Experimental Physics and
Industrial Control System

"Bryce Karnaghan" <[email protected]> · Tue, 18 Sep 2007 10:59:29 +1000

Emmanuel

One simple system I have deployed (not epics related) consisted of the
control system (IOC) using a semiconductor digital output that toggled at a
10 second period square wave(the actual rate depends on the requirements of
the controlled process). 
This 10 second pulse was wired to the input of a simple industrial timer
relay that maintained a closed contact to the plc while the 10 second square
wave was maintained by the control system. 

If it stopped (either high or low for longer that the time out on the relay)
then the contact to the PLC opened and alternate action was taken. Fairly
simple but worked very well.

Regards

Bryce Karnaghan
Control Systems Specialist
Australian Synchrotron
Phone 8540 4135
Mobile 0423844592
E-mail [email protected]

-----Original Message-----
From: [email protected] [mailto:[email protected]]
On Behalf Of Emmanuel Mayssat
Sent: Tuesday, 18 September 2007 09:29
To: [email protected]
Subject: Re: host watchdog and PLC

On Sat, 2007-09-15 at 15:09 -0600, Elder Matias wrote:
> Emmanuel,
> 
> I would agree with John, at the control system design stage you need to
> determine how you are allocating functionality to the sub-systems.  At
> the CLS we configure EPICS so that it can go down and equipment and
> people are still protected.    
> 

Oh, it seems that I misunderstood John's email... My apologies.
I assumed that he was not using a PLC. Yes, we want all the same thing
(no flame war here!) EPICS, computer, IOCs can go down, then the PLC
takes over...

What I am interested in is the detection of the computer failure, not
its resolution. The remote reboot facility which John describes is a
nice feature which I may decide to implement in the future.

To summarize, given my environment, to detect a computer failure, I may
just need a cheap PLC like the Koyo ones and connect it to the other one
I have in place (Siemens PLC not network enabled).

This may not be my final design, but already I would like to thank you
all for you useful suggestions.

Regards,
--
E

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System