On 04/27/2021 02:07 PM, Hirst (US), Stephen wrote:
Thanks Marcus for the quick reply!
I have a B210 with a GPSDO (the TCXO one).
And the system is “offline only” now but I can manually type a
response with something specific if you like.
So I have a little checkbox that I click on my GRC GUI that starts the
recording to a File Sink from a “UHD: USRP Source”. Here’s some
details of what I have specified in the UHD: USRP Source - File Sink
Box Pair.
UHD: USRP Source
GENERAL
Output Type: Complex float32
Wire Format: Complex int8
Stream args: peak=0.003906
Stream Channels: []
Device Address:
Device Arguments: “send_frame_size=8192”
Sync: Unknown PPS
Clock Rate (Hz): Default
Num Mboards: 1
Mb0 Clock Source: Default
Mb0 Time Source: Default
Mb0 Subdev Spec: A:A
Num Channels: 1
Samp rate (Sps): 2000000
ADVANCED
Alias:
Affinity:
Minoutbuf: 8192
Maxoutbuf: 8192
Show LO Controls: OFF
Comment:
RF OPTIONS
Ch0 Center Freq: 2000000000
Ch0 AGC: Default
Ch0: Gain Value: 30
Ch0 Antenna: RX2
Ch0 Bandwidth: 0.2MHz
FE CORRECTIONS
Ch0 Enable DC Offset Correction: true
Ch0 Enable IQ Imbalance Correction: true
[FILE SINK]
GENERAL
File: [Your favorite file name here]
Input Type: complex
Vec length: 1
Unbuffered: OFF
Append file: OFF
Does this help at all?
I tried the ./rx_samples_to_file from within “/usr/lib/uhd/examples”
with similar options but I don’t see an O/U with the selected options.
Is this a GRC thing? I have no idea…
https://files.ettus.com/manual/page_general.html#general_ounotes
https://files.ettus.com/manual/page_transport.html#transport_usb
The 'O' indication is coming from deep within the UHD drivers and has
nothing to do with Gnu Radio, per se.
If the bandwidth of your signal of interest really is only 200kHz, then
you don't need to record at 2Msps. Indeed you could
easily get away with sampling at a much much lower rate. Like
500ksps. This will save disk space while preserving your
data precision (complex floats). You can also write your samples in
a different format, like complex-shorts, halving the
amount of disk space used.
But given that you can use rx_samples_to_file apparently successfully,
and cannot use Gnu Radio successfully to accomplish
more-or-less the same task, then Robs suggestion to initially
eliminate Gnu Radio from the equation is probably a good one.
Also, are you running your machine with the CPU governor set to
"performance" or "power-save"? Are you running inside
a VM, or on the actual raw hardware?
*From:* Marcus D Leech [mailto:[email protected]]
*Sent:* Tuesday, April 27, 2021 1:37 PM
*To:* Hirst (US), Stephen <[email protected]>
*Cc:* [email protected]
*Subject:* [EXTERNAL] Re: [USRP-users] Underflows/Overflows, Filesize
and a Complete N00B
EXT email: be mindful of links/attachments.
If you just use the rx_samples_to_file utility that comes with UHD
what are your results
Can you share your GNuRadio flow graph with us? What type of USRP do
you have? There are about a dozen different models of USRP out there.
Sent from my iPhone
On Apr 27, 2021, at 1:24 PM, Hirst (US), Stephen
<[email protected] <mailto:[email protected]>> wrote:
Hi Guys,
I appreciate you all and thank you for the great community that is
this mailing list! Continue on!
I, however, am a complete N00B and require assistance on some
seemingly minor tasks that I believe you could help me out on. :)
With that being said, I find myself with the Ubuntu 20.04LTS
install on a rather “fast” Dell laptop (specs down below.)
-8th Gen Intel Core i7-8650U Processor (Quad Core, 8M Cache,
1.9GHz,15W, vPro)
-16GB, 2x8GB, 2666MHz DDR4 Non-ECC
-M.2 1TB PCIe NVMe Class 40 Opal 2.0 Self Encrypting Solid State Drive
I have fiddled around with an online install on the laptop but I
haven’t gotten the overflows and underflows to cooperate. And to
be more specific, I can record a sample rate of 2Mhz and 0.2Mhz
bandwidth without issue, but I can’t get a SR or BW greater than
those to work without Under/Overflows developing. (I have been
using UHD 3.15.0.02build5 with GNURadio 3.8.1.0 (Python 3.8.2)
(grc and grc-uhd) and installed everything with “apt”. The first
line in my grc python output labels “Linux GNU C++ version 9.2.1
20200304; Boost_107100; UHD_3.15.0.0-2build5”, USB3.0). Any ideas?
I tried both the num_recv_frame size stuff and setting high
priority with no joy, but then again, I am a N00b, so I may not
have done that correctly as well…
Also if anyone has experience with GRC specifically, I have had
success recording what I wanted to with a file with the 2MHz
Sample Rate with a 0.2 MHz BW, but the files are pretty huge.
We’re talking 1GB/min… Any idea how to make the GRC record to a
less-fine resolution so I can free up some HDD space?
Thanks guys!
Ya Boi Steve
_______________________________________________
USRP-users mailing list -- [email protected]
<mailto:[email protected]>
To unsubscribe send an email to [email protected]
<mailto:[email protected]>
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
