So both should be synchronized (unless I'm missing something) because I'm
externally referencing the B210 to the signal generator's 10 MHz reference. The
EVM over a couple 1000 symbols is good but when one looks at the EVM over ~13e6
samples, it deteriorates dramatically. Looking at the rx waveform as compared
to the tx waveform indicates the source of EVM deterioration is related to
phase instability somewhere. When I got to this point, I started to ask myself
if I had set my expectations for the B210's EVM performance too high so I
paused to pose this question to the forum. If I had to guess based on the
analysis I've done so far, I'd say it appears that the master clock isn't
perfectly sync'd with the external reference on the scale of 0.5 to 1 second.
________________________________
From: Marcus D Leech <patchvonbr...@gmail.com>
Sent: Friday, March 26, 2021 4:54 PM
To: Beaudoin, Christopher J <christopher_beaud...@uml.edu>
Cc: Julian Arnold <jul...@elitecoding.org>; USRP-users@lists.ettus.com
<USRP-users@lists.ettus.com>
Subject: Re: [USRP-users] Re: B210 EVM
This e-mail originated from outside the UMass Lowell network.
________________________________
I would suggest going back to basics. What does the RX spectrum look like
compared to the TX spectrum? Are you doing clock recovery on the RX side, or
assuming both sides are synchronized?
Sent from my iPhone
On Mar 26, 2021, at 4:38 PM, Beaudoin, Christopher J
<christopher_beaud...@uml.edu> wrote:
Hello Marcus,
Sorry for the terse nature of my previous message. To be
more specific, the precise symbol rate is 4.608 MHz so the actual sample rate
is 27.648 MHz; the USRP sets the master clock rate to 27.648 MHz when I command
the sample rate. I'm not seeing any overruns at this rate and we spent a fair
bit of time fine tuning the host machine to sustain this data rate. It will
sustain this rate for as long at 10 minutes without reporting any "O" or "U"
errors. I also embed the time stamps into the recorded data file and post
recording analysis does not indicate any time disruptions.
I'm certain that the mod signal (from my vector signal analyzer) has very low
EVM (~1%) as confirmed with my Rhode Schwartz signal analyzer. I've also
considered saturation of amplifier stages within the AD9361. With 55 dB of
gain, I obtain a rms ADC 16-bit state count of ~15000 for a -40 dBm input
level. As I understand this should be a suitable level given the B210's IIP3
spec is -20 dBm. I've also reduced the input level at constant gain and didnt
observe any net improvement in the EVM.
When properly configured, can I expect the B210 to yield an EVM better than say
5%?
Chris
Creating the usrp device with: num_recv_frames=1024...
[INFO] [UHD] linux; GNU C++ version 5.4.0 20160609; Boost_105800;
UHD_3.11.1.0-0-unknown
[INFO] [B200] Detected Device: B210
[INFO] [B200] Operating over USB 3.
[INFO] [B200] Initialize CODEC control...
[INFO] [B200] Initialize Radio control...
[INFO] [B200] Performing register loopback test...
[INFO] [B200] Register loopback test passed
[INFO] [B200] Performing register loopback test...
[INFO] [B200] Register loopback test passed
[INFO] [B200] Setting master clock rate selection to 'automatic'.
[INFO] [B200] Asking for clock rate 16.000000 MHz...
[INFO] [B200] Actually got clock rate 16.000000 MHz.
Using Device: Single USRP:
Device: B-Series Device
Mboard 0: B210
RX Channel: 0
RX DSP: 0
RX Dboard: A
RX Subdev: FE-RX2
RX Channel: 1
RX DSP: 1
RX Dboard: A
RX Subdev: FE-RX1
TX Channel: 0
TX DSP: 0
TX Dboard: A
TX Subdev: FE-TX2
TX Channel: 1
TX DSP: 1
TX Dboard: A
TX Subdev: FE-TX1
Setting RX Freq: 3199000000.000 Hz...
Actual RX Freq: 3199000000.000 Hz...
Setting RX Rate: 27648000.000 Sps...
[INFO] [B200] Asking for clock rate 27.648000 MHz...
[INFO] [B200] Actually got clock rate 27.648000 MHz.
Actual RX Rate: 27648000.081 Sps...
Setting RX Gain: 55.000000 dB...
Actual RX Gain: 55.000000 dB...
Waiting for "lo_locked": ++++++++++ locked.
Press Ctrl + C to stop streaming...
________________________________
From: Marcus D Leech <patchvonbr...@gmail.com>
Sent: Friday, March 26, 2021 2:45 PM
To: Julian Arnold <jul...@elitecoding.org>
Cc: Beaudoin, Christopher J <christopher_beaud...@uml.edu>;
USRP-users@lists.ettus.com <USRP-users@lists.ettus.com>
Subject: Re: [USRP-users] Re: B210 EVM
This e-mail originated from outside the UMass Lowell network.
________________________________
At 30MSPS are you seeing any overruns? What is your master clock rate?
Sent from my iPhone
On Mar 26, 2021, at 2:41 PM, Julian Arnold <jul...@elitecoding.org> wrote:
Chris,
I would say that your EVM is mainly affected by your SNR and your digital
receiver implementation (AGC / filters / clock recovery / equalizer / ...).
So without more details it’s going to be hard to say if what you are seeing is
within limits.
Cheers,
Julian Arnold, M.Sc
Am 26.03.2021 um 18:29 schrieb christopher_beaud...@uml.edu:
I'm capturing a 3 GHz QPSK signal with 5 MHz symbol rate by sampling the signal
at 6 times the symbol rate. The B210 is externally referenced to a very clean
10 MHz reference. My measurements of the EVM sampling the signal for ~0.5
seconds are pretty poor ~30-40%. I can provide more setup details but I'm
wondering if others can comment on what EVM I can expect. I'm hoping this isn't
a fundamental limitation of this hardware system.
Thanks,
Chris
_______________________________________________
USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-le...@lists.ettus.com
_______________________________________________
USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-le...@lists.ettus.com
_______________________________________________
USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-le...@lists.ettus.com
