| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
- radio_rst was being asserted to reset the capture iface IDELAYs but
that was excessive and had adverse effects on the rest of the radio
- Replaced radio_rst with a localized IDELAYCTRL reset
|
|
|
|
|
|
| |
* Improved toolchan file, accounts for different versions
* Don't force static runtime linking (was causing test errors)
* Include MinGW runtime in packaging
|
| |
|
|
|
|
| |
This reverts commit a12b24027fe1af9ca51949f6a9333ac5451690ef.
|
|
|
|
|
|
|
|
|
|
|
| |
- The following function implementations were moved from x300_impl.cpp
to x300_adc_dac_utils.cpp
- synchronize_dacs
- self_test_adcs
- extended_adc_test
- self_cal_adc_capture_delay
- self_cal_adc_xfer_delay
- This reduces the size of the x300_impl object file
|
| |
|
|
|
|
| |
- We added additional microcode to configure DCI matching,config rate,etc
|
| |
|
|
|
|
|
|
|
|
| |
- The value was originally empirically determined based on self-cal
results. After the fix for uncalibrated IDELAY, the self-cal offset
data was no longer correct.
- The new delay through the ADC ensures that the self-cal nominally
pick the halfway tap of 16
|
|
|
|
|
|
|
|
|
|
|
| |
- Everytime the LMK is configured, we do the following:
- Reset all LMK regs
- Wait for LMK lock
- Reset radio_clk PLL in FPGA
- Wait for FPGA PLL to lock
- Assert radio_rst which resets downstream radio logic
- This address the intermittent self-cal failures due to uncalibrated IDELAY taps
- Bumped FPGA compat to 12
|
| |
|
|
|
|
| |
- This changed with the ADS62P44 -> ADS62P48 design change
|
| |
|
| |
|
|\ |
|
| |\
| | |
| | |
| | |
| | | |
Conflicts:
host/include/uhd/transport/nirio/nirio_driver_iface.h
|
| | | |
|
| | | |
|
| | |
| | |
| | |
| | | |
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
All currently available e3xx daugherboards are equivalent from
a UHD perspective, so make them consistently show up as "E3XX".
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
|
| | | |
|
| | | |
|
| | |
| | |
| | |
| | |
| | | |
- New device arg "ext_adc_self_test" triggers the test
- ext_adc_self_test=<time> runs the test for "time" seconds
|
| | |
| | |
| | |
| | |
| | | |
- product code, revision and revision compat errors are now exceptions
- Added recover_mb_eeprom arg to recover from a corrupt/uninitialized EEPROM
|
| | |
| | |
| | |
| | |
| | | |
- Added new field "revision_compat" to mb_eeprom
- Enforce a revision_compat of 7
|
|/ /
| |
| |
| |
| | |
- Added new PCIe and MB PIDs for new boards
- Added an error if MB PID is invalid
|
| | |
|
|\| |
|
| |
| |
| |
| | |
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
|
| |
| |
| |
| | |
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| | |
* Single class for loading firmware/FPGA images onto devices instead of multiple utilities
* Loading functions are registered for each device, corresponding to their --args="type=foo" name
* Deprecation warnings added to all product-specific image loading utilities
|
|\| |
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
- X300 Rev7 images
- B200 updated firmware (V8)
- OctoClock
- E300 (idle image)
|
| |
| |
| |
| |
| | |
- B2x0: FW compat number (goes with previous firmware update)
- X3x0: Max HW rev number
|
| | |
|
| | |
|
|\|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Conflicts:
fpga-src
host/CMakeLists.txt
host/cmake/Modules/UHDVersion.cmake
host/lib/usrp/b200/b200_impl.hpp
host/lib/usrp/e300/e300_fpga_defs.hpp
host/lib/usrp/x300/x300_fw_common.h
|
| |
| |
| |
| |
| |
| | |
- Updated fpga-src
- Updated version strings
- Updated images package
|
| | |
|
| |
| |
| |
| | |
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
|
| | |
|
| |
| |
| |
| | |
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
This was originally limited because it performed poor,
however, with refactoring that has been done since release,
this now gives better performance.
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This commit will have UHD load the idle fpga image on
destruction of e300_impl.
Note: This requires usrp_e310_idle_fpga.bit to be present
in the UHD images directory.
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
|
| | |
|
| |
| |
| |
| |
| | |
- If the self-cal fails, UHD waits for 2 sec for the ADC temp
to stabilize and retries the self-cal
|
| |
| |
| |
| | |
- Characterized over process and temperature
|