Floating voltage across Field Point measurement points

[944GTR]

My application uses two FP-1601 and eight FP-AI-102 to track numerous voltage sources across a broad range of potential differences. Specifically the voltage difference across individual Industrial Lead Acid Batteries when attached to SCR based chargers. The system was designed to track charge and cooling times, etc.

Recently a problem developed that causes the system to read charger/battery activity when no actitivity is present. The Field Point channel diplays a flucuating voltage, the trace current involved is so minute that using a jumper across the two measurement points nullifies the phantom signal. The channel then drops to a ground level. Because of the localized concentration of chargers a great deal
of electromagnetic radiation is present and it was noticed that the voltage displayed changes as the charger output cables are moved within the radiation fields. The interesting parts include but are not limited to the following details.
1)The voltage being tracked appears to be DC in nature
2)The three problem areas in question are in relativley close proximity to one another.
3)All the voltages appear on channel 5 of different FP modules within the same bank

How prone to noise pollution are the FPs?
What is the minimum amount of current neccasary for readings to be taken by the modules?
What could possibly be rectifying the voltage to make it appear to be DC?
Is it possible that the system has a grounding problem somewhere?

Sincerely,

Ben

[Aaron_G]

Ben,

1) The FieldPoint and Compact FieldPoint lines are tested according to CE and other standards for EMC/EMI vulnerability. For more information, refer to the Declarations of Conformity for each product (available online at ni.com->support->Product Reference->Product Certification->Distributed IO). The D.O.C.'s may specify mounting or shielding conditions required for a product.

2)The input circuitry on the [c]FP-AI-102 modules uses a calibrated resistor divider network (1 Meg Ohm input impedance per channel) on each channel. With the maximum measurable input of 120 Volts, the total expected current flow through a channel is 120 uA. With the 1 Meg Ohm input impedance, each volt of input provides 1 uA of current flow through the input resistor circuitry.

3) All channels of the [c]FP-AI-102 share a common "ground" within the IO module (user accessible via the Com and C terminals). The common for each IO module's inputs is actually isolated from the backplane/terminal base ground (network module C terminals) and can float (within the published specifications) at a different potential from the backplane/terminal base ground. This isolation of commons from module to module will be bypassed if you have wire the C or Com terminals from one terminal base to the C or Com terminal of a different terminal base. Thus, there is the possibility for ground loops to occur if the transducer "grounds" for two different modules are at different levels and the modules have the C or Com terminals tied together. (Side note: some people wire the V/Vsup and C/Com terminals of multiple modules together to share a single power supply, however, many input modules and some output modules do not require external power, so this is unneccessary).

It sounds like you have already localized the problem to certain wires and physical locations for EMC purposes. Depending upon the nature of your current wiring solution, you may want to try either twisted pair or shielded cables in the problem area. Additionally, you can also check the grounding of the chargers within the affected area.

Regards,
Aaron

[Ben]

Ading to what Aaron said...

Have tried to move things around to see if you can narrow down what is happening?

1) You said the errors show up on channel 5. Have you tried swapping channel 5 with another channel? If the issue stays with channel 5 then look towoard the FP bank and its wiring.

2) You did say it was ALL of your channel 5 inputs. Have you tried swapping a "good" 102 with a "bad" 102?

3) As you go through this adventure keep track of what was where and how thing were changed. There is always the possibility a problem with an external device could be damging the FP 102's.

4) "Noise Reduction Techniques in Electronic Systems" (Ott) is an excelent reference if your investigation shows that the problem is in the wiring.

Trying to hel
p,

(another) Ben

[944GTR]

"Aaron"

Thanks for all the information. After additional diagnostic work the following clarifying points were discovered.
The Industrial Battery Chargers used, incorporate long unshielded output cables fabricated from 1/0 line such as would be used for Welders. It seems that the attachment technique used to establish the measurement points make said cables long heavey gauage extensions of the signal wiring which is shielded line.
If similar cabling is attached to the FP wiring and moved within the EMI generated by the Charging system a varying amount of voltage is displayed on the Channel associated with the experiment, but only when in close proximity to a EMI source.
Incidentially, the racking on which all the equipment is mounted shows a positive
potential WRT the Common Points of the FP wiring, which in turn are attached to the numerous voltage sources being tracked. This is not suprising given the tendency of Lead Acid batteries and SCR Chargers to leak somewhat.
I am concerned that the presence and activity associated with the "signal wiring extensions" or any parasitic trace current is giving rise to the voltages being spuriously measured by the FP system. The delta between the amount of voltage that could be generated by the looped wiring within an EMI field and the amount being shown on the FP monitoring gives me reason to question whether the internal circuitry within the FPs is being pushed out of balance.

Any thoughts? Anyone?

Sincerely,

Ben

[944GTR]

"(another)Ben"

Thanks for the insight and the reference.

Ben

[Ben]

"Any thoughts?"

Yes, consult "Ott"!

His book is a real "nuts and bolts" approach to noise reduction.

Not having the book handy I will resort to my memory.

Ott emphesises (sp?);

1) Kill noise at its source (can you twist your 1/0 wire pairs?)

2) Confining the noise (better shielding)

3) Preventing the entry of noise (Shield the FP harware better, Move the FP away from the source [RF decays at 1/r^2 if memory serves me] or posibly some ferite beads on your signal leads just before they enter the FP EMI enclosure).

4) Eliminate noise after the fact (filering).

Ott probably has some other ideas in his book. I recall him going into a discusion of the benefits of using braided ground wire to handle high current spikes in ground
leads.

Another Ben