks much like the
> > > single-USRP API."
> > >
> > > Does this mean a loopback using a single USRP with a BasicRx
> > and Tx boards will not work until the hardware is modified?
> >
> >
> > Loopback is supported. You should be able to send and receive
> > simultaneously just fine. Are you sure you're sending a
> > sinusoid with a
> > frequency high enough to get through the BasicRX/TX
> > transformers? It has
> > to be at least 100kHz. I'm not sure where you're specifying
> > the
> > frequency of the complex vector you generate. I'm also not
> > sure what
> > baseband frequency you're setting the USRP to when you
> > initialize it. I
> > have a feeling the pulses you're seeing are DC filtered
> > through the
> > BasicRX/TX transformers.
> >
> > Incidentally, you might find it easier to work within Gnuradio
> > instead
> > of manually generating the waveforms in C++. You can also use
> > gnuradio-companion to put together a loopback test in minutes.
> >
> > --n
> >
> > > Isaac
> > > _______________________________________________
> > > USRP-users mailing list
> > > USRP-users@lists.ettus.com
> > >
> >
> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
> >
> >
> >
> >
>
>
>
IG
Isaac Gerg
Fri, Jan 28, 2011 9:26 PM
Just to be clear, if I create a 1 KHz complex exponential composed of ~4k
samples and then send it with a tx fc of 1e6, I should be OK?
Isaac
On Fri, Jan 28, 2011 at 4:22 PM, Isaac Gerg isaac.gerg@gergltd.com wrote:
Hi Nick,
I am setting the tx and rx fc's to 1e6.
Thanks for the information on the transformer.
Will write out to disk and plot my samples in just a moment.
Thanks,
Isaac
On Fri, Jan 28, 2011 at 4:18 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:58 -0500, Isaac Gerg wrote:
Hi Nick,
I am generating 2500 samples which compose one full cycle of a complex
exponential. I am using fc=1e6. So to address your first comment:
No, Im not sure if my signal is making it through. I am under the
impression that whatever samples I send out to the BasicTX will be
received at the BasicRX (with low-pass filtering) and that I dont have
modulate anything (the board does that for me). Is this not true? (I
understand there will be attention, carrier offset, etc)
If you're setting the USRP DDC to a center freq of 1e6 then yes,
whatever you send should be modulated on top of that freq. If you
happened to be sending DC instead, you would see a pulse on receive,
which would quickly return to zero. The transformer is a high-pass
filter.
Plot the data you're creating, not the data you're receiving, and make
sure you're actually sending what you think you are.
--n
I am working on a project which is creating a new waveform. I
originally thought about using the GNUradio framework (I have used it
in the past successfully), but decided to use the UHD since it
appeared flexible and simple for the prototype I am building (we have
multiple threads and eventually seek to write a device driver to
communicate between nodes and I dont want the extra overhead of the
GNURadio framework). Additionally, I have written a
simple transceiver class which functions very similar to what the
GNURadio framework would be giving me anyways (as far as an entry
point to get the data with) without all the hassle of dealing with
python, blocks, etc especially if we decide to do GPU processing in
the future (which is likely).
Can you recommend a waveform that I can send that you believe I should
be able to clearly observe (fc=?, etc). It sounded like if I
generated a 100kHz waveform and sent it, that it should get through...
am I correct on this?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
"It also supports multiple channels per USRP (though not
possible with the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx
and Tx boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a
sinusoid with a
frequency high enough to get through the BasicRX/TX
transformers? It has
to be at least 100kHz. I'm not sure where you're specifying
the
frequency of the complex vector you generate. I'm also not
sure what
baseband frequency you're setting the USRP to when you
initialize it. I
have a feeling the pulses you're seeing are DC filtered
through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio
instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
Just to be clear, if I create a 1 KHz complex exponential composed of ~4k
samples and then send it with a tx fc of 1e6, I should be OK?
Isaac
On Fri, Jan 28, 2011 at 4:22 PM, Isaac Gerg <isaac.gerg@gergltd.com> wrote:
> Hi Nick,
>
> I am setting the tx and rx fc's to 1e6.
>
> Thanks for the information on the transformer.
>
> Will write out to disk and plot my samples in just a moment.
>
> Thanks,
> Isaac
>
> On Fri, Jan 28, 2011 at 4:18 PM, Nick Foster <nick@ettus.com> wrote:
>
>> On Fri, 2011-01-28 at 14:58 -0500, Isaac Gerg wrote:
>> > Hi Nick,
>> >
>> >
>> > I am generating 2500 samples which compose one full cycle of a complex
>> > exponential. I am using fc=1e6. So to address your first comment:
>> > No, Im not sure if my signal is making it through. I am under the
>> > impression that whatever samples I send out to the BasicTX will be
>> > received at the BasicRX (with low-pass filtering) and that I dont have
>> > modulate anything (the board does that for me). Is this not true? (I
>> > understand there will be attention, carrier offset, etc)
>>
>> If you're setting the USRP DDC to a center freq of 1e6 then yes,
>> whatever you send should be modulated on top of that freq. If you
>> happened to be sending DC instead, you would see a pulse on receive,
>> which would quickly return to zero. The transformer is a high-pass
>> filter.
>>
>> Plot the data you're creating, not the data you're receiving, and make
>> sure you're actually sending what you think you are.
>>
>> --n
>>
>> >
>> >
>> > I am working on a project which is creating a new waveform. I
>> > originally thought about using the GNUradio framework (I have used it
>> > in the past successfully), but decided to use the UHD since it
>> > appeared flexible and simple for the prototype I am building (we have
>> > multiple threads and eventually seek to write a device driver to
>> > communicate between nodes and I dont want the extra overhead of the
>> > GNURadio framework). Additionally, I have written a
>> > simple transceiver class which functions very similar to what the
>> > GNURadio framework would be giving me anyways (as far as an entry
>> > point to get the data with) without all the hassle of dealing with
>> > python, blocks, etc especially if we decide to do GPU processing in
>> > the future (which is likely).
>> >
>> >
>> > Can you recommend a waveform that I can send that you believe I should
>> > be able to clearly observe (fc=?, etc). It sounded like if I
>> > generated a 100kHz waveform and sent it, that it should get through...
>> > am I correct on this?
>> >
>> >
>> > Thanks,
>> > Isaac
>> >
>> >
>> >
>> >
>> > On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster <nick@ettus.com> wrote:
>> >
>> > On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
>> > > Just saw this from here:
>> > >
>> >
>> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
>> > > "It also supports multiple channels per USRP (though not
>> > possible with the hardware yet) and looks much like the
>> > > single-USRP API."
>> > >
>> > > Does this mean a loopback using a single USRP with a BasicRx
>> > and Tx boards will not work until the hardware is modified?
>> >
>> >
>> > Loopback is supported. You should be able to send and receive
>> > simultaneously just fine. Are you sure you're sending a
>> > sinusoid with a
>> > frequency high enough to get through the BasicRX/TX
>> > transformers? It has
>> > to be at least 100kHz. I'm not sure where you're specifying
>> > the
>> > frequency of the complex vector you generate. I'm also not
>> > sure what
>> > baseband frequency you're setting the USRP to when you
>> > initialize it. I
>> > have a feeling the pulses you're seeing are DC filtered
>> > through the
>> > BasicRX/TX transformers.
>> >
>> > Incidentally, you might find it easier to work within Gnuradio
>> > instead
>> > of manually generating the waveforms in C++. You can also use
>> > gnuradio-companion to put together a loopback test in minutes.
>> >
>> > --n
>> >
>> > > Isaac
>> > > _______________________________________________
>> > > USRP-users mailing list
>> > > USRP-users@lists.ettus.com
>> > >
>> >
>> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>> >
>> >
>> >
>> >
>>
>>
>>
>
IG
Isaac Gerg
Fri, Jan 28, 2011 9:45 PM
Just completed. The data I am creating/sending looks fine. Is there
anything else you would like me to check?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 4:18 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:58 -0500, Isaac Gerg wrote:
Hi Nick,
I am generating 2500 samples which compose one full cycle of a complex
exponential. I am using fc=1e6. So to address your first comment:
No, Im not sure if my signal is making it through. I am under the
impression that whatever samples I send out to the BasicTX will be
received at the BasicRX (with low-pass filtering) and that I dont have
modulate anything (the board does that for me). Is this not true? (I
understand there will be attention, carrier offset, etc)
If you're setting the USRP DDC to a center freq of 1e6 then yes,
whatever you send should be modulated on top of that freq. If you
happened to be sending DC instead, you would see a pulse on receive,
which would quickly return to zero. The transformer is a high-pass
filter.
Plot the data you're creating, not the data you're receiving, and make
sure you're actually sending what you think you are.
--n
I am working on a project which is creating a new waveform. I
originally thought about using the GNUradio framework (I have used it
in the past successfully), but decided to use the UHD since it
appeared flexible and simple for the prototype I am building (we have
multiple threads and eventually seek to write a device driver to
communicate between nodes and I dont want the extra overhead of the
GNURadio framework). Additionally, I have written a
simple transceiver class which functions very similar to what the
GNURadio framework would be giving me anyways (as far as an entry
point to get the data with) without all the hassle of dealing with
python, blocks, etc especially if we decide to do GPU processing in
the future (which is likely).
Can you recommend a waveform that I can send that you believe I should
be able to clearly observe (fc=?, etc). It sounded like if I
generated a 100kHz waveform and sent it, that it should get through...
am I correct on this?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster nick@ettus.com wrote:
On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
"It also supports multiple channels per USRP (though not
possible with the hardware yet) and looks much like the
single-USRP API."
Does this mean a loopback using a single USRP with a BasicRx
and Tx boards will not work until the hardware is modified?
Loopback is supported. You should be able to send and receive
simultaneously just fine. Are you sure you're sending a
sinusoid with a
frequency high enough to get through the BasicRX/TX
transformers? It has
to be at least 100kHz. I'm not sure where you're specifying
the
frequency of the complex vector you generate. I'm also not
sure what
baseband frequency you're setting the USRP to when you
initialize it. I
have a feeling the pulses you're seeing are DC filtered
through the
BasicRX/TX transformers.
Incidentally, you might find it easier to work within Gnuradio
instead
of manually generating the waveforms in C++. You can also use
gnuradio-companion to put together a loopback test in minutes.
--n
Just completed. The data I am creating/sending looks fine. Is there
anything else you would like me to check?
Thanks,
Isaac
On Fri, Jan 28, 2011 at 4:18 PM, Nick Foster <nick@ettus.com> wrote:
> On Fri, 2011-01-28 at 14:58 -0500, Isaac Gerg wrote:
> > Hi Nick,
> >
> >
> > I am generating 2500 samples which compose one full cycle of a complex
> > exponential. I am using fc=1e6. So to address your first comment:
> > No, Im not sure if my signal is making it through. I am under the
> > impression that whatever samples I send out to the BasicTX will be
> > received at the BasicRX (with low-pass filtering) and that I dont have
> > modulate anything (the board does that for me). Is this not true? (I
> > understand there will be attention, carrier offset, etc)
>
> If you're setting the USRP DDC to a center freq of 1e6 then yes,
> whatever you send should be modulated on top of that freq. If you
> happened to be sending DC instead, you would see a pulse on receive,
> which would quickly return to zero. The transformer is a high-pass
> filter.
>
> Plot the data you're creating, not the data you're receiving, and make
> sure you're actually sending what you think you are.
>
> --n
>
> >
> >
> > I am working on a project which is creating a new waveform. I
> > originally thought about using the GNUradio framework (I have used it
> > in the past successfully), but decided to use the UHD since it
> > appeared flexible and simple for the prototype I am building (we have
> > multiple threads and eventually seek to write a device driver to
> > communicate between nodes and I dont want the extra overhead of the
> > GNURadio framework). Additionally, I have written a
> > simple transceiver class which functions very similar to what the
> > GNURadio framework would be giving me anyways (as far as an entry
> > point to get the data with) without all the hassle of dealing with
> > python, blocks, etc especially if we decide to do GPU processing in
> > the future (which is likely).
> >
> >
> > Can you recommend a waveform that I can send that you believe I should
> > be able to clearly observe (fc=?, etc). It sounded like if I
> > generated a 100kHz waveform and sent it, that it should get through...
> > am I correct on this?
> >
> >
> > Thanks,
> > Isaac
> >
> >
> >
> >
> > On Fri, Jan 28, 2011 at 2:46 PM, Nick Foster <nick@ettus.com> wrote:
> >
> > On Fri, 2011-01-28 at 14:11 -0500, Isaac Gerg wrote:
> > > Just saw this from here:
> > >
> >
> http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/2010-October/000040.html
> > > "It also supports multiple channels per USRP (though not
> > possible with the hardware yet) and looks much like the
> > > single-USRP API."
> > >
> > > Does this mean a loopback using a single USRP with a BasicRx
> > and Tx boards will not work until the hardware is modified?
> >
> >
> > Loopback is supported. You should be able to send and receive
> > simultaneously just fine. Are you sure you're sending a
> > sinusoid with a
> > frequency high enough to get through the BasicRX/TX
> > transformers? It has
> > to be at least 100kHz. I'm not sure where you're specifying
> > the
> > frequency of the complex vector you generate. I'm also not
> > sure what
> > baseband frequency you're setting the USRP to when you
> > initialize it. I
> > have a feeling the pulses you're seeing are DC filtered
> > through the
> > BasicRX/TX transformers.
> >
> > Incidentally, you might find it easier to work within Gnuradio
> > instead
> > of manually generating the waveforms in C++. You can also use
> > gnuradio-companion to put together a loopback test in minutes.
> >
> > --n
> >
> > > Isaac
> > > _______________________________________________
> > > USRP-users mailing list
> > > USRP-users@lists.ettus.com
> > >
> >
> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
> >
> >
> >
> >
>
>
>
