J. Kissel

Even though we confirmed that no DAC's "EPICs-channel-to-physical-card-assignment" changed for the model updates that correspond to the 20-bit DAC upgrade of h1susb123 IO chassis this morning (LHO:84506), the MEDM overview screens showed that channels were askew as we started to turn control of the suspensions back on. 

I realize that this is a result how the RCG automatically generates channel numbers for DACs in the USER MODEL.

Let's use ITMX as the demonstration, because that ended up being the most jumbled model.

Physical Slot      IOP Model Name    IOP Card Num     1st IOP Channel    USER Model Name   USER Card Num    GDSTP MEDM Label   1st USER Channel
2                  DAC_0             0                0_0            
4                  DAC_1             1                1_0                DAC_1             1                D1                 0_0
5                  DAC_2             2                2_0            
6                  DAC_3             3                3_0
7                  DAC_4             4                4_0                DAC_4             4                D4                 1_0
8                  DAC_5             5                5_0                DAC_5             5                D5                 2_0
9                  DAC_6             6                6_0     
10                 DAC_7             7                7_0                DAC_7             7                D7                 3_0
So -- even though the IOP model, USER model, and even the automatically generated screens agree to use the call the name of these DAC cards 1, 4, 5, and 7, the actual underlying channels that display the output of these DACs call them DAC 0, 1, 2, and 3.

In a mixed chassis, this USER model DAC channel number counting starts and continues as a simple number of cards called out in the model, first counting 18s, then 20s, then 16s (I think).
So, since the number of 18s in this formerly mixed chassis dropped to zero, that channel count got re-arranged.

This meant that the display on the SUS_OVERVIEW screens (e.g. this screenshot of the ITMX M0 and L1 stages) would not correctly show a given OSEMs signal smoothly transitioning from the "MASTER OUTPUT" column to "USER MODEL DAC OUTPUT" to "IOP MODEL DAC OUTPUT."

These channel assignments are defined in the MEDM macro text files; for this chassis these are 
    /opt/rtcds/userapps/release/sus/common/medm
        susitmy_overview_macro.txt
        susitmx_overview_macro.txt
        susbs_overview_macro.txt
These are nominally common files between sites. Regrettably, because LLO upgraded their 18s to 20s at a different rate than LHO, they've necessarily diverged, so they're not committed / tracked in svn. So at the start of today it was locally modified (due to its difference with LLO, which uses the committed version), and now I've made the updates and it's locally modified (and they're still different from the committed version), so no good version control diffs.

I've updated the macro and validated with simple offsets that each and every output signal goes through the three different MASTEROUT USERDAC and IOP DAC channels for ITMY, ITMX and BS.

In hopes to have *something* in the future, I attach the updated macro files here.