RS-485 Comm Cable Testing

D

Thread Starter

DCM

We have an existing 19awg 2 pair cable that is about 900 meters long that runs between 2 pump houses. 1 pair is currently used but a relay based control system. Wanting to provide digital communications between the two locations. I realize that 485 needs a ground for reference, which I think I can get from one of the conductors in the other pair. My question is this: is there an easy way to determine if this cable will support rs-485, or am I best to go through the process of actually setting up some equipment and testing it? It does operate to the specs of the local telco, but does not interface with their network at all.

Cheers
 
L

Lynn August Linse

Based on past experience with RS-485 over the "wrong kind of cable", your cable won't pass any tests of suitability ... but that doesn't mean it won't work if you drop the baud rate low enough.

So I'd just give it a try, starting at 1200 baud.

I'd also be very concerned about surge/lightning problems with 2 buildings associated with pipes/water/pumps 900 meter apart. Ground potential shifts will be common and very serious. You should optically isolate at least one end, if not both. Plus the pump wiring could be marginal, which means when a pump starts it's local 'N' ground shifts by many volts.
 
Thankfully, we don't get much lightning in the area but it is not unheard of (no pun intended). I was thinking about some surge suppression, but had not gone too far down that road. The other pais has been in service for 24 years now and has not had any "surge" events that I am aware of, that said the controls it is connected to are all relay based and may be more tolerant of lightning surges.

I am worried about ground potential to a degree, in that if the 0 VDC rail on the other circuit is not grounded, then I may need to use a local ground for reference and that may prevent us from using a RS485 solution due to ground potential. Trying to get out there today for some investigation. Failing that our only other solution will be to either send a 24v signal over the pair or use a Ethernet Extenter such as the one from B and B that will run on 1 pair.

Thanks for your input/suggestions.
Cheers
Dave
 
S
> I am worried about ground potential to a degree, in that if the 0 VDC rail on
> the other circuit is not grounded, then I may need to use a local ground for
> reference and that may prevent us from using a RS485 solution due to ground
> potential.

I kinda liked the optical isolator idea to let you transmit the signal even with a difference in ground level (and pick up some surge protection at the same time).
 
Looking for any recommendations for surge/lightning protection. Wondering what is most popular? I was looking at the stuff from Eaton as it is within budget, or do most people like to use MOVs?

Cheers
 
L

Lynn August Linse

MOVs aren't really 'lightning' protection - more heavy transient surge protection.

If your concern is lightning hitting near one end of the RS-485 line and a 20kv to 100kv "common mode surge" surfing down your cable, then your MOV is popcorn. Some form of discharge tube or spark gap is required.

The MOV will save your equipment if you get a 1kv to 5kv differential mode "spike" injected into the line due to a lightning discharge or other event.

Also, make sure the resistance via the surge-device ground is radically better (lower resistance/impedance) than is the RS-485 signal-ground path through your device.

If the surge-ground has about 1 ohm of resistance, and your RS-485 signal-ground also has about 1 ohm, then don't expect the surge to bother using your surge-device :) That is the beauty of galvanic (or optical) isolation - it makes the RS-485 path look like infinite resistance (a dead-end), ergo the surge-ground is always the preferred path for a high potential.
 
Top