Skip to content

Enhanced cluster masking

At the moment, the clusters are masked L1A_MASK_DELAY BX after each L1A during L1A_MASK_WIDTH BX. This is a fully functional solution that would be optimal if the delayed VFAT pulses would be continuous. In practice, this is not the case as shown by the following scan (credits to @cgalloni):

fed1467-slot1-oh3-width1

In that specific case, the clusters should be masked 4-7 BX and 10-11 BX after an L1A. Masking the BX 8-9 would result non-required deadtime (except if the analog front-end of the VFAT is anyway statured).

This issue proposes to optimize the cluster masking by masking a bitmask instead of a window. Upon L1A, the bitmask would be "OR-loaded" into a bitshift register. The output of the bitshift register would decide whether the current clusters should be masked or not. Based on the first measurements, a bitmask of 32-bits should be sufficient.

To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information