Tecan Infinite 200 PRO plate reader backend (Infinite M Plex) #797
Conversation
There was a problem hiding this comment.
Pull request overview
This PR introduces a backend for the Tecan Infinite 200 PRO plate reader (M Plex series), adding support for absorbance and fluorescence measurements with experimental luminescence support. The implementation includes configurable well masking for reading specific wells on plates up to 384-well format.
Key changes:
- New
TecanInfinite200ProBackendclass with USB communication and measurement decoding - Specialized decoders for absorbance, fluorescence, and luminescence measurement streams
- Enhanced USB capture error handling to gracefully manage malformed escape sequences
Reviewed changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 5 comments.
| File | Description |
|---|---|
| pylabrobot/plate_reading/tecan_infinite_backend.py | Implements the complete backend with transport layer, measurement decoders, and plate reading methods for all three modes |
| pylabrobot/plate_reading/tecan_infinite_backend_tests.py | Comprehensive unit tests for decoders, scan geometry, and ASCII frame handling |
| pylabrobot/plate_reading/init.py | Exports the new backend and configuration class for public API access |
| pylabrobot/io/usb.py | Adds error handling to decode operations to prevent crashes from malformed USB data |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| if end == -1: | ||
| return False, None | ||
| text = buffer[1:end].decode("ascii", "ignore") | ||
| del buffer[: end + 2] |
There was a problem hiding this comment.
This deletion assumes the buffer has at least end + 2 bytes, but only checks for bytes up to end. If the buffer ends exactly at ETX (index end), accessing end + 1 would be valid, but end + 2 could exceed buffer length. Add a length check before deletion.
There was a problem hiding this comment.
i have pushed a new commit for extra guard for this after checking the hexdumps. additionally, there has already been guard for absorbance and fluorescence which checked that the consumed
i pushed a new commit (ca73508) for extra guard for this after checking the hexdumps. found cases where ETX arrives at end‑of‑packet (though this does not appear to change the measurement data). nonetheless the extra guard avoids consuming partial frames. also note in current code, we already validate measurement frames: ABS checks wavelength (words[1]), FI checks excitation/emission (words[1]/[2]). only LUM has no wavelength field to match.
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
rickwierenga
left a comment
rickwierenga
left a comment
There was a problem hiding this comment.
apologies for the late response, I was working on slas and then the old nimbus PR
I have two small initial questions / comments for things I think we can parameterize easily
| "TIME 0,LAG=0", | ||
| "TIME 0,READDELAY=0", | ||
| "GAIN 0,VALUE=100", | ||
| "POSITION 0,Z=20000", |
There was a problem hiding this comment.
"POSITION 0,Z=20000",
do you know what these units are? would be nice to parameterize them as focal_height
| "POSITION LUM,Z=14620", | ||
| "TIME 0,INTEGRATION=3000000", |
There was a problem hiding this comment.
do you know what the integration time units are?
- USB._read_packet() now accepts optional size to read exact bytes from wire - USB.read() passes remaining byte count to _read_packet() when size specified - PyUSBInfiniteTransport.read() uses new size parameter instead of slicing - USBValidator.read() updated for interface consistency Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove InfiniteTransport protocol and PyUSBInfiniteTransport class - TecanInfinite200ProBackend creates USB instance internally as self.io - Move reset logic into backend's _recover_transport method - Simplifies the transport layer by removing unnecessary abstraction Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
|
could you please explain what are they just more parameters like this? that we can add as commands? why not add them? |
Replace custom _u16be, _u32be, _i32be helper functions with the Reader class from io/binary using little_endian=False for big-endian parsing. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Remove redundant step_loss local variable and parameter since _end_run can access self._active_step_loss_commands directly. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Inline command lists in _configure_absorbance, _configure_fluorescence, and _configure_luminescence with direct _send_ascii calls - Refactor _cleanup_protocol to use a local helper function instead of building a commands list Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Rename: - _send_ascii -> _send_command - _frame_ascii_command -> _frame_command - _drain_ascii -> _drain - _read_ascii_response -> _read_command_response - _ascii_parser -> _parser Also update related docstrings. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2 participants
EDIT: all three modes now reflect what OEM does with 100% accuracy, and luminescence is no longer experimental ([e92c6db])
Tests: make lint, make format-check, make typecheck, make test.