MoReWeb changelog

v0.7.2 - DB reprocessing

• fix: defective pixel counts in performance parameters dictionary were -1
• fix: remove dependence on numpy module
• navigation on chip result pages between different chips with left/right arrow keys
• BB2 test for baremodules
• additional experimental cut for Fulltest BumpBonding test, not used for grading! and only displayed on the chip pages as dash-dotted green line.
• defects.txt can be placed in test directories to manually grade a module C and additionally specify a machine readable defect type. Modules are then listed together with modules, for which the defect has been found in automatic test analysis. This can be used to replace grade.txt. Possible values are:
  • LeakageCurrentPON
  • LeakageCurrentFQ
  • NotProgrammable
  • DeadROC
  • FulltestPixelDefects
  • FulltestPerformance
  • XrayPixelDefects
  • XrayDoubleColumnDefects
  • XrayEfficiency
  • XrayPerformance
  • ... or any other defect which will then create a new category
• add logfile view to reception test
• speed-up linear PH fitting and Scurve fitting

v0.7.0

Grading

FullQualification

• FullQualification Noise grading: 500/1000 to 200/300
• Exclude threshold defects from Gain and p1 defects

X-ray

• Grade 'C' for bad double coulmns (1 percent or more inefficient pixels)
• Exclude bad double columns in efficiency fit
• Noise: criteria for mean noise 300/400

Features

• Improved PresentationMaker and Production Overview generation
• Readback parameters plot shown on Fulltest/ROC pages
• Reception tests can be analyzed
• Speed-up of analysis of FQ (without fit) takes 45 % less time now
• Analysis of single tests:
• PixelAlive: PixelAlive map per module and per chip
• ReadbackCal: ReadbackParameters table and plots per chip

v0.6.8 - DB reprocessing

• Incomplete/broken/fake test data for not testable modules can now be uploaded to the Pisa DB. HTML pages show more explicitly which tests were missing in the full Qualification
• LeakageCurrentPON tests can now be uploaded.

v0.6.7

• IV grading: also grade on measured value of leakage current at -20 degrees with the same criteria than at +17. This will affect grading of modules that have both a higher leakage current at -20 than at +17 and a leakage current at -20 larger than 10 uA.
  
• show fulltest logfile at HTML on results page
• in the module qualifications list, the test center can be shown with parameter --show-test-center

v0.6.6

• add possibility to place a file containing the grade into each fulltest/x-ray hr subfolder to manually change grade. (comments can be placed in the main folder in a file called comments.txt)

v0.6.4

• modified IV grading which takes I(+17)/I(-20) at 150V into account. If IV at -20 is not available, ratio grading is skipped.
• adds Iana loss test to Fulltest analysis
• adds readout error statistics to x-ray HR tests

v0.6.3

New features

• adds upload of HR test results to global DB

Bug fixes

Fitting

• fixed detection of saturated PH for problematic ROCs which saturate at 254

Fulltest analysis

• fix problems of wrong DAC parameters displayed and trim vcal used for defects grading if dacParameter files or file date were modified after test

v0.6.2

General changes

• use png for chip maps to reduce file sizes

v0.6.1

New Features

Fulltest analysis

• add BB2 test

v0.6.0

New Features

Fulltest analysis

• new grading implemented
• fit Vcal threshold with Gaussian instead of taking RMS
• show separate IV and electrical grades
• readback calibration results added
• improve ROC summary table, include grades, performance parameters and reason of degrading
• add new boxes for TBM parameters
• r/o errors and criticals counting added
• display analysis date, MoReWeb version and branch
• display changes of GradingParameters.cfg wrt GradingParameters.cfg.default in FT summary
• show a more verbose summary of ROC pixel defects in the logfile and also a fulltest summary in the end including grades
• include BB4 test, if BB4 and BB test are both present, BB4 has priority
• include 'tornado' plot

X-ray HR tests

• analysis of HR tests has been implemented, including
  • efficiency maps and interpolation
  • hitmaps to check for missing bump bonds
  • readout uniformity
  • double column uniformity
  • SCurves under X-rays

Vcal calibration

• improve fit of spectra with a large backgroun with SpectrumFitImproveInitialGuess and SpectrumFitAdditionalConstraints options
• improve fit of PH vs Vcal line with LinearFitAdditionalConstraints option
• display module hitmap and hits per pixel distribution on vcal calibration page

Leakage current at startup test

• added LeakageCurrentPON test analysis

Production overview page

• an overview page of tested modules can be produced with ./Controller.py -p, including:
  • total and weekly grading summary
  • cumulative plot of tested modules
  • distribution of ROC performance parameters like RelativeGainWidth, Noise, PedestalSpread etc.
  • distribution of pixel parameters like Gain, Noise, Threshold etc.
  • distribution of DAC parameters and sensor properties
  • module failure overview which shows reason for degrading

General features

• several improvements for analysis of large amounts of data
• improved exception handling, errors in one qualification folder should not stop analysis of other folders but just skip this folder
• show progress of analysis
• added "-r" parameter to force re-fitting of SCurve and PH data
• parallelization of fitting: in the SystemConfiguration ParallelProcessing = 1 can be set to enable parallel fitting, LimitProcesses can be used to limit number of threads
• if fulltests fail, an empty row with grade C is written to database instead of writing nothing
• removed not used or constant column from qualification overview page
• denote which fit function is used if fit curve is displayed in histogram
• improved pXar/pyxar software detection
• add analysis date to overview page
• option to sort the qualification list

Bug fixes

Grading

• fixed bug through which many pixel defects were not leading to a degrading above the intended value
• fixed double counting of pixel defects (maily due to dead pixels)
• threshold defects are now properly counted

Analysis

• fixed white pixels that appeared in vcal threshold map
• fixed bug which caused linear PH fit to fail sometimes leading to Gain defects
• fixed bug where hitmaps for all targets were the same in vcal calibration
• fixed 'ignore least significant trimbit' option in TrimBits test
• fixed bugs where mean values were nan because they included nan's

General

• fixed crashes due to exceeding max number of open file handles if analyzing many folders at once
• exception handling of wrong directory structure, wrong folder names, missing .ini files ...
• exception handling of bad modules, especially dead ROCs or double columns
• fixed serialization of json files
• fixed bug which on qualifications overview page did not display the max number of pixel defects from the fulltests
• fixed bug with not correctly displayed times (month instead of minutes)