Experimental Physics and
Industrial Control System

"Mark Rivers" <[email protected]> · Wed, 21 Sep 2011 09:43:10 -0500

I'll see if I can beat Eric Norum in responding to this!  He solved the
same problem and sent me a solution yesterday.

*****************************************************************
Q. My PLC returns floating point values as two 16 bit integers.  How can
I convert these to a floating point value?

A. Here's a CALC record that can do this.   The expression is longer
than 40 characters but that's a problem only if a client tries to read
the expression as an EPICS string.

#
# Convert to IEEE-754 32-bit float.
# No INF/NAN, no denormalized numbers.
#
record(calc,"$(P)vacuum") {
    field(INPA, "$(P)vacuumHiWord")
    field(INPB, "$(P)vacuumLoWord")
    field(INPC, "0x8000")   # Hi word, sign bit
    field(INPD, "0x7F80")   # Hi word, exponent mask
    field(INPE, "0x00FF")   # Hi word, mantissa mask (incl hidden bit)
    field(INPF, "150")      # Exponent offset plus 23-bit mantissa shift
    field(INPG, "0x0080")   # Mantissa hidden bit
    field(INPJ, "65536")    # Hi/Lo mantissa ratio
    field(CALC, "(A&D?(A&C?-1:1):0)*((G|A&E)*J+B)*2^((A&D)/G-F)")
    field(PREC, "5")
}

--
Eric Norum
[email protected]
*****************************************************************

Eric suggested that I put this in the Modbus documentation, which I will
do.

Note that I don't support Float32 values directly in the driver because
they are not part of the Modbus spec.  There is no guarantee about byte
order, floating point format, etc.  Modbus guarantees to do the
byte-order on 16-bit values correctly, but for any larger data type it
does not guarantee whether the most-significant 16-bit word is the first
or last 16-bit register.

Mark

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Olivier Zimmermann
Sent: Wednesday, September 21, 2011 9:38 AM
To: [email protected]
Subject: Modbus data to float32 ?

Hello,

I'm starting with EPICS and trying to have a PV read a SGL/Float32 value
exposed by a modbus device (the value is stored on two 16bits-wide
holding registers).

To achieve this I'm using "Driver Support for Modbus Protocol under
EPICS".
http://cars9.uchicago.edu/software/epics/modbusDoc.html

Modbus data is retrieved as follows (i.e. using modbus function 4 to
retrieve two 16bit words at address 9) :

drvAsynIPPortConfigure("INSTR","192.168.##.##:502",0,1,1)
modbusInterposeConfig("INSTR",0,5000)
drvModbusAsynConfigure("INSTR:ReadFloat32","INSTR",1,4,9,2,0,50,
"InstrInterface")

What I just can't figure out is :

- How do I typecast this 32bit modbus data into a Float32 ?

- What "EPICS device support" should I use (asynFloat64, asynInt32...) ?

- Should my PV be an "ai" and where can I cast the right data type
within ?

Thanks for any help,
oz

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System