Empathy List Archives

R

RizThon

Tue, Jul 24, 2018 7:18 AM

Hi all,

I'm still having issues with a B210 with sample rates greater than 28MS/s.
At 28 I get a few overruns out of a million blocks of 1024 samples, but as
I sample faster, the number of overruns grow fast.

I have an AMD Ryzen 7 2700 (8 physical cores @ 3.2GHz), 16GB RAM, Gigabyte
GA-AB350N, Windows 10. The USB driver is from
http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip .

I've tried adding "num_recv_frames=256" but I get an error:

Creating the usrp device with: num_recv_frames=256...
[INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_106300;
UHD_3.11.0.1-37-g2c9087d1
[INFO] [B200] Detected Device: B210
[INFO] [B200] Operating over USB 3.
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1

I would have expected to be able to stream with barely any overrun at
56MS/s on that computer. I stream at that speed in dual channel on a
LimeSDR (in SC12 format).

Any advice on how to get better results?

Thanks.

Hi all, I'm still having issues with a B210 with sample rates greater than 28MS/s. At 28 I get a few overruns out of a million blocks of 1024 samples, but as I sample faster, the number of overruns grow fast. I have an AMD Ryzen 7 2700 (8 physical cores @ 3.2GHz), 16GB RAM, Gigabyte GA-AB350N, Windows 10. The USB driver is from http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip . I've tried adding "num_recv_frames=256" but I get an error: Creating the usrp device with: num_recv_frames=256... [INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_106300; UHD_3.11.0.1-37-g2c9087d1 [INFO] [B200] Detected Device: B210 [INFO] [B200] Operating over USB 3. [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: LIBUSB_ERROR_CODE -1 I would have expected to be able to stream with barely any overrun at 56MS/s on that computer. I stream at that speed in dual channel on a LimeSDR (in SC12 format). Any advice on how to get better results? Thanks.

MD

Marcus D. Leech

Tue, Jul 24, 2018 7:06 PM

On 07/24/2018 03:18 AM, RizThon via USRP-users wrote:

Hi all,

I'm still having issues with a B210 with sample rates greater than
28MS/s. At 28 I get a few overruns out of a million blocks of 1024
samples, but as I sample faster, the number of overruns grow fast.

I have an AMD Ryzen 7 2700 (8 physical cores @ 3.2GHz), 16GB
RAM, Gigabyte GA-AB350N, Windows 10. The USB driver is from
http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip .

I've tried adding "num_recv_frames=256" but I get an error:

Creating the usrp device with: num_recv_frames=256...
[INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_106300;
UHD_3.11.0.1-37-g2c9087d1
[INFO] [B200] Detected Device: B210
[INFO] [B200] Operating over USB 3.
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1

I would have expected to be able to stream with barely any overrun at
56MS/s on that computer. I stream at that speed in dual channel on a
LimeSDR (in SC12 format).

Any advice on how to get better results?

Thanks.

You cannot stream faster than about 30Msps from the B210 when in
dual-channel mode. This is a limitation of the data bus architecture on
the AD9371 chip, and not strictly a limitation of the B2xx FPGA.

On 07/24/2018 03:18 AM, RizThon via USRP-users wrote: > Hi all, > > I'm still having issues with a B210 with sample rates greater than > 28MS/s. At 28 I get a few overruns out of a million blocks of 1024 > samples, but as I sample faster, the number of overruns grow fast. > > I have an AMD Ryzen 7 2700 (8 physical cores @ 3.2GHz), 16GB > RAM, Gigabyte GA-AB350N, Windows 10. The USB driver is from > http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip . > > I've tried adding "num_recv_frames=256" but I get an error: > > Creating the usrp device with: num_recv_frames=256... > [INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_106300; > UHD_3.11.0.1-37-g2c9087d1 > [INFO] [B200] Detected Device: B210 > [INFO] [B200] Operating over USB 3. > [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: > LIBUSB_ERROR_CODE -1 > > I would have expected to be able to stream with barely any overrun at > 56MS/s on that computer. I stream at that speed in dual channel on a > LimeSDR (in SC12 format). > > Any advice on how to get better results? > > Thanks. > You cannot stream faster than about 30Msps from the B210 when in dual-channel mode. This is a limitation of the data bus architecture on the AD9371 chip, and not strictly a limitation of the B2xx FPGA.

MB

Martin Braun

Tue, Jul 24, 2018 8:54 PM

FYI, we fixed a couple of sc12/8 related issues on our latest master
branch. As Marcus says, if you're doing dual channel on the B210, you
can only go to 30.72 Msps though.

-- M

On 07/24/2018 12:06 PM, Marcus D. Leech via USRP-users wrote:

On 07/24/2018 03:18 AM, RizThon via USRP-users wrote:

Hi all,

I'm still having issues with a B210 with sample rates greater than
28MS/s. At 28 I get a few overruns out of a million blocks of 1024
samples, but as I sample faster, the number of overruns grow fast.

I have an AMD Ryzen 7 2700 (8 physical cores @ 3.2GHz), 16GB
RAM, Gigabyte GA-AB350N, Windows 10. The USB driver is
from http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip .

I've tried adding "num_recv_frames=256" but I get an error:

Creating the usrp device with: num_recv_frames=256...
[INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_106300;
UHD_3.11.0.1-37-g2c9087d1
[INFO] [B200] Detected Device: B210
[INFO] [B200] Operating over USB 3.
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1

I would have expected to be able to stream with barely any overrun at
56MS/s on that computer. I stream at that speed in dual channel on a
LimeSDR (in SC12 format).

Any advice on how to get better results?

Thanks.

You cannot stream faster than about 30Msps from the B210 when in
dual-channel mode. This is a limitation of the data bus architecture on
the AD9371 chip, and not strictly a limitation of the B2xx FPGA.

USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

FYI, we fixed a couple of sc12/8 related issues on our latest master branch. As Marcus says, if you're doing dual channel on the B210, you can only go to 30.72 Msps though. -- M On 07/24/2018 12:06 PM, Marcus D. Leech via USRP-users wrote: > On 07/24/2018 03:18 AM, RizThon via USRP-users wrote: >> Hi all, >> >> I'm still having issues with a B210 with sample rates greater than >> 28MS/s. At 28 I get a few overruns out of a million blocks of 1024 >> samples, but as I sample faster, the number of overruns grow fast. >> >> I have an AMD Ryzen 7 2700 (8 physical cores @ 3.2GHz), 16GB >> RAM, Gigabyte GA-AB350N, Windows 10. The USB driver is >> from http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip . >> >> I've tried adding "num_recv_frames=256" but I get an error: >> >> Creating the usrp device with: num_recv_frames=256... >> [INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_106300; >> UHD_3.11.0.1-37-g2c9087d1 >> [INFO] [B200] Detected Device: B210 >> [INFO] [B200] Operating over USB 3. >> [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: >> LIBUSB_ERROR_CODE -1 >> >> I would have expected to be able to stream with barely any overrun at >> 56MS/s on that computer. I stream at that speed in dual channel on a >> LimeSDR (in SC12 format). >> >> Any advice on how to get better results? >> >> Thanks. >> > You cannot stream faster than about 30Msps from the B210 when in > dual-channel mode. This is a limitation of the data bus architecture on > the AD9371 chip, and not strictly a limitation of the B2xx FPGA. > > > > > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com >

R

RizThon

Wed, Jul 25, 2018 3:42 AM

I have used the b2x0 radios for some applications myself, we have had no
luck sampling much higher than 30MS/s. I think real-world USB3 speeds max
out around 100MB/s.

Will

Thanks Will, that matches my experience. At 28MS/s it tells me I need to
stream at 112MB/s.

You cannot stream faster than about 30Msps from the B210 when in

dual-channel mode. This is a limitation of the data bus architecture on the
AD9371 chip, and not strictly a limitation of the B2xx FPGA.

Sorry for the misunderstanding. On the B210, I'm in single channel, RX
only, and I start having overruns at 28MS/s.

I was just giving my experience with a Lime where I can stream in dual
channel, RX only, at 56MS/s. In that case though the Lime sends its data
back as SC12, meaning it can pack a bit more samples.

FYI, we fixed a couple of sc12/8 related issues on our latest master

branch. As Marcus says, if you're doing dual channel on the B210, you
can only go to 30.72 Msps though.

Should I play with the OTW format? Does the B210 handle SC12 (
https://files.ettus.com/manual/structuhd_1_1stream__args__t.html mentions Only
some devices for SC12).

I tried rx_samples_to_file.exe from
http://files.ettus.com/binaries/uhd/latest_release/Windows7/uhd_3.12.0.0-release_Win32_VS2014.exe
. If I choose SC8, then I can stream faster. It seems I can go up to 49MS/s
without error, over 20 seconds.

At 50MS/s and up I get an error
rx_samples_to_file.exe --null --time 20 --freq 2450e6 --gain 50 --spb 10000
--rate 50e6 --wirefmt sc8
[ERROR] [STREAMER] The receive packet handler caught a value exception.
ValueError: bad vrt header or packet fragment
Error: Receiver error: ERROR_CODE_BAD_PACKET

I modified the code to allow --wirefmt sc12, I also get an error, even at
28MS/s
[ERROR] [STREAMER] The receive packet handler caught a value exception.
ValueError: bad vrt header or packet fragment
but again I'm not sure if the B210 handles that format .

UHD 3.13 is just out, so I tried
http://files.ettus.com/binaries/uhd/uhd_003.013.000.000-release/Windows-7-x86_64/uhd_3.13.0.0-release_Win32_VS2014.exe
. I now get 2850 overruns at 28MS/s in SC16, while I'd get up to 5 with the
previous version. In SC8 I can go up to 40MS/s only, instead of 49. With
the x64 version I get similar results. So for now I reverted back to 3.12.

> > I have used the b2x0 radios for some applications myself, we have had no > luck sampling much higher than 30MS/s. I think real-world USB3 speeds max > out around 100MB/s. > > Will Thanks Will, that matches my experience. At 28MS/s it tells me I need to stream at 112MB/s. You cannot stream faster than about 30Msps from the B210 when in > dual-channel mode. This is a limitation of the data bus architecture on the > AD9371 chip, and not strictly a limitation of the B2xx FPGA. Sorry for the misunderstanding. On the B210, I'm in *single channel*, RX only, and I start having overruns at 28MS/s. I was just giving my experience with a Lime where I can stream in dual channel, RX only, at 56MS/s. In that case though the Lime sends its data back as SC12, meaning it can pack a bit more samples. FYI, we fixed a couple of sc12/8 related issues on our latest master > branch. As Marcus says, if you're doing dual channel on the B210, you > can only go to 30.72 Msps though. > Should I play with the OTW format? Does the B210 handle SC12 ( https://files.ettus.com/manual/structuhd_1_1stream__args__t.html mentions *Only some devices* for SC12). I tried rx_samples_to_file.exe from http://files.ettus.com/binaries/uhd/latest_release/Windows7/uhd_3.12.0.0-release_Win32_VS2014.exe . If I choose SC8, then I can stream faster. It seems I can go up to 49MS/s without error, over 20 seconds. At 50MS/s and up I get an error rx_samples_to_file.exe --null --time 20 --freq 2450e6 --gain 50 --spb 10000 --rate 50e6 --wirefmt sc8 [ERROR] [STREAMER] The receive packet handler caught a value exception. ValueError: bad vrt header or packet fragment Error: Receiver error: ERROR_CODE_BAD_PACKET I modified the code to allow --wirefmt sc12, I also get an error, even at 28MS/s [ERROR] [STREAMER] The receive packet handler caught a value exception. ValueError: bad vrt header or packet fragment but again I'm not sure if the B210 handles that format . UHD 3.13 is just out, so I tried http://files.ettus.com/binaries/uhd/uhd_003.013.000.000-release/Windows-7-x86_64/uhd_3.13.0.0-release_Win32_VS2014.exe . I now get 2850 overruns at 28MS/s in SC16, while I'd get up to 5 with the previous version. In SC8 I can go up to 40MS/s only, instead of 49. With the x64 version I get similar results. So for now I reverted back to 3.12.

MD

Marcus D. Leech

Wed, Jul 25, 2018 3:55 AM

On 07/24/2018 11:42 PM, RizThon via USRP-users wrote:

 I have used the b2x0 radios for some applications myself, we have
 had no luck sampling much higher than 30MS/s. I think real-world
 USB3 speeds max out around 100MB/s.

 Will

Thanks Will, that matches my experience. At 28MS/s it tells me I need
to stream at 112MB/s.

 You cannot stream faster than about 30Msps from the B210 when in
 dual-channel mode.  This is a limitation of the data bus
 architecture on  the AD9371 chip, and not strictly a limitation of
 the B2xx FPGA.

Sorry for the misunderstanding. On the B210, I'm in single channel,
RX only, and I start having overruns at 28MS/s.

You can try using the "num_recv_frames" option in the device
argument--I'd suggest trying 128 and 256.

I'll also note that, near as I can tell, LimeSDR drivers dont' reliably
report overruns, so you may be getting them without knowing about it.

I was just giving my experience with a Lime where I can stream in dual
channel, RX only, at 56MS/s. In that case though the Lime sends its
data back as SC12, meaning it can pack a bit more samples.

 FYI, we fixed a couple of sc12/8 related issues on our latest master
 branch. As Marcus says, if you're doing dual channel on the B210, you
 can only go to 30.72 Msps though.

Should I play with the OTW format? Does the B210 handle SC12
(https://files.ettus.com/manual/structuhd_1_1stream__args__t.html
mentions/Only some devices/for SC12).

I tried rx_samples_to_file.exe from
http://files.ettus.com/binaries/uhd/latest_release/Windows7/uhd_3.12.0.0-release_Win32_VS2014.exe
. If I choose SC8, then I can stream faster. It seems I can go up to
49MS/s without error, over 20 seconds.

At 50MS/s and up I get an error
rx_samples_to_file.exe --null --time 20 --freq 2450e6 --gain 50 --spb
10000 --rate 50e6 --wirefmt sc8
[ERROR] [STREAMER] The receive packet handler caught a value exception.
ValueError: bad vrt header or packet fragment
Error: Receiver error: ERROR_CODE_BAD_PACKET

I modified the code to allow --wirefmt sc12, I also get an error, even
at 28MS/s
[ERROR] [STREAMER] The receive packet handler caught a value exception.
ValueError: bad vrt header or packet fragment
but again I'm not sure if the B210 handles that format .

UHD 3.13 is just out, so I tried
http://files.ettus.com/binaries/uhd/uhd_003.013.000.000-release/Windows-7-x86_64/uhd_3.13.0.0-release_Win32_VS2014.exe
. I now get 2850 overruns at 28MS/s in SC16, while I'd get up to 5
with the previous version. In SC8 I can go up to 40MS/s only, instead
of 49. With the x64 version I get similar results. So for now I
reverted back to 3.12.

USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

On 07/24/2018 11:42 PM, RizThon via USRP-users wrote: > > I have used the b2x0 radios for some applications myself, we have > had no luck sampling much higher than 30MS/s. I think real-world > USB3 speeds max out around 100MB/s. > > Will > > Thanks Will, that matches my experience. At 28MS/s it tells me I need > to stream at 112MB/s. > > You cannot stream faster than about 30Msps from the B210 when in > dual-channel mode. This is a limitation of the data bus > architecture on the AD9371 chip, and not strictly a limitation of > the B2xx FPGA. > > > Sorry for the misunderstanding. On the B210, I'm in *single channel*, > RX only, and I start having overruns at 28MS/s. You can try using the "num_recv_frames" option in the device argument--I'd suggest trying 128 and 256. I'll also note that, near as I can tell, LimeSDR drivers dont' reliably report overruns, so you may be getting them without knowing about it. > > I was just giving my experience with a Lime where I can stream in dual > channel, RX only, at 56MS/s. In that case though the Lime sends its > data back as SC12, meaning it can pack a bit more samples. > > FYI, we fixed a couple of sc12/8 related issues on our latest master > branch. As Marcus says, if you're doing dual channel on the B210, you > can only go to 30.72 Msps though. > > > Should I play with the OTW format? Does the B210 handle SC12 > (https://files.ettus.com/manual/structuhd_1_1stream__args__t.html > mentions/Only some devices/for SC12). > > I tried rx_samples_to_file.exe from > http://files.ettus.com/binaries/uhd/latest_release/Windows7/uhd_3.12.0.0-release_Win32_VS2014.exe > . If I choose SC8, then I can stream faster. It seems I can go up to > 49MS/s without error, over 20 seconds. > > At 50MS/s and up I get an error > rx_samples_to_file.exe --null --time 20 --freq 2450e6 --gain 50 --spb > 10000 --rate 50e6 --wirefmt sc8 > [ERROR] [STREAMER] The receive packet handler caught a value exception. > ValueError: bad vrt header or packet fragment > Error: Receiver error: ERROR_CODE_BAD_PACKET > > I modified the code to allow --wirefmt sc12, I also get an error, even > at 28MS/s > [ERROR] [STREAMER] The receive packet handler caught a value exception. > ValueError: bad vrt header or packet fragment > but again I'm not sure if the B210 handles that format . > > UHD 3.13 is just out, so I tried > http://files.ettus.com/binaries/uhd/uhd_003.013.000.000-release/Windows-7-x86_64/uhd_3.13.0.0-release_Win32_VS2014.exe > . I now get 2850 overruns at 28MS/s in SC16, while I'd get up to 5 > with the previous version. In SC8 I can go up to 40MS/s only, instead > of 49. With the x64 version I get similar results. So for now I > reverted back to 3.12. > > > > > > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

R

RizThon

Wed, Jul 25, 2018 5:56 AM

You can try using the "num_recv_frames" option in the device argument--I'd
suggest trying 128 and 256.

I get tons of errors right after seeing [INFO] [B200] Operating over
USB 3. Would
this explain why I can't seem to stream fast?

Here's what I get for 128
Creating the usrp device with: num_recv_frames=128...
[INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_105900;
UHD_3.12.0.0-release
[INFO] [B200] Detected Device: B210
[INFO] [B200] Operating over USB 3.
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1
[...]
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1
[INFO] [B200] Initialize CODEC control...
[...]

With 256 the program crashes and I only see
Creating the usrp device with: num_recv_frames=256...
[INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_105900;
UHD_3.12.0.0-release
[INFO] [B200] Detected Device: B210
[INFO] [B200] Operating over USB 3.
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1
[ERROR] [USB] libusb_sessi

Can this be a USB driver problem? I'm using the one from
http://files.ettus.com/binaries/misc/ dated 29-Mar-2016 19:52.

Can I run any test to make sure everything is fine (B210, drivers,
software)?

I'll also note that, near as I can tell, LimeSDR drivers dont' reliably

report overruns, so you may be getting them without knowing about it.

I always check the timestamps returned, so I'm pretty sure I only get very
few overruns with the Lime.

> > You can try using the "num_recv_frames" option in the device argument--I'd > suggest trying 128 and 256. > I get tons of errors right after seeing [INFO] [B200] Operating over USB 3. Would this explain why I can't seem to stream fast? Here's what I get for 128 Creating the usrp device with: num_recv_frames=128... [INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_105900; UHD_3.12.0.0-release [INFO] [B200] Detected Device: B210 [INFO] [B200] Operating over USB 3. [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: LIBUSB_ERROR_CODE -1 [...] [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: LIBUSB_ERROR_CODE -1 [INFO] [B200] Initialize CODEC control... [...] With 256 the program crashes and I only see Creating the usrp device with: num_recv_frames=256... [INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_105900; UHD_3.12.0.0-release [INFO] [B200] Detected Device: B210 [INFO] [B200] Operating over USB 3. [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: LIBUSB_ERROR_CODE -1 [ERROR] [USB] libusb_sessi Can this be a USB driver problem? I'm using the one from http://files.ettus.com/binaries/misc/ dated 29-Mar-2016 19:52. Can I run any test to make sure everything is fine (B210, drivers, software)? I'll also note that, near as I can tell, LimeSDR drivers dont' reliably > report overruns, so you may be getting them without knowing about it. > I always check the timestamps returned, so I'm pretty sure I only get very few overruns with the Lime.

MD

Marcus D. Leech

Wed, Jul 25, 2018 2:57 PM

On 07/25/2018 01:56 AM, RizThon wrote:

 You can try using the "num_recv_frames" option in the device
 argument--I'd suggest trying 128 and 256.

I get tons of errors right after seeing [INFO] [B200] Operating over
USB 3. Would this explain why I can't seem to stream fast?

Here's what I get for 128
Creating the usrp device with: num_recv_frames=128...
[INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_105900;
UHD_3.12.0.0-release
[INFO] [B200] Detected Device: B210
[INFO] [B200] Operating over USB 3.
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1
[...]
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1
[INFO] [B200] Initialize CODEC control...
[...]

With 256 the program crashes and I only see
Creating the usrp device with: num_recv_frames=256...
[INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_105900;
UHD_3.12.0.0-release
[INFO] [B200] Detected Device: B210
[INFO] [B200] Operating over USB 3.
[ERROR] [USB] libusb_session_impl::libusb_event_handler_task:
LIBUSB_ERROR_CODE -1
[ERROR] [USB] libusb_sessi

Can this be a USB driver problem? I'm using the one from
http://files.ettus.com/binaries/misc/ dated 29-Mar-2016 19:52.

Can I run any test to make sure everything is fine (B210, drivers,
software)?

 I'll also note that, near as I can tell, LimeSDR drivers dont'
 reliably report overruns, so you may be getting them without
 knowing about it.

I always check the timestamps returned, so I'm pretty sure I only get
very few overruns with the Lime.

Presuming you followed the instructions here for UHD installation:

https://files.ettus.com/manual/page_install.html

Everything should be fine.

I can't remember if we covered this or not. Are you externally-powering
the B210? Some USB ports don't quite provide enough power.

On 07/25/2018 01:56 AM, RizThon wrote: > > You can try using the "num_recv_frames" option in the device > argument--I'd suggest trying 128 and 256. > > > I get tons of errors right after seeing [INFO] [B200] Operating over > USB 3. Would this explain why I can't seem to stream fast? > > Here's what I get for 128 > Creating the usrp device with: num_recv_frames=128... > [INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_105900; > UHD_3.12.0.0-release > [INFO] [B200] Detected Device: B210 > [INFO] [B200] Operating over USB 3. > [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: > LIBUSB_ERROR_CODE -1 > [...] > [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: > LIBUSB_ERROR_CODE -1 > [INFO] [B200] Initialize CODEC control... > [...] > > With 256 the program crashes and I only see > Creating the usrp device with: num_recv_frames=256... > [INFO] [UHD] Win32; Microsoft Visual C++ version 14.0; Boost_105900; > UHD_3.12.0.0-release > [INFO] [B200] Detected Device: B210 > [INFO] [B200] Operating over USB 3. > [ERROR] [USB] libusb_session_impl::libusb_event_handler_task: > LIBUSB_ERROR_CODE -1 > [ERROR] [USB] libusb_sessi > > Can this be a USB driver problem? I'm using the one from > http://files.ettus.com/binaries/misc/ dated 29-Mar-2016 19:52. > > Can I run any test to make sure everything is fine (B210, drivers, > software)? > > I'll also note that, near as I can tell, LimeSDR drivers dont' > reliably report overruns, so you may be getting them without > knowing about it. > > > I always check the timestamps returned, so I'm pretty sure I only get > very few overruns with the Lime. Presuming you followed the instructions here for UHD installation: https://files.ettus.com/manual/page_install.html Everything should be fine. I can't remember if we covered this or not. Are you externally-powering the B210? Some USB ports don't quite provide enough power.

R

RizThon

Wed, Jul 25, 2018 4:18 PM

On Wed, Jul 25, 2018 at 10:57 PM Marcus D. Leech mleech@ripnet.com wrote:

Presuming you followed the instructions here for UHD installation:

https://files.ettus.com/manual/page_install.html

Everything should be fine.

Yes I installed
http://files.ettus.com/binaries/uhd/latest_release/Windows7/uhd_3.12.0.0-release_Win32_VS2014.exe
as well as the driver
http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip
In the device manager, I see it under USRPs / Ettus Research LLC B200/B210

I can't remember if we covered this or not. Are you externally-powering
the B210? Some USB ports don't quite provide enough power.

I only tried it bus-powered.

Unfortunately it seems my power supply doesn't work well. When I plug it,
the power led turns on then off. If I then connect the USB, it's still off.

What's the min/max voltage I can use? Maybe I can use a different power
supply.

Thanks.

On Wed, Jul 25, 2018 at 10:57 PM Marcus D. Leech <mleech@ripnet.com> wrote: > Presuming you followed the instructions here for UHD installation: > > https://files.ettus.com/manual/page_install.html > > Everything should be fine. > Yes I installed http://files.ettus.com/binaries/uhd/latest_release/Windows7/uhd_3.12.0.0-release_Win32_VS2014.exe as well as the driver http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip In the device manager, I see it under *USRPs / Ettus Research LLC B200/B210* > I can't remember if we covered this or not. Are you externally-powering > the B210? Some USB ports don't quite provide enough power. > I only tried it bus-powered. Unfortunately it seems my power supply doesn't work well. When I plug it, the power led turns on then off. If I then connect the USB, it's still off. What's the min/max voltage I can use? Maybe I can use a different power supply. Thanks.

MD

Marcus D. Leech

Wed, Jul 25, 2018 4:36 PM

On 07/25/2018 12:18 PM, RizThon wrote:

On Wed, Jul 25, 2018 at 10:57 PM Marcus D. Leech <mleech@ripnet.com
mailto:mleech@ripnet.com> wrote:

 Presuming you followed the instructions here for UHD installation:

 https://files.ettus.com/manual/page_install.html

 Everything should be fine.

Yes I installed
http://files.ettus.com/binaries/uhd/latest_release/Windows7/uhd_3.12.0.0-release_Win32_VS2014.exe
as well as the driver
http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip
In the device manager, I see it under /USRPs / Ettus Research LLC
B200/B210/

 I can't remember if we covered this or not.  Are you
 externally-powering the B210?   Some USB ports don't quite provide
 enough power.

I only tried it bus-powered.

Unfortunately it seems my power supply doesn't work well. When I plug
it, the power led turns on then off. If I then connect the USB, it's
still off.

What's the min/max voltage I can use? Maybe I can use a different
power supply.

Thanks.

Use a 5-6V supply capable of 2 or 3 amps.

On 07/25/2018 12:18 PM, RizThon wrote: > > > On Wed, Jul 25, 2018 at 10:57 PM Marcus D. Leech <mleech@ripnet.com > <mailto:mleech@ripnet.com>> wrote: > > Presuming you followed the instructions here for UHD installation: > > https://files.ettus.com/manual/page_install.html > > Everything should be fine. > > Yes I installed > http://files.ettus.com/binaries/uhd/latest_release/Windows7/uhd_3.12.0.0-release_Win32_VS2014.exe > as well as the driver > http://files.ettus.com/binaries/misc/erllc_uhd_winusb_driver.zip > In the device manager, I see it under /USRPs / Ettus Research LLC > B200/B210/ > > I can't remember if we covered this or not. Are you > externally-powering the B210? Some USB ports don't quite provide > enough power. > > I only tried it bus-powered. > > Unfortunately it seems my power supply doesn't work well. When I plug > it, the power led turns on then off. If I then connect the USB, it's > still off. > > What's the min/max voltage I can use? Maybe I can use a different > power supply. > > Thanks. Use a 5-6V supply capable of 2 or 3 amps.

AL

Anon Lister

Wed, Jul 25, 2018 4:52 PM

Used to have to compile the driver for the lime in debug mode to get
underrun reporting. Though I believe they may have changed some of that
recently.

On Tue, Jul 24, 2018, 23:57 Marcus D. Leech via USRP-users <
usrp-users@lists.ettus.com> wrote:

On 07/24/2018 11:42 PM, RizThon via USRP-users wrote:

I have used the b2x0 radios for some applications myself, we have had no

luck sampling much higher than 30MS/s. I think real-world USB3 speeds max
out around 100MB/s.