Skip to content

v5.0.0
Compare
Choose a tag to compare
@github-actions github-actions released this 29 Mar 19:03
· 448 commits to main since this release
v5.0.0
b95a6dc
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.

Major changes that affect users

Number of subbands changed from 128 to 512

To lower the current draw on the FPGA, the number of channels in the polyphase filter bankw was increased from 128 to 512. This means there are 512 subbands instead of 128. This mainly affects find_freq. We now provide start_freq and stop_freq optional arguments which allow then user to define the frequency range to sweep. This is more convenient to use than defining subbands.

Ref phase delay change

The low current firmware has longer processing times. This means the ref_phase_delay is larger than in past releases. The user can remeasure the full delay of the system with estimate_phase_delay. Note that this only changes the value in the current Rogue session. If the Rogue session is killed, it will not be reloaded. The config file can be updated with this new value. This will set the correct value for phase delay on setup(). The user can manually set the processing delay with set_processing_delay().

Tracking setup change lms_gain to feedback_gain

The optional argument lms_gain is now an “expert only” variable. Use feed_back gain instead. See this jupyter notebook and the docstring for more details.

Example script

Mitch's example script that walks users through using the system.
slaclab/pysmurf

Pull Requests Since v4.3.0

Client

Interface-change

  1. #621 - Find freq eta scan speedup

Bug

  1. #621 - Find freq eta scan speedup
  2. #606 - Fixed no_dir error during setup
  3. #604 - Adds a few register_file calls for noise plots

Enhancement

  1. #621 - Find freq eta scan speedup
  2. #616 - take_noise_psd return_noise_params expanding

Pull Request Details

Find freq eta scan speedup
Author: Edward [email protected]
Date: Wed Mar 24 14:05:36 2021 -0700
Pull: #621 (2144 additions, 217 deletions, 12 files changed)
Branch: slaclab/find-freq-eta-scan-speedup
Labels: bug, client, enhancement, interface-change

Notes:

Issue

This PR updates the pysmurf software for the low current FW.

Description

[Describe what this PR does here]

Does this PR break any interface?

  • Yes
  • No

Which interface changed?

Number of subbands changed from 128 to 512

To lower the current draw on the FPGA, the number of channels in the polyphase filter bankw was increased from 128 to 512. This means there are 512 subbands instead of 128. This mainly affects find_freq. We now provide start_freq and stop_freq optional arguments which allow then user to define the frequency range to sweep. This is more convenient to use than defining subbands.

Ref phase delay change

The low current firmware has longer processing times. This means the ref_phase_delay is larger than in past releases. The user can remeasure the full delay of the system with estimate_phase_delay. Note that this only changes the value in the current Rogue session. If the Rogue session is killed, it will not be reloaded. The config file can be updated with this new value. This will set the correct value for phase delay on setup(). The user can manually set the processing delay with set_processing_delay().

Tracking setup change lms_gain to feedback_gain

The optional argument lms_gain is now an “expert only” variable. Use feed_back gain instead. See this jupyter notebook and the docstring for more details.

What was the interface before the change

[Describe here how was the original interface]

What will be the new interface after the change

[Describe here how is the new interface, and what the difference with the original interface]

take_noise_psd return_noise_params expanding
Author: Edward [email protected]
Date: Fri Mar 19 14:33:52 2021 -0700
Pull: #616 (21 additions, 11 deletions, 1 files changed)
Branch: slaclab/issue614
Issues: #614
Labels: client, enhancement

Notes:

Issue

This resolves #614

Description

Returns all the fit parameters

Does this PR break any interface?

  • Yes
  • No

The dimension of the return parameter is now larger.

Which interface changed?

[Indicate here what is the interface that changed, preferable with a link to the code where the interface is defined]

What was the interface before the change

[Describe here how was the original interface]

What will be the new interface after the change

[Describe here how is the new interface, and what the difference with the original interface]

Fixed no_dir error during setup
Author: Edward [email protected]
Date: Tue Feb 16 15:46:38 2021 -0800
Pull: #606 (49 additions, 49 deletions, 1 files changed)
Branch: slaclab/issue600
Issues: #600
Labels: bug, client

Notes:

Issue

Resolves #600 and https://github.com/simonsobs/smurf-issues/issues/24

Description

Figured out there was an error during setup in this branch that happened because some code was indented once to many times, which caused some of the setup not to be run if no_dir=True. This PR de-indents it.

Does this PR break any interface?

  • Yes
  • No

Adds a few register_file calls for noise plots
Author: Edward [email protected]
Date: Tue Feb 16 15:45:37 2021 -0800
Pull: #604 (12 additions, 0 deletions, 1 files changed)
Branch: slaclab/register_noise_plots
Issues: #603
Labels: bug, client

Notes:

Issue

Issue #603

Description

Adds a few register_file statements to take_noise_psd and correctly sets the publisher action.

I'm not exactly sure which branch I should merge this into... I think we have some people running on the main branch, but if find-freq-eta-scan-speedup will be merged soon then I think it's ok.

Does this PR break any interface?

  • Yes
  • No
Assets 2
Loading