I have modified the c1lsc model so that I have access to the AS110 channels in the triggering and power normalization matrices.
I also put in a few blocks so that we can have triggering on the violin notches that we moved to the LSC model a week or so ago.
Here is my svn comment, so I don't have to retype things:
2 changes: AS110 channels added, and Violin filter triggers added.
AS110: We recently installed a new demod board and PD for an AS110 signal. Since we will not be using AS11 in the forseeable future, the AS110 demod board outputs use the former AS11 channels. I have left the AS11 channels in the model so that we can easily add them back if we want to, but they are grounded rather than connected to the ADC. I've added digital demodulation for AS110, power normalization, and then added the I&Q signals to both the trigger matrix and the main power normalization matrix. NOTE that these slide the matrix columns around. Since the AS110 is also using the former AS11 whitening channels, swapped those on the BIO block also.
Violins: Recently, Rana and I moved the SUS LSC violin filters from the individual suspension models over to the LSC model. Giving every optic every optics' violin notch helps eliminate bad cross-coupling between servo loops. Here, I have enabled triggering for these notches, so that the violin filters can come on after a cavity is locked. Since the filter banks SHOULD BE THE SAME for all LSC_SUS banks, the "mask" is common to all optics.
I also edited several medm screens, to show the new changes: the lsc overview screen has a button to the violin notch triggering screen, in addition to being able to get to the new screen from the regular triggering matrix screen. I made the trigger and normalization matrix screens bigger, since there are now 2 new columns.
I added AS110 to both the LSCoffsets script, and Masayuki's new, better, LSCoffsets2.py.
I added new lines to the .req files for the ifo configure burt restores for the new matrix columns, and the violin triggering.
I restored, checked out, and saved the Xarm, Yarm, MICH, PRM_sb, and DRM configurations.
I tried locking the DRMI, but haven't really been successful. I'm not 100% sure how to do the phasing for AS110, so that could be a problem. For POP, I can watch POPDC to see if something is a carrier or a sideband flash, but I don't have something quite as convenient at the AS port. I have set the AS110 phase to 60 degrees for now, since during free swinging DRMI flashes, it looks like most of the buildup is in the I phase with 60 degrees. Even with the same configurations as a week or so ago, I'm not getting much more than ~1 second locks.
I also tried locking the SRMI, but am not getting anything at all. I think I need to go back to simulation-land to figure out what good signals might be.
Other thoughts:
Stefan modified the LSC filter module triggering blocks, so now we have a new epics variable, "_INVERT", which sends the trigger through a NOT or not. I think that we want to keep this variable set to 0 to be the same as things were, but I do need to expose this new variable on the screens.
The trigger and normalization matrices pictured on the LSC overview screen need to be expanded by the 2 new columns. The actual matrix screens are good, but I forgot to fix up the little Kissel buttons.
When I have a free swinging SRMI, MICH and SRCL should have the same sign for the gain, if I'm using AS55 I&Q for locking.
LLO is using REFL 9I for SRCL, and ASDC for MICH for the SRMI, but I don't have any REFL beam with a misaligned PRM, so I don't think I can copy what Den and Lisa did on Monday night.
I have figured out / rediscovered why the "sqrt" buttons on the power normalization screen aren't restored when you restart the LSC model - They are controlled by momentary epics records, which go to embedded c-code to do some toggling. I don't know yet of a good way to save the configuration of these guys for burt restore-type restoration. This will be a problem for anything that is using these toggle c-codes. |