I worked on the mitigation of c1mcs time-over issue this afternoon.
The timing for the c1mcs is successfully reduced from >60us to 45us.
The previous models are svned in redoubt as follows:
MCS rev. 6696
RFM rev. 6697
IOO rev. 6698
What I changed was:
- Remove connection from ALS (on c1ioo) to MCS (on c1sus). This should be all done in LSC. (# of RFM IPC in MCS -1)
- MC2 trans QPD filters are moved from IOO to MCS to reduce the RFM channels in MCS.
Previously the signals for the 4 segments are sent. Now the processed siganls (pit/yaw/sum) are sent. (# of RFM IPC in IOO -1, MCS -1)
- WFS MC3 feedback channels are moved from MCS to RFM to distribute the RFM channels (# of RFM IPC in MCS -2, in RFM +2)
model prev. timing[us] current timing[us] diff in time[us] diff in ch#
c1mcs >60 45 -15 -4
c1rfm 47 53 + 6 +2
c1ioo 47 36 -11 -1
Revisions of the new models:
MCS rev. 6702
RFM rev. 6701
IOO rev. 6700 |