From: Bob Kibrick To: Greg Wirth Subject: an alternate proposal for unswapping the CCD 1A & 1B videos Date: Tue, 19 Aug 2003 11:33:06 -0700 (PDT) Greg and Hector, I have looked at the results of Greg's IDL test script that ran last night. Unfortunately, after 375 exposures, we failed to see even a single case of the intermittent dropout problem from CCD 1A/1B. I'm running additional tests this morning in which I'm alternated between spectral/single:b readouts and direct/dual:a+b readouts, and I have yet to see a single recurrence of the intermittent problem. Either we are just unlucky (e.g., the "watched water never boils" syndrome), or we may have inadvertenty "fixed" the problem by messing with the SMA/SMB connectors that pass the video signals into the front of the video processing board. (These are the most delicate connectors on the portion of the electronics that is external to the dewar and its dewar electronics.) If the latter is the case, then we might potentially re-introduce the problem if we swap these video cables and their delicate SMA/SMB connectors back to their original configuration. Instead, I propose that we swap the CCD 1A and 1B video signal coaxes at their respective BNC connectors on the BACK panel of the DEIMOS science mosaic controller (the back panel faces toward the telescope). The back panel is easier to get to, and the BNC connectors are much easier to remove and re-attach. The wiring would then look like this: Dewar Preamplifier Preamplifier Electronics Output for Output for Box CCD 1A amp CCD 1B amp ^ ^ \BNC BNC/ \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / Swap these two \ BNC/ cable ends today \BNC V V Back Panel CCD 1A amp CCD 1B amp BNC Connectors ^ ^ \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ SMA/ \SMA V V SMA/SMB CCD 1A amp CCD 1B amp Connectors on front of The cables to these two connectors Video Board were the ones swapped on August 5, 2003. For CCD 1 Leave these alone. If we agree to proceed with this plan, then I suggest we print a copy of this e-mail (or make a reduced copy of it) and prominently affix it to the side of the DEIMOS science CCD controller so that techs will understand why the wiring/labelling was perverted in this manner. We should also commit a copy of the text of this e-mail to a documentation file that is checked into the document directory of the DEIMOS CVS repository. Regards, Bob P.S. If we do this, my apologies to Steve Allen, as it will completely mess up his very nice drawings of this cabling.