Wireless Connectivity

P

Thread Starter

Pete Robinson

I am in the process of dropping in a highly automated line using 20+ nodes on Ethernet. I am using 5/05's and was pondering over slapping a
wireless LAN card into my notebook and using that to get into the 5/05's. Has anybody done this and has expierience of it? Guess I have to
try it anyway but was looking for some do's and don'ts. It may have it's problems but thought it may give me something to play with :) We
all like that!
Anybody been there, done that and got the T-Shirt?
TIA
 
J

Jake Brodsky

There aren't many who have "been there, done that, got the T-shirt" --AND FOR GOOD REASON!

"Wireless" is a term that marketeers have revived to use instead of "radio" so that users get a feeling of "it's just like being wired, but
with out the wire." Get over it. You are using a radio. Radio signals go everywhere. You don't put anything on a radio that you wouldn't
put on a TV station to say to the world.

Yes, I know, there are silly countries such as the USA who legislate ridiculous things trying to put clothes on the naked emperor. THERE IS
NO SUCH THING AS PRIVACY ON THE AIR, and there never was, nor is it likely there ever will be.

If you have something critical to say on the air, use encryption. 802.11b has provisions for minimalist encryption, but that's been cracked. To make matters worse, most users don't even enable those features. Although the likelihood of interception or hacking is low, it's not zero. If you had an industrial accident and the PLCs were involved, most competent lawyers would have a field day with your "wireless" programming port scheme.

So, yes, there is good reason why you don't see that much "wireless" (ugh) stuff in use on a plant. I wouldn't go there if I were you...
 
M
802.11b is probably your best bet but there are problems. First, it is not a secure link so it is prone to bandwidth thieves. They might not be able to hack into your PLCs but they certainly will be able to use the link. I routinely visit a company who has an 802.11b LAN in their offices. I use it for e-mail and browsign with no problem and no local account. The second point is that of drop outs. Don't rely on it for critical access. You will have holes in coverage and drop off occassionally.

Other than that, it is wonderful.

Mitch
 
This has been done many times and is relatively easy. You can use either 802.11b (or A) or even a product that offers proprietary wireless protocols over TCP/IP. I have several customers that use both technologies connecting to PLC-5s and SLC-5/04s on DH+ via a ControlLogix Gateway w/DH+ and ENet Modules.

 
H

Hans van Wijk

Oke, we have been there and, done that .... and got it working!!
First you must must realize that you are working with very low transmission power, 0,1 to 1 mW so the range in which you are able to receive anything isn't that big!!!
Then ofcourse you could upgrade the encryption to work more safely.

BUT..... much more important is that you have a lot more disturbence on your factoryfloor then you have in a neat and clean office. So getting a good and stable connection is much more of a problem.
Luckely if you use 802.11 you work with packets of information which are checked upon receival and incase of a failure are send again.
We have build a 802.11 connection between a control system and two rail-guide transfer cars and we accept the fact that the connection is off for a maximum of 10 seconds.
This system is up and running (24hrs/day) for two years now without any failures.

So please stay to the facts instead of talking about hacking and more of that nonsense which are all theoretical but practical very diffult.

So in an industrial environment you can get it working.
Regards

 
I tried some radio transmiting for my project even it's not done on 5/05. I used spread spectrum 900Hz transmiting with the equipment (several overhead crane.) Basically, it works fine. But, from time to time, you may lost signal
for unknown reason (maybe message conflict, signal interrupted, or others.) If your project were really physically impossible transmitted by hard-wire, then you could try it wireless, otherwise, don't waste your time to troubleshoot on losing communication. It could drive you crazy in no time.
 
Hans van Wijk:
> This system is up and running (24hrs/day) for two years now without any
> failures.

> So please stay to the facts instead of talking about hacking and more of
> that nonsense which are all theoretical but practical very difficult.

The trouble with hacking is that it's very difficult to evaluate the risks accurately.

If you can get the connection working, then so can an attacker just outside the wall, or on a nearby hill with a directional antenna (illegal, but easy to get). Why would anyone want to? No idea. Make your own threat model.


Of course, to some extent the same applies to wired connections, at least for listening - the network cable will work as a very long antenna - but that's less useful (though it might provide passwords for dial-in)...

Jiri
--
Jiri Baum <[email protected]> http://www.csse.monash.edu.au/~jirib
MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools
 
Haven't been there, done that, but seriously considering it for debug only. Mainly due to the fact you could get your notebook right where you need to be to watch something with your own eyes while you were watching logic. Doubt if I would leave the wireless connection running other than during the debug. Also worried about noise interference, especially from VFDs. Any comments on that? Rick Lamb
 
Top