4-20mA wiring

NetNathan

Lifetime Supporting Member
Join Date
Nov 2011
Location
Corona, CA
Posts
2,191
I have 15 signals that are going from signal conditioner to PLC Input module.
They are all 4-20mA and there is nothing else in the wireway. Length is less than 2 feet.
Should I run twisted pair shielded or is just straight wire okay (I use Blue + and Blue/White -)?

I know the color codes may not be to everyone's liking :).
 
Use shielded wire! There my not be anything else in the wireway now but may be in the future. Use shields wire and sleep better every night.
 
Despite what others will say, in my experience these sorts of short runs are not significantly affected by not using screened cable, provided you don't have high current / high noise generating equipment in close proximity.

The PLC input card isn't installed in a faraday cage after all.

Field wiring, for sure - screened cable all the way to the device.
 
Hardly ever bother connecting the shield to ground with 4-20ma signals these days anyways unless there is a problem - pretty immune I have found.
 
4-20 ma was developed specifically to reduce noise from external influences so is reasonably stable as it's the current that is monitored rather than the voltage so more immune to spikes, saying that it is always best to use screened cable (not convinced twisted makes a great deal of difference) and screened one end only, this stops earth loops as you should also ensure the sensor equipment is earthed and at the same potential as the PLC ground.
 
Despite what others will say, in my experience these sorts of short runs are not significantly affected by not using screened cable, provided you don't have high current / high noise generating equipment in close proximity.

The PLC input card isn't installed in a faraday cage after all.

Field wiring, for sure - screened cable all the way to the device.

I've seen this done in numerous applications without any problems in control panels that only contain PLC equipment, but I'd personally use shielded cable anyway simply because I'm a stickler for doing things by the book.




4-20 ma was developed specifically to reduce noise from external influences so is reasonably stable as it's the current that is monitored rather than the voltage so more immune to spikes, saying that it is always best to use screened cable (not convinced twisted makes a great deal of difference) and screened one end only, this stops earth loops as you should also ensure the sensor equipment is earthed and at the same potential as the PLC ground.

This is not true. The advantage of a current loop over a voltage signal is immunity to voltage drop on long cable runs. 4-20 mA current loops are immune to noise because of using twisted pair cable and shielding, and frankly, the twisted pair is the most effective of the two in reducing EMF interference with the signal. The shield primarily helps to keep static charges off of the wire and attenuate high frequency interference. If you are going to expose the wiring to an electrically noisy environment, then a shield alone will not provide adequate protection. The cable should be run inside a galvanized steel conduit in such an environment, since you need a ferromagnetic barrier to attenuate the EMF noise.




4-20mA isn't a differential signal so using a twisted pair cable shouldn't make much difference.

Some are. Some aren't.
 
Last edited:
I've never had an issue with 4-20mA without a shield, ran many between 50-100m without a shield even in the same cable way with 1000A local db supply running alongside it.

People say not to share a common as well between multiple signals, well, I break that rule as well, never had an issue. I have 20 4-20mA signals sharing a common in a 25 core unshielded. The only thing I bother to shield is high speed signal or data
 
When I was doing an instrumentation course the instructor told us that the twist is important with 4-20 ma and voltage signals as well as thermocouples and the like. The theory was that if something was induced into a cable it would be induced in the reverse direction at the next twist and almost always effectively cancel the induction out - not sure if this is true or not. That was a long time ago LOL.
 
despite what others will say, in my experience these sorts of short runs are not significantly affected by not using screened cable, provided you don't have high current / high noise generating equipment in close proximity.

The plc input card isn't installed in a faraday cage after all.

Field wiring, for sure - screened cable all the way to the device.

+1
 
When I was doing an instrumentation course the instructor told us that the twist is important with 4-20 ma and voltage signals as well as thermocouples and the like. The theory was that if something was induced into a cable it would be induced in the reverse direction at the next twist and almost always effectively cancel the induction out - not sure if this is true or not. That was a long time ago LOL.

Did you ever look inside a Reliance DC drive? All the signal wires, not many 4-20 I would say, were highly twisted. It was explained the the high twists per inch would cancel noise. My guess there were 10 or more twists per inch and no shielding. They were a decent reliable DC drive.
 
Did you ever look inside a Reliance DC drive? All the signal wires, not many 4-20 I would say, were highly twisted. It was explained the the high twists per inch would cancel noise. My guess there were 10 or more twists per inch and no shielding. They were a decent reliable DC drive.

The twists are what exposes both poles of the wiring equally to the electromagnetic field, thereby inducing a voltage that cancels itself out.
 
If you think about it all are true, reduced voltage increase or drop do not influence current as it is a constant source current and by default small voltage spikes from interference also will not affect it and I agree that twisted pair will cancel out noise. However, I have seen installations on PT100 probes that are 100m long with a resolution of 0.5 degree resolution using screened untwisted 3 core (including Silfex 3 core cable) without any issues however I personally don't like it.
 

Similar Topics

Hi All, I have a question about the input wiring on a HMISCU8B5. I am so close to having what I am after yet it seems so far away. I have...
Replies
34
Views
7,130
Ok, so I have 2 boxes with separate power supplies they are going to be about 300 to 400ft apart. the Main Box as the PLC in it with the Analog...
Replies
19
Views
5,130
Just wondering if it is possible to wire in a Cynergy UF08B100 Flowmeter into an Allen Bradley 1734-IE4C. I have tried every way I can think of...
Replies
13
Views
3,138
Hi guys, I have a Metler Toledo M800 that will be used to read mainly my ph and conductivity sensor on my system, i saw on the terminal block i...
Replies
1
Views
1,676
Hello, I have a float switch that is labeled as below and has only 2 wires (red,black): Input 10-30VDC Output 4-20mA Resolution 5mm I am...
Replies
11
Views
3,292
Back
Top Bottom