|
Vol
8 Issue 6
Home
Contents
Next
Comment
|
|
|
 |
|
Following
postgraduate study in mechanical engineering design at Cambridge
University, Jon Severn worked as a product and machine design
engineer, before becoming Contributing Editor for European
Design Engineer
|
|
JON SEVERN
Two-thirds of the way
to heaven
"Three steps
to heaven" was, somewhat ironically, a posthumous hit for the rock
singer Eddie Cochran in 1960, and a conversation I had recently
started me thinking along the lines of three steps to an automation
heaven. For a designer of machinery, the ideal would be to have
a single fieldbus network that could be used to convey all information
necessary for control. While conventional fieldbuses are fine for
transmitting data relating to standard machine I/O, there are limitations
- especially when it comes to closed-loop motion control of either
position or velocity.
To operate
a number of synchronised motor drives, as might be required in,
for instance, a high-speed paper mill where a number of sets of
rollers have to be controlled simultaneously, determinism is vital,
whether the motion or position control function is centralised or
distributed. In the example of a paper mill, all the motors need
to accelerate or decelerate in unison, otherwise the paper is likely
to be torn or be spewed out due to the tension straying outside
the acceptable limits.
Most fieldbus
protocols either do not have the determinism or speed to control
high-speed synchronised drives and, although there are several bus
systems that have been specifically developed for operating drives,
these tend not to be capable of handling all of the additional general
machine I/O data. In many cases, proprietary protocols have been
developed for linking drives but, if an OEM chooses to specify one
of these, it severely limits the choice of drive supplier and is
not a favoured option for those who are keen to utilise open control
systems.
Now, ladies
and gentlemen, please welcome on stage, the headline act: Profibus!
Within Profibus is a profile, Profidrive, that enables the closed-loop
control of drives via the same fieldbus that already handles the
machine I/O data. This means that the machine builder has less design
to do, there is only one cable to route and connect, and the system
is likely to be more reliable than a conventional set-up wherein
encoders or other feedback devices are wired back to a drive (less
wire equals fewer cabling errors during installation and fewer wires
to break during normal operation and abuse). With OEMs and end users
both working towards reducing the number of suppliers they use today,
the concept of a single fieldbus for transmitting I/O and drive
data is very attractive.
Although Profidrive
has its limitations (typically a maximum of 14 drives on a network
and cable lengths that need to be kept short so as to maintain a
high data transmission rate, for example), the important point is
that it is based on the Profibus standard. As such, the system is
open and any manufacturer of drives can design and manufacture compatible
products if they desire, which gives fans of open control systems
a far broader choice of products.
So far, so
good: we are being offered an open fieldbus protocol that is capable
of conveying two types of data - which is good progress towards
our automation heaven. But Eddie Cochran told us that there are
three steps to heaven, not two. So what is the third step? In the
June 2002 issue of this magazine I wrote about Profisafe, the profile
within the Profibus standard that allows safety-related control
data to be transmitted on the same fieldbus as the machine I/O.
My question, therefore, is when will Profibus be ready to convey
all three types of data (machine I/O, motion control and safety)?
Regrettably,
the Profibus expert with whom I was recently conversing would not
discuss this topic, but it would be great if such an all-encompassing
fieldbus could be created. Meanwhile, as Meatloaf has been singing
since 1977, "Don't be sad, 'cause two out of three ain't bad."
|