https://github.com/diamondlightsource/dls-mbf
Diamond Multi-Bunch Feedback Processor Firmware and Software
https://github.com/diamondlightsource/dls-mbf
Last synced: 22 days ago
JSON representation
Diamond Multi-Bunch Feedback Processor Firmware and Software
- Host: GitHub
- URL: https://github.com/diamondlightsource/dls-mbf
- Owner: DiamondLightSource
- Created: 2018-03-12T10:58:58.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2025-03-11T10:40:09.000Z (11 months ago)
- Last Synced: 2025-05-01T12:44:06.522Z (10 months ago)
- Language: VHDL
- Size: 7.69 MB
- Stars: 3
- Watchers: 7
- Forks: 0
- Open Issues: 3
-
Metadata Files:
- Readme: README.rst
Awesome Lists containing this project
README
Diamond Light Source Multi-Bunch Feedback Processor (MBF)
=========================================================
This repository contains the firmware and software sources for running bunch by
bunch feedback on a synchrotron. The firmware and software was developed at
Diamond Light Source and is designed to run on specific hardware documented
below.
Full documentation for using the system can be found at the `MBF Documentation`_
pages. See the `Bringing up MBF`_ page for instructions on setting up MBF for
first operation.
.. _MBF Documentation: https://diamondlightsource.atlassian.net/wiki/x/GAAN
.. _Bringing up MBF: https://diamondlightsource.atlassian.net/wiki/x/FgAN
LICENSING
---------
The firmware in this repository (all files under the ``AMC525`` directory) has
been developed using commercial tools made available through the Europractice_
service. As a consequence, this firmware is only available subject to an
agreement with Diamond Light Source that the firmware and its source may not be
redistributed and may not be commercially exploited.
This restriction does not apply to any of the other files in this repository.
.. _Europractice: http://www.europractice.stfc.ac.uk/welcome.html
MBF Components
--------------
This repository contains the following components.
Firmware
The firmware is in the ``AMC525`` directory. The targeted hardware is an
AMC525 MicroTCA carrier card with an FMC500 ADC/DAC digitiser, and the
firmware is built with Vivado 2019.2.
Kernel Driver
The control system is designed to be run on a separate processor card on the
same backplane as the AMC525 carrier card. The register interface to the
firmware is managed by a kernel driver in the ``driver`` directory.
EPICS Driver
The control system is implemented as an EPICS IOC, all sources in the
``epics`` directory. Note that currently only EPICS 3.14 is supported. To
build the EPICS driver and run the helper GUI scripts the following
dependencies are required:
=============== ======= ====================================================
Component Version Download from
=============== ======= ====================================================
EPICS 3.14 https://epics.anl.gov/base/index.php
epics_device 2.0 https://github.com/Araneidae/epics_device
epicsdbbuilder 1.2 https://github.com/Araneidae/epicsdbbuilder
cothread 2.14 https://github.com/dls-controls/cothread
=============== ======= ====================================================
Note that epics_device must be at least the version shown.
Tune Fitter
A separate EPICS IOC is provided for computing machine tune and sidebands
from the result of tune sweeps. This IOC is written in Python and uses the
following tool:
=============== ======= ====================================================
Component Version Download from
=============== ======= ====================================================
pythonIoc 2.15 https://github.com/Araneidae/pythonIoc
=============== ======= ====================================================
Release Notes
-------------
1.0.0 May 2018
..............
This was the first working release deployed at DLS.
1.0.1 June 2018
...............
Minor bug fixes and refinements.
1.1.0 August 2018
.................
Bug fixes, some structural renaming: at this point LMBF was renamed to MBF
throughout the project.
1.1.2 January 2019
..................
Minor changes:
* Some improvements to the tune fitting algorithm to improve robustness.
* Add selector to tune fitter to just take maximum value as proxy for peak.
* Bunch-by-bunch control waveform editing has been improved.
1.2.0 March 2019
................
The firmware for this release is strictly incompatible with the 1.0 and 1.1
releases, and this is now checked during startup. The following features have
been added:
* Tune PLL tracking is now available. This is the major change in this release.
* NCO frequencies are now 48 bits wide (rather than 32 bits). This means that
frequency resolution will no longer be an issue anywhere.
* Bunch by bunch gain control now has option for up to +30dB of extra gain.
* New bunch rejection filter added for detector data.
* Two changes to sequencer: now have optional holdoff at start of dwell, and can
optionally apply tune PLL offset as extra offset to sweep frequency.
1.2.1 June 2019
...............
Minor tweaks, no change to firmware. ``mbf_memory`` Python library merged.
``mbf_read_tune_pll.m`` script rewritten and API changed. Minor PV changes:
* Remove ``BUN:n:ALL:SET_S`` PV: too easy to press by mistake and really
somewhat redundant.
* Change semantics of ``ADC:THRESHOLD_S`` (LMBF only).
1.3.0 September 2019
....................
Mostly small changes to tune fitting, however the graphs have changed to show
magnitude and phase rather than power, and data weighting during fit is
configurable.
One incompatible change is to correct the calculation of detector phase, which
now correctly advances in the negative direction with increasing frequency.
This affects the detector IQ and phase waveforms as well as the tune fitting.
This software is compatible with version 1.2.0 of the firmware which has not
changed.
1.3.1 January 2020
..................
Move help pages to public confluence server. No other change.
1.4.0 February 2020
...................
This is a place-holder release, and is essentially untested. This release will
either be followed with bug fixes or with further significant firmware changes.
Major firmware and control system changes, implementing the following:
* New DAC MMS trigger source available.
* Generated FIR data now available as input to DAC MMS, and for trigger
generation.
* DRAM FIR source now comes from output scaled FIR data, so the memory FIR gain
control and overflow events have been removed.
* Add second NCO, rename original NCO so now have NCO1 and NCO2.
* Add option for fine gain control over all four NCOs (NCO1, NCO2, SEQ, PLL).
* Add option for tune PLL offset tracking to NCO1 and NCO2.
* Increase available FIR gain by 3dB and recalibrate gain settings so that 0dB
corresponds to true unity gain with a trivial unity FIR. This means that
gain settings on 1.3 will need to be reduced by 12dB for 1.4.
* Controls are provided to make it easy to copy bunch banks.
* Super sequencer frequency offsets are now 48 bits.
Note that the firmware is now built with Vivado 2019.2.
1.4.1 March 2020
................
This release will be deployed at Diamond for the run starting at the end of
March 2020. This release incorporates a major change to the bunch by bunch
control of output data.
The firmware for this version is incompatible with earlier releases. The
following changes have been implemented:
* Separate bunch by bunch gain controls for FIR and all four NCOs.
* Tango support has been moved into a separate repository, MBF-Tango_.
* The required version of epics_device has been bumped to 2.0.
* The lowest FIR gain control setting now sets FIR output to zero.
* The screens now refer to the ADC and DAC FIRs as the Compensation (COMP) and
Pre-Emphasis (PEMPH) filters respectively to reduce confusion with the bunch
by bunch FIR.
1.4.2 May 2020
..............
This release contains a bug fix to the string displayed in the ``:SEQ:MODE`` PV.
Other changes include making the core Python scripts work with Python 3 and
moving the ``mbf_memory`` module to a separate repository PyMBF_read_. There
are no firmware changes in this release.
.. _MBF-Tango: https://github.com/DLS-Controls-Private-org/MBF-Tango
.. _PyMBF_read: https://github.com/DLS-Controls-Private-org/PyMBF_read