Induced Voltage Causing False Readings

[dlfrmtn]

We are presently installing a system using FP-DI-330 input modules over an ethernet network using Lookout software on Widows XP systems.  All of the signals to the input modules are 120 volt signals going to dry contact alarms in the filed at different buildings at our plant, varying in distance from approximately 10 to 2000 feet.  Where we have a single pair of wires goint to a particular alarm this arrangement works fine.  Where we have several alarms run together we have false alarms that toggle on and off 8-10 times per second on adjacent channels along with the expected channel alarm.  Can you offer any suggestions as to how we might be able to overcome this problem?

Thanks in advance for any assistance that you might offer.

[Aaron_G]

How do you have the inputs wired? The FP-DI-330 has differential inputs that can be wired fully independently, or you can wire some of the terminals together to share a common supply or return. From the information in your posting, I would suggest that you make sure you are only using InA and InB for each channel and not wiring the InA's or InB's to each other. It seems that with some runs being 10 feet and some being 2000, you may have some ground loops being created which may be causing the false readings. If you connected the InA's or InB's together, and have voltage differentials due to the run lengths, that could cause the false readings. Also, make sure that the C or COM terminals are not connected to C or COM terminals on other terminal bases. Please note that the FP-DI-330 inputs are differential and isolated from each other, but the module does not have a safety rating on the isolation voltage of one channel to another, only from the input channels to the backplane. 

Regards,
Aaron

[dlfrmtn]

Alarms have a common neutral that is tied to InB on all modules.  The COM terminals on the bases are not connected together.  I beleive that voltage high enough to trigger alarms is being induced into the signal wires for the adjacent channels that are run along with the "legitimate" alarm wire.  I should mention that the channels with the short wiring runs do not have this problem.  I was hoping there was a solid state device that is commonly used in this situation that could be connected to the circuit that would require the voltage to exceed a set threshold before the channel would receive the signal voltage, say 100 volts for instance. 

[Henrik_Volkers]

you can use Z-Diodes (80V rating) together with a RC network

120V DC ? what load is acceptable?

[Aaron_G]

The [c]FP-DI-330 input circuitry has the following thresholds per the manual:

Off State Max:

DC  +/-1 V

AC +/-1 V

ON State

DC   Min +/-4V

50/60 Hz AC 15VAC

1 kHz AC  3 VAC.

So it's not that there aren't thresholds, they are just lower than you want. As for solid state electronics to change the threshold, you could always look at adding a resistor to decrease some of the induced voltage. You may want to use a DMM to measure just how much voltage is being induced on the longer wire runs.

Regards,

Aaron

Message Edited by Aaron G on 12-29-2005 06:57 AM

[dlfrmtn]

We currently have each circuit fused at .25 amps.  We have tried using a resistor network reducing the voltage to 25 volts without success.  The diode sounds like it might do the trick.

[dlfrmtn]

We have tried a resistor networks reducing the voltage on the active alarms to 25 volts and still had the problem previously described.  The voltage appears to vary with the number of alarms coming in.  We have recorded readings from millivolt range to as high as 50 volts.  Sounds like a diode may be worth a try.  Thanks for your response.

[Henrik_Volkers]

Another trick:

the fasle alarms are due to a high dI/dt (inductive coupling) or dU/dt (capacitiv coupling) in your lines. Sloving down things can help.

Another approach: I think  Weidmueller / Schneider/ Phoenix etc sell (galvanic) level decoupler, maybe that's worth a try