Just a data point. Here's a 256QAM constellation from an Ettus B210. I
don't know what the exact EVM percentage is, but the constellation looks
pretty good. This is at 429 MHz, 10.721074 Msps and 42.884296 MHz master
clock rate with a commercial digital television demodulator evaluation
board.
Ron
On 3/26/21 14:08, Beaudoin, Christopher J wrote:
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 <[email protected]>
*Sent:* Friday, March 26, 2021 4:54 PM
*To:* Beaudoin, Christopher J <[email protected]>
*Cc:* Julian Arnold <[email protected]>;
[email protected] <[email protected]>
*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
<[email protected]> 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 <[email protected]>
*Sent:* Friday, March 26, 2021 2:45 PM
*To:* Julian Arnold <[email protected]>
*Cc:* Beaudoin, Christopher J <[email protected]>;
[email protected] <[email protected]>
*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 <[email protected]>
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 [email protected]:
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 -- [email protected]
To unsubscribe send an email to [email protected]
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
