I am onsite in Alabama for a water plant start-up, I have in one of my Remote panels 18 120 Vac inputs into Opto22 I/O for alarms, the inputs come in from a piece of equipment located Approximateley 400 feet away. we supply the 120 vac from our system to the equipment an on the equipment end they use interposing relays to give us a "Dry Contact" input back to our Opto I/O. the problem is I am getting from 34 to 70 volts on my inputs even when the the input is off, can anybody give any help on bleeding this voltage off to ground?
First of, what conduit did you use?? If it is plastic/PVC you will have to change it to steel. the steel conduit can be earthed every here and there and it should bleed the voltage. And try to space the steel conduit and any high voltage cables at least 250mm (10 inches) away from each other. This will minimize the effect EMI.
I would be looking for a defective ground or neutral connection. I would also verify that they have not mixed the neutrals between your panel and their panel. If I am understanding you correctly, your panel provides the 120vac that is switched by their relays, which then goes to the input of the opto-22 devices.
I would run one 120 vac circuit over as the supply to the relay contacts (you will not need a neutral for this, then daisy chain this remote power to one side of all the relay NO contacts (common side), then you bring back a switched hot from each realy contact NO contact (contact side). I would try this with one relay first. On the Opto-22 inputs the neutral side can be daisy chained from your panels power source. Verify you have solid ground and neutral connections with in your panel. It is unlikely you have a induced voltage from another source such as high voltage AC (would have to be fairly high current as well to induce a significant voltage on another ac circuit).
Contact the electrician who did the wire pulls and conduit work.
go to: http://www.almegcontrols.com/
then click FAQ then click "what a reed switch can do" You will see a circuit at the bottom of that page 1 and a description above it. Reeds are dry contacts and it may be the same situation.
Are your conductors properly isolated/shielded from other power circuits? I recall a project using 2k ohm, 10 watt or so resistors across plc inputs for bleed off.
The Radio Shack fix is to shunt a resistor to ground or neutral should the system be ungrounded. This should "bleed" off the induced voltage. This was always a problem in older PLC's when connecting solid state outputs to solid state inputs. False triggering of inputs due to "leakage" current. Experiment and find the largest value resistor that will do the trick. I would probably start with a 4k and if the voltage is still too high lower the value. Try not to make it too small though as this will increase the current load on the panel.
Hope this helps,
Another approach would be if you're already using "dry contacts" then simply make the contacts form "c". (1 NO, 1 NC, 1 COM) Wire them in such a way that when the input is not true it is tied to neutral or common.
First off, I assume that the "off" voltage is due to parasitic coupling. To verify this, you could carefully connect a load (resistor) to the off circuit and note the drop in voltage after applying load. You may find that some minimal load eliminates the induced voltage.
Secondly, the induced voltage may be common-mode, in which case it could be eliminated by input line filters. This could be verified by repeating the above and actually measure the current in the load resistor (via DMM or current probe). If the current is near zero (i.e. << V/R, where V is the apparent measured voltage and R is the value of the applied load) then the voltage is common mode and can be filtered.
Hope this helps.
fs your signals are in a screened cable then connect the screen to earth at only one end. The same for any steel walled conduit. Then check if there is an current in this screen by using a clamp on current meter. There should not be a current.
Then run a decent sized earth connection between the two points. 6 mm square copper would be good.
Again check for an earth current.
You have not said if the cables that you are running in the conduit are singles or not. If you are using a multi-core cable then make sure that the screen is terminated to earth at the source end. then sleeve the other end.Have you tried running a seperate earth wire inside the conduit?
Verify that the other guy's "dry contact" outputs are actually relays and not isolated triacs or other solid state outputs (like Opto 22 modules). Solid state outputs often leak enough current to create these kind of voltage levels on a high impedence input load. If this is the case, you can either persuade the other guy to change the modules to true mechanical dry contact output modules or add interposing relays, or you can add a load resistor in parallel with the input on your end.
I suspect you are either getting your signal back through a multicore cable carrying used power or that your signal cable is laid near a power cable. You need to load the return signal at your end. You could do it with a resistor to neutral (in parallel to the input) or by making the line work by asking it to switch a 'chunky' relay or contactor.