| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
- Created AD936x manager class
- Moved functionality from B2x0 and E310 into manager
- Separated property tree + perifs initialization in both device classes
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes issue #828 "B200: Tx quadrature calibration regression in master"
Following commit added new gain table settings to reflect updated values
from ADI. Gain indices used by Tx Quad Cal were not matched to
accommodate the new tables.
2b06c38 "b2xx: dc offset and iq imbalance correction control"
Requirement for Tx Quad Cal is for TIA gain and analog LPF gain to be
set at 0 dB, or 0x20 in the gain table. Final effect is a dramatic
decrease in Tx DC offset and quadrature image.
Signed-off-by: Tom Tsou <tom.tsou@ettus.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Patch fixes a portion of #807
"B210: severe distortion on In-phase data for some gain settings"
ADI recommends that the "Prevent Pos Loop Gain" setting be enabled to
prevent the Rx quadrature tracking loop from becoming unstable at low
power levels. ADI Linux kernel driver also reflects this setting.
We do not follow the ADI recommendation. Adjust accordingly.
Signed-off-by: Tom Tsou <tom.tsou@ettus.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch resolves issue #823
"B200: Receive RF DC calibration makes calibration worse below 34 dB"
According to ADI reference documents, enabling any of the 3 LNA's in the
receive path causes a 180 degree phase shift. Correspondingly, we invert
the LNA bypass path (gain indices below 34 dB) to match. Testing,
however, reveals that one of these statements or the polarity inversion
setting itself is false. Disabling the switch results in expected
behavior and proper phase alignment.
Overall effect is up to 60 dB of DC offset suppression ahead of the Rx
analog LPF. This reduces the problematic dependency on active baseband
tracking and may resolves multiple tracking stability issues.
Signed-off-by: Tom Tsou <tom.tsou@ettus.com>
|
|\ |
|
| | |
|
| | |
|
|\|
| |
| |
| |
| |
| |
| | |
Conflicts:
host/lib/usrp/common/ad9361_ctrl.hpp
host/lib/usrp/common/ad9361_driver/ad9361_device.h
host/lib/usrp/e300/e300_remote_codec_ctrl.hpp
|
| |
| |
| |
| | |
When the LO is tuned it changes the frequency on both channels. The frequency value read back for the first channel was not updated when the LO frequency for the other channel was tuned to a different value.
|
| | |
|
|\|
| |
| |
| |
| | |
Conflicts:
host/CMakeLists.txt
|
| | |
|
|\|
| |
| |
| |
| |
| |
| | |
Conflicts:
host/lib/usrp/b200/b200_io_impl.cpp
host/lib/usrp/common/ad9361_driver/ad9361_device.cpp
host/lib/usrp/common/ad9361_driver/ad9361_device.h
|
| | |
|
|\|
| |
| |
| |
| | |
Conflicts:
host/CMakeLists.txt
|
| | |
|
|\| |
|
| |
| |
| |
| |
| | |
- Factored out MAX287x code into common header file
- Added necessary code for MAX2871 synchronization.
|
|\| |
|
| |
| |
| |
| |
| | |
This creates a wb_iface child class called timed_wb_iface, which
adds support for timed commands.
|
|\| |
|
| |
| |
| |
| |
| |
| |
| | |
* CMake now not applying C++ flags to C files
* GCC 4.4: anti-aliasing rules
* MSVC: narrowing, differences in subclass function parameters
* Clang: uninitialized variables
|
| |
| |
| |
| | |
RX/TX filter information and writing of analog and FIR filters
|
| | |
|
| |
| |
| |
| |
| |
| | |
- Use ULL and LL etc. so compiler doesn't need to decide constant's types
- define the empty 'while' loop to include {} to make older compilers happy
- Use explicit type names for BOOST_CHECK_EQUAL
|
| | |
|
| | |
|
| | |
|
|\| |
|
| |
| |
| |
| | |
Issues #649 & #691
|
|\|
| |
| |
| |
| | |
Conflicts:
host/examples/rx_samples_to_file.cpp
|
| | |
|
|\|
| |
| |
| |
| | |
Conflicts:
host/docs/usrp_e3x0.dox
|
| |
| |
| |
| |
| |
| | |
- Fix ported from He. The current TX filter does not need
the additional -6dB of headroom. Set it to zero so we
meet our max power specs.
|
| | |
|
|\ \ |
|
| |/ |
|
| | |
|
|/
|
|
| |
This allows to read a relative temperature from an AD9361 device.
|
| |
|
|
|
|
| |
AD9361 driver can now select coeffs for different interpolation ratios.
|
| |
|
|
|
|
| |
- Made the methods in adf4001_ctrl virtual
|
|
|
|
|
|
|
|
|
| |
- Split niriok_proxy interfaces to support NI-RIO <=13.0 and >=14.0 kernel interfaces
- Fixed multi-session race conditions by synchronizing niriok_proxy access
- Fixed bug switching from NI LV-FPGA access to UHD access by changing how devices are hashed into a reservation table
- Fixed calculation of FRAC values for CBX and SBX LO tuning by rounding instead of truncating
- Fixed bug that was not setting two MSBs for band select configuration of CBX LO
- Submitting on behalf of Patrick Sisterhen, Matthew Crymble
|
| |
|
| |
|
| |
|
| |
|
| |
|