Surge protector and grounding`

[guilio2010]

I'm not familiar with instrument grounding and I am experiencing some issues with our RS485 surge protector and trying to get ideas on what could be causing it. When it storms, the 485 channel on our device is burned up. I want to say that the ground resistance on our SP is high enough where the surge just goes through the 485 channel rather than through the ground wire. I'm looking at the resistance of that leg that is connected to the SP that goes back to our ground bus. The total system resistance is low, but is it possible that the leg of that ground wire for the SP is too high? I'm trying to look at this as a equivalent circuit and that leg is parrallel with other devices ground wire and that is tied all back together.

Also, one soluation is to use an optical isolator. Any thoughts on those?

Thanks,

 

[IRstuff]

It's possible that the two sides of the RS485 interface wind up at different voltages. The typical RS485 input can only tolerate about 7V of common mode voltage. A large common mode can cause the inputs to get fried, especially if it's a bipolar RS485 receiver

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[racookpe1978]

We had a recent situation where the ground wires within each instrument sensor and at the terminal control cable at the computetr console were grounded - but the different ground wires inside the terminal box between those two panels were NOT connected to each succeeding wire -> also, several ground wires inside the junction box were wired to "other" instrument wire coming in: D/P Instrument "A" was grounded to D/P Instrument "B" control wires, D/P Instrument B was grounded to D/P Instrument C control wires, etc.

Do a wire-by-wire check of EVERY connection.

 

[djs]

Surge suppressors do go bad. Generally they should be replaced after each "hit". That being said there are several things to check:
1. Ground for the surge suppressor and the device being protected must be the same. Many RS-485 devices provide a RS-485 port that is isolated from the rest of the device. So device ground is not necessarily the same as the RS-485 ground. And yes RS-485 does have a ground with must be interconnected between the various communications ports.
2. Surge suppressors have a protected side and an un-protected side. Make sure that yours have the protected side toward the local communications port.
3. Make sure that you surge suppressors are three stage deices and have both line to line and line to ground protection paths.
4. Use LARGE grounding wires (the bigger the better) and make sure that it is solidly grounded to the electrical ground, not the local lighting ground.

 

[guilio2010]

I checked all the basics. The SP is connected correctly with equipment side to the RS485 and the Line size going back to our controller. The equipment side is connected at a short distance. The wiring is correct. The box is non metallic. The only thing I keep going back to is the ground wire. The manurfacture states 3' 10AWG max. When I went out to look, we have a 2' 12AWG wire connected to a ground bus bar that has another 50' 12 AWG wire from the busbar back to our controller cabinet. Again, I am not good at the grounding, but when I look at the equivilance of the circuit, that particular ground leg from the SP going back to our cabinet where all our other grounds are connect to has a resistance of ~.08 ohms (value based on wire resistance calculator) where the manufacture resistance on their recommended 3' 10AWG is .003 ohms. Yes they are small resistance, but our is +2000% more than what the manufacture is recommending. Is this right to look at it that way? I feel like that isn't enough and if there was a surge, that is low enough for the SP to cut the voltage and send the rest to earth ground.

 

[djs]

If the problem persists, then a viable option is to use fiber between the various units.

 

[guilio2010]

thats not a bad idea. Just have a converter and run fiber. I'll look into that.

What abouve comments on the thought above concerning the length run and resistance? Is this plausable or in the right line of thinking?

Thanks,

 

[djs]

The idea is to tie all of the grounds at one location together with as low as a resistance as possible. Also you might place an isolated grounding strip (bar) and connect all of the ground wiring associated with the communications to that bar and then run a large cable from the grounding strip to the system ground. The idea is that during a surge event, all of the local equipment remains at the same "ground potential". To the extent that this is accomplished, the less damage will occur during an event. An event can have 10KA or more current. So a resistance of 0.001 ohms will have a voltage difference of 10 volts. Very close to causing damage to the equipment.

 

[LionelHutz]

Wire resistance means little during a surge where the wire impedance will be much higher.

The 485 device in the field is the part which is failing? Is this field device grounded to the same ground bus bar as the SP? It could be possible the surge is hitting the device where it's mounted and not coming through the RS-485 wires from the control cabinet meaning the SP is useless where it's applied.

 

[guilio2010]

It is the 485 devices that fails in the field. The field device is not grounded in the same bus bar as the SP. The device ground is part of the power input it receieves which I believe is correct as if there is a short, this will return to the power source. I don't think the device is getting hit as this is a problem on multiple devices we have.

The SP must be the shortest point of contact to earth ground. This is the path from the ground lug on the SP to earth ground. Where does earth ground begin? If I have a ground rod 10' in the ground with a 20' 2 AWG wire, earth ground to me would be anywhere on that wire, therefore I need to run that into the bus bar in the cabinet and then have a 10 AWG wire less than 3' to that bus bar based on manufactures installations. Any surge will disipate that power into that ground. I'm looking at this first as some of the SP have 12AWG tied ot the bus bar with 10AWG that is 5-10' in lenght that is then tied into the ground wire. I been told that doesn't matter, but to me, I think it does.

 

[djs]

For best protection the SP and the field device must be tied to the same ground bus bar.