I repeated the study I did on the L1 ADC hold times ( aLOG entry 16455 ) using date from H1. It is running the tagged RCG 2.9 and has higher thresholds (MAX_ADC_WAIT_CARD_0 of 20, MAX_ADC_WAIT_CARD_S of 3).
Here, I record values from the IOP model proc file (i.e. cat /proc/h1ioplsc0/status on h1lsc0)
FE | Max Hold | ADC#0 Max | Max Delta |
---|---|---|---|
h1lsc0 | 22 | 21 | 1 |
h1susb123 | 23 | 21 | 2 |
h1sush34 | 25 | 21 | 2 |
h1sush56 | 24 | 21 | 2 |
Again, in most cases ADC#0 Max + Max Delta = Max Hold. It does appear that better limits of 23 and 5 for MAX_ADC_WAIT_CARD_0, MAX_ADC_WAIT_CARD_S (from L1 data) would work here. The current CDS bugzilla issue is Bug 791