05-04-2012, 12:06 PM
A Universal Smart Transducer Interface
10.1.1.28.7258.pdf (Size: 49.86 KB / Downloads: 106)
Introduction
The point-to-point connection of process input/output
devices to a control system is expensive, both from the
installation point of view and from the engineering point of
view. One approach to reduce these costs is the
introduction of the emerging smart transducer technology.
A smart transducer is an intelligent subsystem consisting
of a sensing or actuating device (sometimes already an
integrated device on a silicon die), a micro-controller with
the necessary software, and a communication network
interface (CNI) to a field bus.
What is an Interface?
An interface is a common boundary between two
subsystems. An information exchange across an interface
is only possible if the engaged subsystems share a
common background of concepts and a common coding
system. In the context of a distributed control system, the
smallest area of concern is a cluster, consisting of a set of
sensor, actuator, and processing nodes connected by a
broadbus communication medium. The set of all nodes of
the cluster must thus share the same concepts and must
agree on common code spaces.
Observations
In the abstract, the purpose of a smart sensor interface is
the timely exchange of "observations" of real-time entities
between the engaged subsystems across the provided
interfaces. An RT entity is a state variable of interest that
has a name and a value. An observation [3] is thus an
atomic triple
Smart Sensors
The 4-20 mA interface technology was primarily developed
to interface an analog sensing element with a small amount
of local analog processing logic to an (remote) analog
controller. The noise pickup on the analog transmission
line is one of the limiting factors for the accuracy of 4-20
mA signals. A smart sensor is the combination of an
analog or digital sensor or actuator element and a local
microcontroller that contains the interface circuitry, a
processor, memory and a network controller in a single
unit. The smart sensor transforms the raw sensor signal to
a digital representation, checks and calibrates the signal,
and transmits this digital signal via a secure
communication protocol to its users.
The Generic TTP/A Protocol
The generic TTP/A communication protocol is a timetriggered
protocol for the communication among smart
transducer nodes within a cluster. It is controlled by an
active master who establishes the common time base
within the cluster. In case the master fails, a secondary
master can take over control. In TTP/A it is assumed that
every node has a unique personal identification number,
e.g., an eight byte integer that is used to assign a (short)
logical name to the node after power up. The scope of the
logical name is a single cluster.
Communication System
As described in Section 3, the communication of TTP/A is
organized into rounds. Every round starts with the
fireworks frame from the master. The fireworks frame of a
multipartner round consists of two UART bytes, the
fireworks frame of a master-slave round consists of four
UART bytes.
Namespaces
We have tried to map all names into single or multiple byte
objects, such that these objects can be communicated
efficiently by an UART protocol and stored in a single
byte memory location of an 8-bit microcontroller
architecture.
File-operation: The generic TTP/A protocol requires 3 file
operations: read, write, and execute. The operation code
is assigned to the first 2 bits of one byte.
File-name: It has been decided to provide a name space
for filenames of 64 interface files in each node. This
requires a six-bit name field.
Conclusions
The low-level programming of the input/output routines
for the diverse I/O devices is cumbersome and error prone.
The specification of a universal smart transducer interfaces
makes it possible to hide the idiosyncrasies of the various
I/O devices behind such an interface and to provide to the
programmer a unified view of the I/O devices.