On 07/18/2018 03:55 PM, Veytser, Leonid - 0665 - MITLL wrote:
I switched from using txrx but instead took an arbitrary waveform
generator that I have and was able to generate a sine wave there and
see it on N310. I did have to bump up the receive gain though. Not
sure what the problem was with transmitting the waveform from the N310
but since I am only concerned about RX at the moment it’s a moot issue
for me.
However, I am trying to see if I can synchronize all 4 channels on
N310. I am using an external Octoclock and it seems like channel 0 and
1 are synchronized if I plot the received sine wave. Also, channel 2
and 3 are symphonized. However there is a phase shift between channels
0,1 and 2,3.
I know each of these pairs of channels have own AD9371 and own
oscillator. Should I be able to synchronize across both? I tried
various suggestions on this page
(https://files.ettus.com/manual/page_sync.html) that talked about
synchronizing channel phase but no luck.
Lenny
My understanding is that there's no way to get predictable mutual phase
offset between the AD9371, unless you use an external-LO, which I haven't
used at this point, but it requires 2XLO, so you'd need to put
3600MHz into that port and select external LO.
*From: *"Marcus D. Leech" <[email protected]>
*Date: *Wednesday, July 18, 2018 at 11:14 AM
*To: *Leonid Veytser <[email protected]>,
"[email protected]" <[email protected]>
*Subject: *Re: [USRP-users] Setting N310 TX and RX bandwidth
On 07/18/2018 09:57 AM, Veytser, Leonid - 0665 - MITLL wrote:
The output samples file is parsed into raw i/q values and plotted
in gnuplot.
Currently, the loopback is done via just direct RF cable between
RF0 TX/RX radio and RF1 RX2 radio with 30 db of attenuation inline.
Lenny
So, try this command:
./txrx_loopback_to_file --tx-args
"addr=192.168.20.2,clock_source=external,time_source=external"
--rx-args
"addr=192.168.20.2,clock_source=external,time_source=external"
--tx-rate 1.25e6 --rx-rate 1.25e6 --tx-freq 1800e6 --rx-freq 1800e6
--tx-channels "0" --rx-channels "0" --tx-subdev A:0 --rx-subdev B:0
--wave-type SINE --wave-freq 1e3
*From: *"Marcus D. Leech" <[email protected]>
<mailto:[email protected]>
*Date: *Tuesday, July 17, 2018 at 8:54 PM
*To: *Leonid Veytser <[email protected]>
<mailto:[email protected]>, "[email protected]"
<mailto:[email protected]> <[email protected]>
<mailto:[email protected]>
*Subject: *Re: [USRP-users] Setting N310 TX and RX bandwidth
On 07/17/2018 02:13 PM, Veytser, Leonid - 0665 - MITLL wrote:
I tried various gain settings, including both 0 and 30 for TX
and 0 and 30 for RX.
I also tried setting clock and time sources to “internal” but
it didn’t seem to make a difference.
How are you inspecting the file that it produces to see if the
signal is there or not?
How are you doing the loopback?
*From: *"Marcus D. Leech" <[email protected]>
<mailto:[email protected]>
*Date: *Tuesday, July 17, 2018 at 1:02 PM
*To: *Leonid Veytser <[email protected]>
<mailto:[email protected]>, "[email protected]"
<mailto:[email protected]>
<[email protected]> <mailto:[email protected]>
*Subject: *Re: [USRP-users] Setting N310 TX and RX bandwidth
On 07/17/2018 12:22 PM, Veytser, Leonid - 0665 - MITLL wrote:
I tried both of your suggestions – increasing RX gain and
offsetting RX frequency. However, neither seem to be working.
Lenny
What gain setting are you using for both TX and RX?
What happens if you don't use external time and clock sources?
*From: *"Marcus D. Leech" <[email protected]>
<mailto:[email protected]>
*Date: *Monday, July 16, 2018 at 10:08 PM
*To: *Leonid Veytser <[email protected]>
<mailto:[email protected]>, "[email protected]"
<mailto:[email protected]>
<[email protected]>
<mailto:[email protected]>
*Subject: *Re: [USRP-users] Setting N310 TX and RX bandwidth
On 07/16/2018 06:53 PM, Veytser, Leonid - 0665 - MITLL wrote:
Hi Marcus,
Thanks for your answer. Perhaps this is not my issue
then. I am having issue with sending and receiving a
simple sine wave using N310. In the simplest case, I
can take X310 run txrx_loopback_to_file with the
arguments below, plot reals and imaginaries of the
stored files and I can see the sine wave. If I do the
exact same command against a N310, I appear to just
see noise, which seems to be highly quantized. Just
values in the range between -4 and 4.
./txrx_loopback_to_file --tx-args
"addr=192.168.20.2,clock_source=external,time_source=external"
--rx-args
"addr=192.168.20.2,clock_source=external,time_source=external"
--tx-rate 1.25e6 --rx-rate 1.25e6 --tx-freq 1800e6
--rx-freq 1800e6 --tx-channels "0" --rx-channels "1"
--wave-type SINE --wave-freq 1e3
Any help or suggestion is greatly appreciated!
Thanks,
Lenny
The gain-control range on the AD9371 is much larger than
on cards you'll find on the X310.
Try increasing the RX gain, and offset the RX frequency a
bit -- you may be losing some in DC-offset removal for a
carrier that is right on top of
the "DC" region.
*From: *USRP-users
<[email protected]>
<mailto:[email protected]> on behalf
of "Marcus D. Leech via USRP-users"
<[email protected]>
<mailto:[email protected]>
*Reply-To: *"Marcus D. Leech" <[email protected]>
<mailto:[email protected]>
*Date: *Monday, July 16, 2018 at 3:20 PM
*To: *"[email protected]"
<mailto:[email protected]>
<[email protected]>
<mailto:[email protected]>
*Subject: *Re: [USRP-users] Setting N310 TX and RX
bandwidth
On 07/16/2018 02:46 PM, Veytser, Leonid - 0665 - MITLL
via USRP-users wrote:
I am unable to set either RX or TX bandwidth on
N310. When attempting to set, I get the following
warning:
Setting TX Bandwidth: 40.000000 MHz...
*[WARNING] [0/Radio_0] *set_tx_bandwidth take no
effect on AD9371. Default analog bandwidth is 100MHz
Actual TX Bandwidth: 0.000000 MHz...
and
Setting RX Bandwidth: 40.000000 MHz...
*[WARNING] [0/Radio_0] *set_rx_bandwidth take no
effect on AD9371. Default analog bandwidth is 100MHz
Actual RX Bandwidth: 100.000000 MHz...
When looking through the UHD code, I tracked down
to these FIXME comments:
https://github.com/EttusResearch/uhd/blob/master/host/lib/usrp/dboard/magnesium/magnesium_radio_ctrl_impl.cpp#L322
Is this some sort of limitation with the N310 and
the AD9371 tranceiver? Does this mean I am unable
to set TX and RX bandwidth at all?
Thanks,
Lenny
Based purely on the comment, I'm guessing that there
are notionally registers to control this in the
AD9371, but they don't apparently
work as documented, hence the warning message.
Keep in mind that *internally*, the AD9371 samples the
analog mixer outputs at several hundred MHz, so if the
internal anti-alias filters
are set-up for 100MHz by default, there's no danger
of aliases appearing in the outputs, regardless of
your ultimate sample-rate delivered
to the host.
_______________________________________________
USRP-users mailing list
[email protected]
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
