Experimental Physics and
Industrial Control System

Stefan Heim <[email protected]> · Fri, 13 Jul 2007 18:31:20 +0200

Hello all,

I have a device that came with a controller from the vendor. This controller 
offers a TCP/IP server you can connect to and send a certain set of commands 
to operate the device.

Using EPICS, asyn and streamdevice, it has been close to trivial to set up a 
basic softIOC to operate the device (essentially a getPosition/setPosition 
operation).

In the streamdevice protocol file I have (among others)

getX {
        out "Joe Crds ?";
        in "Joe Crds %f %*f %*f %*f %*f %*f %*f ";
}

setX {
        out "Joe AxisAbs 1 %(VAL)f";
        in "Joe started";
}

In the database I have (among others)

record (ai, "MyIOC:X")
{
        field (SCAN, ".1 second")
        field (DTYP, "stream")
        field (INP,  "@device.proto getX MYPORT")
	# EGU, HOPR, LOPR, PREC, warning limits, ... left out for brevity
}

record (ao, "MyIOC:setX")
{
        field (SCAN, "Passive")
        field (DTYP, "stream")
        field (INP,  "@device.proto setX MYPORT")
	# EGU, DRVL, DRVH, PREC, warning limits, ... left out for brevity
}

This works like a charm, I can do "caget MyIOC:X" to get the current position 
and "caput MyIOC:setX <targetpos>" to drive to a target position.

Now I want these records to operate in different, linearly transformed 
coordinates rather than in the device's very own coordinates, let's say for 
example

myCoords = -20.5 * devCoords + 5.3 .

Therefore, I added

        field (ASLO, "-20.5")
        field (AOFF, "5.3")

to the ai and ao records described above; for the ai record, this had the 
desired effect, i.e. the record now delivers the position in transformed 
coordinates.

However, the ao record _ignores_ ASLO and AOFF with a streamdevice device 
support. Is this the intended behaviour? Or a bug in streamdevice's device 
support for the ao record?

How are you supposed to solve the problem of linear coordinate transformations 
without adding a bunch of helper records (calc & calcout, or even motor) to 
the database? 

The LINR="Slope" ESLO/EOFF scaling of the ao record would be sufficient, if 
only RVAL was a double rather than a long; atm if you use it, it 
truncates "devices coordinates" to whole numbers after doing the correct 
transformation.

Any suggestions?

-Stefan

-- 
Stefan Heim, PhD Student X-ray microscopy
BESSY mbH, Albert-Einstein-StraÃe 15, D-12489 Berlin, Germany
voice: +49 30 6392 3177 fax: + 49 30 6392 4757 e-mail: [email protected]
BESSY mbH - Mitglied der Leibniz Gemeinschaft
Vorsitzender des Aufsichtsrates: Prof. Dr. Dr. h.c. mult. Joachim Treusch
GeschÃftsfÃhrer: Prof. Dr. Dr. h.c. Wolfgang Eberhardt, Prof. Dr. Eberhard 
Jaeschke 
Sitz Berlin, AG Charlottenburg, HRB 14635

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System