Best PLC Wiring

[Randyg08]

Question: The 4-20 mA signal my PLCs operate on is in a very noisy enviornment. The noise is primarily due to large magnetic fields. The first and best thing to reduce this noise I can think of is to use shielded wire.
What I'm looking for is there a best kind of shielding? ground or unground it?
Does having the wire be a twisted pair improve the signal? how tightly twisted?
Has anyone had particular success in this kind of situation? what manufacture for the wire did you use?

Thanks!

 

[Giz99]

I generally used a two conductor shielded cable with an overall shield such as a Beldon or equivalent 18 or 20 awg. Ground the shield of the cable at one end only to prevent ground loops.

Giz

 

[mordred]

I've used the Belden cables before in a situation like you described for a 900 ft length, Tying one end to ground as described previously and had no problems sometimes the ferrule cores help as well but most often I find that I don't need them make sure you are tying your grounds to one point and if your tying it to the backplane make sure you remove the paint may sound obvious but I've seen lots of equipment that forgets that simple step

 

[PeterW]

The first and best thing to reduce this noise I can think of is to use shielded wire.

The first and best thing is to try and avoid the noise, run in seperate conduits away from the noisy areas (minimum gaps etc).

The second thing is as they say.

 

[mordred]

I believe metal conduits is the better choice for noise as opposed to PVC

 

[MASEngr]

I believe metal conduits is the better choice for noise as opposed to PVC

The laws of physics agree with you.

 

[mordred]

gee thanks I guess gravity must therefore also work lol

 

[rootboy]

You're two for two!

 

[maintenanceman38]

All of the above I agree, Ground the shield usually a seperate wire surronded by the the shield itself, But your noise may be due to the fact that the wire is so long. If you think about this 10 milliamps over a long run of wire makes resistance and 10 miliamps becomes 2 miliamps and can vary with tempature. I think you should put the current source and the power source close. And send an output to the input you are seeking.

 

[jimbo3123]

All of the above I agree, Ground the shield usually a seperate wire surronded by the the shield itself, But your noise may be due to the fact that the wire is so long. If you think about this 10 milliamps over a long run of wire makes resistance and 10 miliamps becomes 2 miliamps and can vary with tempature. I think you should put the current source and the power source close. And send an output to the input you are seeking.

This doesn't make any sense. If your analog loop is truely a loop (and it typically is), then the current in all parts of the loop must be equal, regardless of resistive losses. That's one of the chief advantages of using a current signal over a voltage signal.

I don't see what the distance between the power source and current source has to do with anything.

 

[Buletjie]

If you have the budget, use a media converter 4-20mA/OpticalFibre/4-20mA

 

[plcsoftvsd]

a)Use Belden twisted pair shielded cable

b)Ground it at PLC panel end only.

c)Do not Pigtail the shield, instead strip open the insulation, and use a clamp to ground the shield to panel.

d)Keep atleast 300 mm distance from power cables ( especially VSd output leads)

Cheers.

 

[Bob A.]

In my experience, you should use steel conduit that is effectively grounded. Run shelded and twisted pair conductors with the shield connected at the end where the signal is used (PLC end).

If you have "noise" issues with 4 - 20 mA signals, it is more likely due to non-isolated inputs and dis-similar power sources that are floating with regard to each other. 4 - 20 mA is very low impedence and is therefore difficult to effect with induced noise by stray magetism.

Note that non-ferrous metals are transparent to magetism and are therefore of no use as magnetic shielding. This includes conduit and wire shielding. Their primary goal would be to shield against electromagnetic noise which is basically RF in nature like contact arcs and welding, etc.

Best Regards,

Bob A.

 

[monkeyhead]

I'd look into what Bob A. has said. Induced noise on a 4-20mA signal should be a rare thing.

Shielding, using twisted pairs, and remebering to only ground one end of the sheild are all good tips for reducing noise on any type of electrical signal cabling.

Another tip that has arguable consequences is to not coil up the excess cable at the ends of the run since coils of wire are inductors.

 

[mordred]

I'd look into what Bob A. has said. Induced noise on a 4-20mA signal should be a rare thing.

Shielding, using twisted pairs, and remebering to only ground one end of the sheild are all good tips for reducing noise on any type of electrical signal cabling.

Another tip that has arguable consequences is to not coil up the excess cable at the ends of the run since coils of wire are inductors.

S lay the excess not in a loop coil