Repository navigation
Conversation
…8/LAB-303) The SiPM detector settings (bias, enable, offsets, bandwidth, temperature) are not in ScanImage or its TIFF header. This adds a node that reads them through the vendor ThorSiPM.dll and publishes them as a connector, so they land in the run log and bag. - nodes/thorsipm.py: ctypes driver, no ROS. Binds only FindDevices, SelectDevice, GetParam, GetParamString, GetParamInfo and TeardownDevice (signatures from the vendor GUI's P/Invoke metadata); no write path. CLI: scan / snapshot / watch. FakeBackend simulates the controller for tests. - nodes/sipm_node.py, nodes/sipm_cmd.py: /sipm/cmd, /sipm/reply, /sipm/status (1 Hz), std_msgs/String JSON like the BIAS and ScanImage connectors; commands ping, status, get, info, scan, disconnect, reconnect. Keeps reporting (and retrying) while the vendor GUI holds the COM port. - bridge_node, mcap_convert, rig (--sipm), pixi tasks (sipm-node, sipm-cmd, sipm, test-sipm): the sipm connector is recorded and derived like the others. - tests/test_sipm.py: 28 checks against the simulated controller (driver, node, busy port, recording). The existing suites still pass (test-frames 39/39, test 54/54). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
mbreiser
changed the title
… into claude/sipm-node
… into claude/sipm-node # Conflicts: # rig/nodes/mcap_convert.py # rig/nodes/sipm_cmd.py # rig/nodes/sipm_node.py # rig/nodes/thorsipm.py # rig/tests/test_sipm.py
This was referenced Oct 27, 2026
Merged
On the real controller the serials came back as '0' and the firmware as '5':
the DLL writes wide strings ("000027" = 30 00 30 00 ...), and buf.value stopped
at the first NUL. Decode the raw buffer as UTF-16LE and allocate 2 x size bytes
in case size counts wide characters. Now reads 000027 / 000026 and 5.4.6, as the
GUI shows. test-sipm still 28/28 on Windows.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
mbreiser
force-pushed
the
claude/scanimage-frames-wrapper
branch
from
October 28, 2026 21:23
a54bd8d to
1d52b0a
Compare
mbreiser
added a commit
that referenced
this pull request
Oct 28, 2026
…nodes layout (LAB-319, draft) The read-only node, its command tool, the ctypes driver and the harness moved to reiserlab/rig-nodes 0.3.0 as rig_nodes.sipm_node / sipm_cmd / thorsipm (marked EXPERIMENTAL there: never run against the controller). What stays here: the /sipm/cmd|reply|status connector in bridge_node (recorded and relayed like BIAS and ScanImage), the converter's derivation into connector events, rig.py --sipm / --sipm-args, the pixi tasks, and two static harness checks. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
mbreiser
force-pushed
the
claude/sipm-node
branch
from
October 28, 2026 21:23
a4442ad to
f4302a0
Compare
mbreiser
changed the base branch from
claude/scanimage-frames-wrapper
to
main
October 28, 2026 21:23
mbreiser
force-pushed
the
claude/sipm-node
branch
from
October 28, 2026 21:44
f4302a0 to
a4442ad
Compare
mbreiser
changed the base branch from
main
to
claude/scanimage-frames-wrapper
October 28, 2026 21:44
5 tasks
5 tasks done
Contributor
|
Closing without merging: the read-only SiPM driver, node, command tool and harness (test-sipm 25/25 there, on the real controller in Block S) now live in the private rig repository. This branch also carried #261, which is closed the same way. What happened to this PR's work:
Michael decided on 2026-10-28 that the imaging work is private; the details are in the §3.3 amendment in 🤖 Generated with Claude Code |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Stacked on #261 (
claude/scanimage-frames-wrapper), so this diff is only the SiPM commit. Retarget tomainonce #261 merges.Implements LAB-319 (split out of LAB-308, ScanImage documentation for agents, Bergamo III); uses the LAB-303 connector pattern.
Why
The SiPM detector settings (bias, enable, offsets, bandwidth, temperature) are not in ScanImage or its TIFF header. This node reads them through the vendor
ThorSiPM.dlland publishes them as a connector, so they land in the run log and bag next to ScanImage and the arena.What
ros2-poc/nodes/thorsipm.py— ctypes driver, no ROS. Binds onlyFindDevices,SelectDevice,GetParam,GetParamString,GetParamInfo,TeardownDevice(signatures from the vendor GUI's own P/Invoke metadata). No write path. CLI:scan/snapshot/watch.FakeBackendsimulates the controller.ros2-poc/nodes/sipm_node.py,sipm_cmd.py—/sipm/cmd,/sipm/reply,/sipm/status(1 Hz),std_msgs/StringJSON like the BIAS and ScanImage connectors. Commands:ping,status,get,info,scan,disconnect,reconnect. Keeps reporting (and retrying) while the vendor GUI holds the COM port.bridge_node.py,mcap_convert.py,rig.py --sipm, pixi tasks (sipm-node,sipm-cmd,sipm,test-sipm) — thesipmconnector is recorded and derived like the others.Open question the rig test answers
The GUI's
Paramsenum gives the IDs for enable, gain, offset, bandwidth, safety, saturations and serial numbers. The DLL also namesPARAM_SIPM1_BIAS,…_TEMPERATURE,…_TEMPSTATUS, but their numeric IDs aren't in the enum. Block S finds them with a read-onlyGetParamInfoscan and writes them toconfig/sipm_params.jsonin scanimage-2026-bergamo.Testing
pixi run test-sipm: 28 / 28 (driver, node, busy port, recording) on macOS.test-frames39 / 39,test54 / 54.--fake: 16"src":"sipm"connector events in the run log, bag closed cleanly, controller released on exit.🤖 Generated with Claude Code