| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Signed-off-by: Josh Morman <jmorman@gnuradio.org>
|
|
|
|
| |
Signed-off-by: Liu, Andrew Z <liu.andrew@vast-inc.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The DC offset and IQ balance functions were incorrectly implemented in
the GRC blocks: It was possible to force-enable, but not force-disable
them. Also, the underlying Python/C++ API allowed changing the actual
correction value, but the GRC did not.
Now, there are four options for IQ and DC:
- Default: Whatever UHD does by default
- Auto: Enable auto-correction
- Disabled: No correction
- Manual: Allows setting the correction value separately
Signed-off-by: Martin Braun <martin@gnuradio.org>
|
|
|
|
|
|
| |
also reorder USRP initialization to match Python template
Signed-off-by: karel <5636152+karel@users.noreply.github.com>
|
| |
|
|
|
|
|
|
| |
Gain type hiding was inverted: It was supposed to hide gain type for
default (absolute/dB), and not hide for normalized and absolute/dBm.
Fixed by this commit.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
UHD 4.0 is introducing a power reference level API, which will now
become available in GNU Radio. Its availability is conditionally
compiled into GNU Radio based on whether or not the feature is available
in the UHD version GNU Radio was compiled against.
USRP source and sink blocks receive the following new APIs:
- has_power_reference()
- set_power_reference()
- get_power_reference()
The command interface now understands a new `power_dbm` dictionary key.
The GRC bindings can now switch between absolute gain, normalized gain,
and absolute power settings.
If GNU Radio was compiled against an older version of UHD (prior to UHD
4.0) which doesn't have this API, an error or warning is printed, the
has_power_reference() will return false, and the others will throw an
exception.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
A USRP block offers the option to set a `clock_rate` which in turn sets
`master_clock_rate` in UHD. There were some common options
(153.6MHz, 125MHz, 122.88MHz) missing for N310s. Now these options are added.
Instead of a separate call to `set_clock_rate`, `master_clock_rate` is
now set as part of the device address string. This behavior makes it
more consistent with UHD examples. Flowgraphs that set MCR as part of
their `device_args` option, are currently overruled if clock rate is set
too. This behavior may easily be reversed.
According to the UHD manual, X300s don't support
`set_master_clock_rate`, thus in this case it is strictly necessary to
set MCR as part of the device address string.
|
|
|
|
|
| |
With pybind11 `uhd.time_spec()` doesn't work anymore because default
values are not defined.
|
|
|
|
|
|
|
|
|
| |
Instead of relying on the pyuhd pybind11 bindings, create new gr::uhd
pybind11 bindings that wrap the underlying types. Trying to use the
pyuhd bindings directly has many issues
Also, push some of the overloading into the constructors as opposed to
in the __init__ functions with constructor interception
|
| |
|
| |
|
| |
|
|
|
|
| |
- Adds set_auto_dc_offset and set_auto_iq_balance calls to GRC bindings
|
|
|
|
|
|
|
| |
- Duplicated existing logic for Python generation to create updated
*.yml to produce C++ code for USRP source and sink.
- Updated to handle either string literals or string parameters
- Concatenate dev_addr and dev_args as std::string
|
|
|
|
|
|
|
| |
In the 3.8 gen_uhd_usrp_blocks.py generated an invalid statement for the
set_gain callbacks, so they were never rendered to the output py.
Fixes #2791
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Due to the way the template interprets the stream_chans value as string,
it always returns True and the '[]' from stream chans was used as the
channels parameter.
With this change, channels in the case where stream chans is empty,
chans populates as before, which is range(num_chans)
python3 outputs a range object, so it is necessary to add a list
surrounding the range command as well
Fixes #2789
|
|
|
|
| |
This adds the set_rx_agc() call from UHD to the USRP source.
|
|
|
|
|
|
| |
For some attributes in the GRC bindings, we'd skip the setters at make()
time when no value was given. This also adds this behaviour for the
antenna.
|
|
|
|
|
|
|
|
| |
- The callback for gain would always use normalized gain, which would
cause errors since the non-normalized gain range would cause
exceptions
- The "Show Command Port" menu was removed, it didn't work, but no other
block has this.
|
|
|
|
|
|
|
|
|
| |
- Subdev spec wasn't being applied at all
- time/clock source, neither
- Sync wouldn't stay the same value when it was set to "don't sync"
- Bandwidth is no longer set when the value is 0, allowing for a default
behaviour
- OTW format wasn't being applied properly
|
|
|
|
|
| |
The YML file generated during CMake was simply not correct. Frequency,
gain, etc. weren't propagated properly to the underlying C++ object.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
UHD asynchronous messages are needed in order to obtain information
about issues happening on Tx side (Underflows, Sequence Errors,
Time Errors).
This change adds handling of these messages to the usrp sink by adding
async_event_loop() function executed in a separate thread (_async_event_thread)
where these messages are:
- received from the USRP,
- transformed into PMT messages,
- sent to the new "async_msgs" message output port.
On the top level the output PMT messages are (message_type, content) pairs:
- message_type is always "uhd_async_msg",
- content is a dictionary with key names taken from
uhd::async_metadata_t.
- there is no 'has_time_spec' as in PMT it's not needed.
There just won't be 'time_spec' in the PMT message
when 'has_time_spec' is false.
- there is no 'user_payload' - but it can be easily added in the
future as pmt blob.
Example message from the 'async_msgs' output is below:
(uhd_async_msg (channel . 0) (time_spec 4 . 0.0608192) (event_code burst_ack))
Reviewed-By: Martin Braun <martin.braun@ettus.com>
|
| |
|
|
|
|
|
|
|
|
|
| |
- Show/Hide LO settings now works the right way round
- Gain Type changed from bool to enum
- Fix Pylint warning on the YML generation code
Reviewed-by: Marcus Müller <mmueller@gnuradio.org>
Reviewed-by: Håkon Vagsether <hauk142@gmail.com>
|
| |
|
| |
|
|\
| |
| |
| |
| | |
Conflicts:
gr-uhd/grc/gen_uhd_usrp_blocks.py
|
| |
| |
| |
| |
| |
| | |
GRC bindings for the USRP sink and source suggest some antenna names.
This adds the antenna name RX1 (TwinRX) into the list of antenna name
suggestions.
|
|/ |
|
| |
|
|
|
|
|
| |
This reverts commit 8820340ae2cb119e177a445f8d9b079759ed2dd7, reversing
changes made to 71a7aeb97b215210c8ee1148b2d1a96b0b29e97f.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
This reverts commit 5d29a17c12d19ebc5f4494b8ae1be77568b235e7.
Well, also fixes a bug where older FGs would throw a template
error.
|
|
|
|
| |
This reverts commit 24bf23d75fb88e77f86f4ddd1990995fd41ecdfa.
|
|
|
|
|
|
|
| |
- Better preselect options
- RF Options moved to own tab
- FE Corrections tab for rx corrections
- sc12 preselect
|
| |
|
| |
|
| |
|
| |
|
|\ |
|
| |
| |
| |
| |
| |
| | |
In some cases, there could be an error when reducing the number
of chans for UHD device in GRC (e.g. going from 2 chans to 1).
This adds another check which will alert the user.
|
| | |
|
|/ |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
From: Timo Juhani Lindfors <timo.lindfors@iki.fi>
Date: Tue, 9 Jul 2013 09:23:06 +0300
Subject: [PATCH] uhd: "scalar" stream arg is now called "peak"
uhd commit d46c176af34b728fd43b3dd46485b38623a7335e (dsp rework: tx
trailer, scaling work (peak)) renamed "scalar" stream arg to "peak"
and changed its semantics. Using 1/256=.003906 captures upper 8-bits.
|