leonarski_fandClaude Opus 5
c619d25c96
CI / lint (push) Skipped
CI / test (3.12) (push) Skipped
CI / test (3.13) (push) Skipped
CI / test-with-beamline-plugins (pxi_bec) (push) Skipped
CI / test-with-beamline-plugins (pxii_bec) (push) Skipped
CI / test-with-beamline-plugins (pxiii_bec) (push) Skipped
Take the DAQ inputs at face value and solve the collection per beamline instead of applying stacked fudge factors. - Resolution is used as reported: the DAQ estimate is no longer biased, so the -0.2 A shift and the 1/(1/d + 0.1) reciprocal-space push are gone. d_tar now drives both the dose budget and the detector distance. - The crystal is assumed to fill the beam, so the sqrt(xtal_y * xtal_z) dilution of the dose rate is dropped. Under that assumption the beam dose rate is the diffraction-weighted rate, which is the damage-relevant one; averaging over unilluminated volume was under-reporting it. The redundant "Calculated Dose (xtal size)" row goes with it -- it was also being fed a rate under an MGy label. - Flux comes from DAQStatus at 100% transmission, replacing the hardcoded 2.5e11 that disagreed with the flux the panel displayed. - Beamline split: X10SA pins the detector at 100 Hz and solves for transmission; X06DA holds transmission at 100% and solves for exposure time, falling back to transmission at the exposure floor. When the budget cannot be spent at full beam we cap at 100% and flag the under-dose rather than stretching the frame. - Respect bl.dtz_min / bl.dtz_max (replacing the hardcoded 108 mm) and the per-beamline exposure floor, with 2 ms stated explicitly for X06DA. - Move the cryo/room-temperature switch from 250 K to the protein-solvent glass transition at 200 K. Warkentin & Thorne fit two Arrhenius regimes either side of ~200 K; applying the cryo budget in the 200-250 K band overdosed there by close to an order of magnitude. - Snap detector distance to 10 mm, transmission to two significant digits and exposure to 1 ms. Every snap rounds in the direction that under-spends the dose budget; dtz floors because calc_dtz_mm returns the largest distance that still puts d_tar on the detector edge. Dose constants are now named and carry their citations (Holton 2009 for kdose, Howells for the 10 MGy/A half-dose, Owen for the room-temperature limit) instead of appearing as bare literals. Fixed while in here: - "Total angle" was wired to the resolution field's newValue, so editing it never recalculated. - self.transmission was only assigned inside update_calculated_labels, which early-returns without a DAQ status, so "Run rotation" could raise AttributeError before the first status arrived. - "Total measurement time" reported the beam-on budget, not wall-clock; those differ whenever transmission is below 100%. - run_measurement re-read the widgets instead of sending the last calculated solution, so dtz, exposure and transmission could be mixed across edits. - Dropped a dead wavelength_angstrom None branch: it is a computed property on DiffractionGeometry and cannot be None. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018JVrLMLGNUXhfJjWxUcTBj
AareDAQ
AareDAQ is the Data Acquisition system for MX beamlines at PSI.