Andrés Alejandro Plazas Malagón added a comment - 25/Jan/19 12:55 PM - edited People attending: Merlin Fisher-Levine, Pierre Astier, Pierre Antilogus, Andrei Nomerotski, Andrés Plazas Initial discussion: stability of pedestal e2V: about 25k ADU total, stable Two contributions to pedestal as seen in overscan: Main contribution of about 22k ADU after analog amplification (video chain): very stable Clock injection at CCD level of about 3K: 1% variation (30 ADU out of 25k ADU) ITL: about 10k-30k? total: not that clear Two contributions to pedestal: Main contribution of about 22k ADU after analog amplification (video chain): very stable (same as above for e2v) Clock injection at CCD level: could have large variations! (e.g., 15k!). From Pierre: "This spread is associated with a bad ground connection of the amplifier in the CCD, in particular for the large offset, and it’s associated to noise and CTI issues.  DM is not set up yet to monitor/deal with variability in overscan and 2D bias. Overscan: for ITL, not flat in both (column and row) directions. Currently in DM: fit spline per row. Need a way of checking variation We would benefit from storing overscan values: Good to monitor drift of pedestal Make diagnostic plots Being able to try different overscan corrections inside DM framework would be interesting. 2D bias frames: during bootcamp, biasing a bias frame still left residual structure. Reported to Robert, not sure if something was done about it. To move forward: Create a list of sensor effects with priorities, tasks, and deliverables. Already there (use the page created for DM-16552 ):  https://confluence.lsstcorp.org/display/DM/Sensor+Characterization+and+ISR Use the list to open the discussion to other members of DESC/SAWG and get their help Let’s meet again in two weeks (2/7/18), same time and day of the week (Wed., 9 a.m EST/ 3 p.m Paris) Joint DM-SAWG session at next DESC meeting in Berkeley.