Grounding shield of 4-20mA signal adds noise?

kolyur

Lifetime Supporting Member + Moderator
K
Join Date
Oct 2004
Location
Wooster, Ohio
Posts
1,601
I'm really stumped on this one. I have a Turck linear transducer connected to a 1762-IF4 analog module on a MicroLogix 1400, configured for 4-20mA. The run is about 25 feet of 2-conductor shielded twisted pair (Belden 9463). I grounded the shield drain wire on the panel side only. Looking at the sampled value in RSLogix, the signal is very stable (+/- a few counts) until I turn on any of the 480VAC motors, at which point the noise increases to a few hundred counts. Portions of the 4-20 cable run along 480V cables so I pulled them out--no improvement.

On a whim I disconnected the analog cable shield from ground and the noise immediately disappeared, even with many motors running. I don't have a lot of experience with noise on analog signals but this is completely opposite of what I would expect to see. Any suggestions?
 
Have you checked the reliability of your ground connection? A good policy is to provide a separate "clean earth bar" for screens at each panel and a separate safety earth bar / stud for metalwork earthing.
 
If it is ungrounded check your DC common relative to ground.
I have found AB analog inputs on various modules get unhappy if DC common is more than .5 volts from ground either way.

Have had to tie several systems DC common to ground.
 
Have you confirmed that the cable you have doesn't ground the shield through the connector on the field end?

Where exactly did you ground your shield on the panel side?

Keith
 
We normally don't ground our DC common but in this case I tried it both ways... no effect on noise.

Ground quality might be questionable so I'm checking that now.
 
kamenges said:
Have you confirmed that the cable you have doesn't ground the shield through the connector on the field end?
Click to expand...
Good point, yes I did verify that the shield is not connected to the plug body on the transducer end.

kamenges said:
Where exactly did you ground your shield on the panel side?
Click to expand...
Same location where all my VFDs are grounded :oops:
 
Have you tried connected the shield at the transducer end only?
I have found some transducers that require the shield connected at the transducer end of the cable.
 
if cable/quality ground condition ok,

try take out the transducer from touching earth/metal or put side and manually test/simulate the transducer.

also check vibration spec
 
I really don't have a way to ground the shield at the transducer end. It's a M12 molded plug.

I'm pretty much convinced that this is a problem with my ground. The signal gets noisy even if I run a motor that is nowhere near the transducer. At this point I have the analog cable draped along the floor so there should be no 480 interference. The common denominator is that everything is grounded on the same backplane. All of the VFDs are grounded together in a series of terminal blocks and this is where I was initially grounding the analog shield. I have since moved the shield ground to other locations on the backplane with no improvement.
 
I've run into this before. Although it defies common practice, sometimes grounding the shield at both ends actually helps. In other cases not grounding the shield helps. It is easy to try these - if it works, then go for it.

One case that drove me nuts was traced back to a Siemens VFD in the same building. We knew that the problem only occurred when the pump on that VFD ran. The VFD was grounded properly, but a control card internal to the drive wasn't grounded properly to the main drive control. Ran a two cent jumper and a $100,000 problem disappeared.
 
Shielding is to reduce noise issues, if shielding makes it worse, why would you use shield?

On another note. We run grounding on all piping etc. Zero noise problems.
 
TurpoUrpo said:
Shielding is to reduce noise issues, if shielding makes it worse, why would you use shield?

On another note. We run grounding on all piping etc. Zero noise problems.
Click to expand...

In reality that's the approach you have to use on site, customers aren't usually interested in paying you any more money if they think that you've "fixed" the problem. It's not really right though...
 

Similar Topics

B
Shield Grounding Info
Happy New Year to all! More interesting opinions about shield grounding. ------------- Answers to shield grounding question...
Replies
1
Views
4,787
krk
K
J
4-20mA loop grounding
Hi all, I'm connecting several 4-20mA sensors together in parallel (only one shown below) The enclosure is ABS plastic with metal backplate DAQ...
Replies
5
Views
261
Jenny8711
J
T
Power Supply Grounding
I have Rhino 120 to 24 dc power supply and it has 2 Positive and 2 negative terminals on the DC side but no ground terminal. Should I connect one...
Replies
9
Views
432
lfe
L
R
EMV grounding machine
Hi, Our customer has bought a machine a CNC wood cutting machine. It states that it need a seperate 16mm2 EMV"ground" in addition to our 5G16mm2...
Replies
2
Views
598
rQx
R
F
Beckhoff AI module - huge bouncing! Grounding issue?
I have a Beckhoff PLC with many AI modules. Most devices are loop powered using the same +24 VDC power as the Beckhoff Bus Coupler and all is...
Replies
5
Views
662
I_Automation
I_Automation
Back
Top Bottom