Experimental Physics and
Industrial Control System

Michael Davidsaver via Tech-talk <tech-talk at aps.anl.gov> · Mon, 26 Jun 2023 14:02:12 -0700

On 6/26/23 12:34, Leblanc, Gregory via Tech-talk wrote:

Hi folks,

I'm still working on getting these Heinzinger PCU 50-300 supplies doing everything I want using EPICS.  Today I thought I'd try reading the status word from the supply, so I added an mbbiDirect record.  The custom sections of my .db file now read:

record(ai, "$(P)$(R)GetCurrent") {
     field(DESC, "Read output current")
     field(DTYP, "stream")
     field(INP, "@devPCU50_300.proto getCurrent $(PORT) $(A)")
     # field(SCAN, "1 second")
     field(SCAN, "Passive")
     # field(PINI, "YES")
     field(EGU, "Amps")
     field(ESLO, "0.001")
     field(LINR, "SLOPE")
}

record(ao, "$(P)$(R)SetCurrent") {
     field(DESC, "Write output current")
     field(DTYP, "stream")
     # field(SCAN, "1 second")
     field(SCAN, "Passive")
     field(OUT, "@devPCU50_300.proto setCurrent(%d) $(PORT) $(A)")
     field(PINI, "YES")
     field(VAL, "0")
     # field(FLNK, "$(P)$(R)GetCurrent")
     field(EGU, "Amps")
     field(ESLO, "0.001")
     field(LINR, "SLOPE")
}

record(mbbiDirect, "$(P)$(R)StatusRegister") {
     field(DESC, "Status Register")
     field(NOBT, "24")
     field(DTYP, "stream")
     field(INP,  "@devPCU50_300.proto getStatus $(PORT) $(A)")
     field(PINI, "YES")
     field(SCAN, "Passive")
}

But I'm seeing some unexpected behavior, somehow related to my other records.  Above, the PINI field for the GetCurrent record is commented out.  When I start the IOC in this state, the StatusRegister record is properly populated.  If I set PINI to YES and start the IOC, the StatusRegister record is not properly populated.  Running the command 'dbtr PCU50_300testStatusRegister' on the "broken" startup will update it so that it has the correct status.  I've watched the console output when launching st.cmd, and browsed the logfiles that I generated, but I can't spot what I'm doing wrong.  I've attached a logfile where I have the problem, and another where I don't.  My .db, .proto, and st.cmd are at: https://github.com/leblancOUAL/PCU_50-300/

I feel like the root cause is that somehow the replies to the different serial commands are getting mixed up, but I can't see how that's possible.  It is more likely that I've done something silly along the way.  Any hints for troubleshooting are greatly appreciated!

I suspect what you are seeing results from the asyn port not being
connected at the time of the IOC's initial scan.

Firstly, the name "StatusRegister" suggests to me that this value
should be read more than once on IOC startup.

imo. for robust operation some care should be taken anytime when
IOC and device don't share a power switch.

I try to avoid PINI with such devices, and instead use SCAN="10 second"
as a starting point for values which change slowly or only on device
power cycle.

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System