Re: Rate of change calculation

[Donny Domagoj Cosic via Tech-talk <tech-talk at aps.anl.gov>]

Thank you to everyone for your help and the detailed explanation of the 
benefits and merits of each approach. I really appreciate it and it gave 
me a deeper understanding on how records are processed and what to keep 
in mind when writing new databases.
Regards,
Donny

---
Donny Domagoj Cosic
 Institut Ruđer Bošković, Bijenička cesta 54, 10000 Zagreb, Croatia

Dana 29.03.2021 18:42, Johnson, Andrew N. je napisao(la):
> On Mar 29, 2021, at 7:41 AM, Pietryla, Tony wrote:
> > You could change your calc record to this:
> >
> > record(calc, "$(KEIT):CalcCurrent")
> >  {
> >     field(SCAN, "1 second")
> >     field(INPA, "$(KEIT):Charge PP")
> >     field(CALC, "LA-A")
> > }
> Unfortunately the CALC expression doesn’t have access to the fields LA
> through LL, although they can be read by other records through DB
> links or via Channel Access.
>
>
> On 3/29/21, 7:33 AM, Donny Domagoj Cosic wrote:
>
> >    However, when I corrected this error as you suggested, the result 
> > was
> >    still always 0. Changing the order of NPP before PP should have 
> > solved
> >    this problem but it didn't.
> Marty’s suggestion has been used as the standard “rate of change”
> database in EPICS training courses for well over 20 years, but it
> doesn’t work if the record which is the source of the data (your
> "$(KEIT):Charge” record) is asynchronous, which yours is since it uses
> Streams. When the PP link processes that record Streams sends the
> request for data to your Keithley 6514 device, but the device support
> is not allowed to wait for a response because that would stop the IOC
> until the reply comes back. What happens instead is that after the I/O
> operation has been started the PP link just returns whatever value was
> in the ai.VAL field at the time and the calculation uses that.
>
> >    Using Marks suggestion of setting B to A after the calculation
> >    ("B-A;B:=A") works as expected and yields the correct result. This
> >    solves my problem, but I still dont understand why Marty's 
> > suggestion
> >    didn't work.
> Mark’s approach still doesn’t solve the problem that the PP link to
> your ai record is always returning the previous reading. That might
> not matter in your case, but if it could this database solves that
> problem:
>
> record(ai, "$(KEIT):Charge") {
>   field(DESC, "Measure Charge")
>   field(DTYP, “stream")          # Typo fixed, this was "steam"
>   field("@devKeithley6514.proto measureCharge() $(PORT)")
>   field(SCAN, "1 second")
>   field(FLNK, "$(KEIT):CalcCurrent")
> }
> record(calc, "$(KEIT):CalcCurrent") {
>   field(SCAN, "Passive")
>   field(INPA, "$(KEIT):Charge NPP")
>   field(CALC, "B-A;B:=A")
> }
>
> - Andrew
>
>
>
> >    Dana 29.03.2021 13:48, Smith, Martin via Tech-talk je napisao(la):
> > > Hi Donny,
> > >
> > > Because the Charge record is Passive scan and is being processed
> > > (presumably) only by
> > > the calc record what you are saying makes sense. the calc record
> > > processes the charge
> > > record first for link A using a PP link causing it to get a new
> > > value, then it gets that new
> > >
> > > value again using link B without processing (NPP) the charge record.
> > >
> > > I think what you really want is this:
> > >
> > > record(calc, "$(KEIT):CalcCurrent")
> > > {
> > > field(SCAN, "1 second")
> > > field(INPA, "$(KEIT):Charge NPP") # Get the current value of Charge
> > >
> > > field(INPB, "$(KEIT):Charge PP") # Process charge record then get the
> > > new value
> > > field(CALC, "B-A") # You may need to change this to A-B if you want a
> > > positive value assuming the charge is always increasing
> > > }
> > >
> > > You need to get the current value first then process the record to
> > > get the new value.
> > >
> > > Marty
> > >
> > > -------------------------
> > >
> > > FROM: Tech-talk <tech-talk-bounces at aps.anl.gov> on behalf of Donny
> > > Domagoj Cosic via Tech-talk <tech-talk at aps.anl.gov>
> > > SENT: Monday, March 29, 2021 6:35 AM
> > > TO: Tech-talk <tech-talk at aps.anl.gov>
> > > SUBJECT: Rate of change calculation
> > >
> > > Hello,
> > >
> > > I am trying to calculate the rate of change of a record using a calc
> > > record as follows:
> > >
> > > record(ai, "$(KEIT):Charge")
> > > {
> > > field(DESC, "Measure Charge")
> > > field(DTYP, "steam")
> > > field("@devKeithley6514.proto measureCharge() $(PORT)")
> > > field(SCAN, "Passive")
> > > }
> > >
> > > record(calc, "$(KEIT):CalcCurrent")
> > > {
> > > field(SCAN, "1 second")
> > > field(INPA, "$(KEIT):Charge PP")
> > > field(INPB, "$(KEIT):Charge NPP")
> > > field(CALC, "B-A")
> > > }
> > >
> > > The calc record always returns 0. If I change the calc to B+A, to
> > > check
> > > what is going on, I always get double the charge. It seems like the
> > > PP
> > > and NPP are both returning the same value even though the actual
> > > charge
> > > is changing. Any ideas on why this would be happening or suggestions
> > > would be greatly appreciated.
> > >
> > > Thank you in advance,
> > > Donny
> > >
> > > --
> > > Donny Domagoj Cosic
> > > Institut Ruđer Bošković, Bijenička cesta 54, 10000 Zagreb,
> > > Croatia