Dave, Jim, Jeff, Filiburto

Software Watchdogs

I created new h1iopsusex and h1iopseiex models which use the new RCG2.8 Watchdog and DACKILLIOP parts. This permits the timed shutdown of SEI and SUS DACs if the excitation of the suspension(s) persits over a period of time. These parts are available with RCG2.8.3, so initially I built the models against this. For test purposes the countdown timers have been kept short, between 10 and 30 seconds.

I put together a new software watchdog MEDM in userapps cds/h1/medm/H1CDS_SOFTWARE_SUS_TARGETEDWD_EY.adl (targeted watchdog). We are starting with EY since this is the easiest model (one chamber, simple mapping between SUS and SEI, one IPC channel).

Jeff setup SUS ETMY to drive the DACs, I then tested the watchdogs using oscillators I built into the IOP for test purposes. There were no problems found.

I then set the release pointers to build out of RCG branch-2.8 to get the final version of the WD parts (with trip levels in mV) and the reset-cached bug fixed.

Hardware Watchdogs

To support in monitoring-only testing of the hardware watchdog (HWWD) in EY, I modified the h1susetmy model to add the HWWD part, and link it to the Binary input and output channels which are connected to the unit. We were able to read back the status of the front panel LEDs, but were unable to set the WD parameters. I then installed the HWWD part the h1sustmsy as a test of using different Binary chassis, but the problem remains. I backed out the change in h1sustmsy, and made a copy of etmy called h1susetmx_hwwd.mdl. By reverting the models I did not need to do a DAQ restart.

The release pointers were set back to 2.8.3 and the sus models were recompiled and restarted.

Situation at the end of the day

So at the end of this, the IOP models are the new targeted/timed DACKILL version built against Branch2.8, the SUS ETMY,TMSY are back to the original and the SEI ISI/HPI models were not changed today.