Why PLC is preffered over Industrial PC's and controllers for automation?

I call "bullshit" on 5+ years of experience with XP and 2000.

As for the 2 day MTBF comments, those are derived from the manufacturer's published figures. Such as they are. I suspect that as far as most people are concerned, the manufacturer's "5% crash often" figure will weigh more heavily than your anecdotal report.

Jiri
--
Jiri Baum <[email protected]> http://www.csse.monash.edu.au/~jirib
MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools
 
C
Hi Matthew

You should apply at Microsoft, They obviously could use your talents. I'm sure Bill Gates doesn't see that kind of reliability, so you must know something they don't.

Regards

cww
 
C
Hi anonymous,

Now that's silly too. Lots of QNX, VX Works and RTLinux ueers as well as othere RTOS users would dispute that. After they finished arguing what near real time means :^) Even "normal" Linux can meet 10mSec deadlines with kernel premption to a very high degree of confidence. In fact Wind River just linked up with RedHat and is aggressively
persuing the Linux business:

http://www.windriver.com/announces/linux/

This is the very perception I'm talking about. Most would dispute the "not built for reliability part as well" especially when you configure a PC for reliability. Or use an embedded class PC.

Regards

cww
 
Sounds like you encountered the BSOD screensaver in the xscreensaver package... "Systems depicted include Microsoft's Windows 95 and Windows NT, Commodore-Amiga's AmigaDOS 1.3, SPARC Linux, SCO UNIX, the Apple Macintosh (both the MacsBug debugger and the rarer "Sad Mac"), and the Atari ST."

http://gd.tuwien.ac.at/linuxcommand.org/man_pages/bsod1.html
http://www.jwz.org/xscreensaver/

XWindows may have crashed, but the system was probable still Ok. If you couldn't recover control from the console, you could always telnet/ssh in to kill the offending process.
 
B

Blunier, Mark

So you want us to believe that windows 98 will run continuously for 4 years in a row, and the the way to do this is maintain it by keeping the software up to date. Yet when I look at the patches released for windows 98, there are several patches that have been release in the last 4 years, and they required that you reboot your computer to impalement them. Thus you are contradicting yourself.

Mark
 
J

Joe Jansen/TECH/HQ/KEMET/US

Win 2000 and XP for 5+ years? If this is 2004, did I miss some early release of 2000 or XP that would have come out 5 years ago? Or maybe a time warp?
 
D
And I am sure that the average machine operator has
root access to do this... I suspect that the action
taken would be the same as if/when Windows freezes -
reboot and keep on gettin' it.

Davis Gentry
 
C
Hi Jiri

I agree, but I was waiting for the Windows folks to call him on that one. I don't think anyone would believe that since almost everybody uses Windows.

Regards

cww
 
M

Michael Griffin

Actually, I believe it was Windows 95 and 98 which crashed after 49.7 days. The crash was caused when the 32 bit milli-second uptime counter rolled over. The only way to avoid the crash was to re-boot before that happened. It took years for anyone to notice this problem. Not too many people ever had to worry about reaching that limit in practice though.

--

************************
Michael Griffin
London, Ont. Canada
************************
 
Personally, I prefer using a PC (although I have used PLC's for years). With a PC and comm card (such as Profibus master) you can program in a single language (such as VB.Net) and have direct fast access to the world (plant devices on a network, database access, operator interfacing, animation, great debugging environment, a multitude of built-in functions and more).

I might also add that I have heavy duty applications running a standard Dell PC that have been running for several years with a glitch. These applications are running with Profibus master card to I/O blocks, 8 server axis, vision lighting controllers, pnuematic valve blocks and AC variable speed drives. Additionally, the PC has an imaging card for vision with live video, is running animated graphics and reads and writes parameters to a remote database server (MS SQL Server).

If you try to do this with a PLC, comparitively you pay a much higher cost in hardware and software packages. The PC with VB or C offers seamless integration in a single lanquage, where a PLC (which I have used for such applications) means many pieces to integrate. This often means a debugging nightmare when one vendor is pointing the finger at the other because of bug in software which you have no debugging access to determine who is the offender.

Just my take after over 25 years experience from relays to PLCs, to PCs.
 
C

california bob

???? This is the dumbest thing I've heard all week, - it's now going to rattle around in my brain all day.

There are tens of THOUSANDS of PCs doing 'real-time I/O servicing' - I have three on my desk now, two of them 'embedded' PCs, one is even synching several drives at 1ms cycle times and only about 3us jitter.

Get your facts straight before making global generalizations, especially ones that are dead wrong.

Robert Trask, PE
Los Angeles, CA
[email protected]
 
E

Edwin van den Oetelaar

<p>http://www.microsoft.com/ntserver/nts/downloads/management/uptime/default.asp

<p>The place for the uptime command for windows. However even this simple program is broken. It reports my system up for 267 days but I can see in the event-log that I rebooted a few days ago.
<pre>
[snip from screendump]
06/04/2004 11:43:18 AM Boot Prior downtime:0d 0h:1m:37s
06/04/2004 2:35:10 PM Boot

Current System Uptime: 267 day(s), 10 hour(s), 15 minute(s), 9 second(s)

Estimate based on last boot record in the event log.
See UPTIME /help for more detail.

------------------------------------------------------------------------


Since 09/15/2003:

Total Reboots: 119
Mean Time Between Reboots: 2.25 days
Total Bluescreens: 3


Notes:
There is insufficient data in the event log to calculate
system availability. Please see UPTIME /help for more detail.
</pre>
 
C
They've declared their system Open, they've declared their system costs less, they've declared their system secure. What reason would they have to doubt that people will believe the uptime figure?

Regards

cww
 
Top