J. Kissel

Though folks found a temporary solution that worked for the time being while I was off-site last week (see LHO aLOG 8894 and LHO aLOGs 8901), I wanted to try sixteen more ways arranging signals in the h1susim.mdl and/or h1sushtts.mdl front-end simulink models to get what I want out of the calibrated, Cartesian ISI GS-13 signals projected to each optic's EULER basis. Still no dice. Gosh darn it.

I attach to pages of pdf. These are cartoon mimicry of the top level of a given suspension type's front-end Simulink model. I drew the cartoon so you can see a simplified view of the important parts for this discussion, and you can see into the subsystem blocks.

Pg 1 -- a cartoon of what every other suspension type (which only have one optic per model) has. Note, that here, goto/from tags and bus creators/selectors are used going straight from the IPC blocks at the top level, and skipping down two levels (both of which are separate library parts) to a cdsFilt bank. We've been told that skipping levels like this shouldn't work. But it has and does, in this configuration.

Pg 2 -- a cartoon of what I *want* to do with the HSSS, which have many suspensions per model. Here, I send the GS13 signals into a (non-library) subsystem named after the suspension type (HTTS or IM), and calibrate the GS13s once. Yes, I skip a level, but if pg 1 works, so should this. Also, the subsystem in the (HTTS or IM) block has an underscore in the block name -- trying to preserve the channel name format across sus types -- so maybe this is fooling the RCG. From there, the calibrated signal is sent up and out to the top level again, sent to the common HSSS_MASTER block renamed to reflect each of the optics in the model.

As drawn, the model doesn't compile, complaining that just about every bus selectors is not connected.

Where you see roman numerals in double square braces, e.g. [[ii]], I've tried inserting 
- dummy epicsOutputs
- Simulink gains of 1.0
- Test points
- Terminators and Grounds
and can only get it to compile in something equivalent to the temporary configuration where the IPC is fed directly to the HSSS_MASTER.mdl block. 

Very frustrating.