4-20 ma cabling


Thread Starter

Pat Levinson

Hello list members, I have a question. What would be your comments on using "normal" house-wiring or non-instrumentation-grade cables for 4-20 ma links? Normally we use special shielded cables (belden 8761 for example) as recommended by PLC manufacturers, grounding one end only, etc..

I realize (and have experienced) the importance of
special cables for noise reasons in analog voltage
signals, but for 4-20ma loops in a light industrial scenario, where not much noise is expected from large motors, contactors, or VFDs, do you think we could do without special cables, as long as the loop resistance limitations are met?? Any comments, recommendations, or links to the subject would be very welcome. By the way, I remember hearing one time (possibly on this list) that twisted pairs of cables are NOT recommended
in 4-20ma loops, or that separate single cables should not be twisted together in 4-20ma loops. Any comments on this also?

Thanks and regards, Pat

Darold Woodward

You should use shielded twisted pair cables. The twisting helps prevents loops
for induced currents that can corrupt the signal.

You do not need to use anything fancy for 4-20ma as the overall impedance characteristics are not important for 4-20ma as it is not digital data. If you are using some type of system that multiplexes digital communication on the same
wire as the 4-20ma signal, however, the cable type is more important.

I generally use #16AWG TSP for instrumentation as it is much less likely to be damaged during installation by electricians used to handling 500 MCM electrical conductors.

The only consideration is the insulation level. 300V is adequate for most installations, but if your cable routes through or starts in a VFD or other cabinet that contains higher voltages, you should use 600V insulated cables.

You should also verify that the cable jacket is suitable for the installation. While many people use it, PVC jacketing materials are generally not rated for wet locations. Underground conduits are considered wet locations by the electric code and often contain standing water that can degrade the jacket. You also need to choose the correct jacketing material if you are routing through
cable trays.

You should find, however, that 4-20ma signals operate well over very reasonably priced cables.

Darold Woodward PE
SEL Inc.
[email protected]
Hello Pat

I think for the field wiring upto the control room we require special cable , that is individual pair shielded and with twisting, now this depends upon the what is the possibility of noice pickups in the field, and when we use twisted pair noise pickups and magnetic field developement cross cancel eachother. and that's why European countries goes for Higher Twistings per meter.
Braided and twisted signal cabling is not enough today with the walky-talky around and other RF gadgets. The braided type + continuous aluminium foil provides for cleaner signals.
Your two ordinary wires will act as antenna.
They don't have enough common mode noise rejection. Try it but don't cry after ! ! !
Please climb into my time machine, back to ~1960. There was a brand new oil refinery wired up with this new fangled 4-20mA (Taylor) stuff. The installers used 24-core, and found a great convenience - they could clip a phone onto any wire in the cable, and listen on any other one (it had a comms core). Came startup day, and all the temperature transmitter signals were reading the noise on the flow transmitters. Three months later, the plant had twisted pair and the electrical contractor hasn't been seen since. You may get away with no shielding, but dont try to get away without twisted pair.


Jake Brodsky

Ummmm, if I'm not mistaken, most instrument cabling is twisted, even inside a shield. I don't know of any good reason NOT to use a twisted pair for a 4-20 mA loop. In fact, I think it would be an improvement over using something like, say, zip cord in a conduit.

I would hesitate to use stuff like CAT 5 cable, though, simply because I don't know enough about its current handling characteristics over longer distances. Use the wire gauge to try and determine how much resistance to expect.

And to answer your question, yes, shielding does help. How much it helps, on the other hand, is something I suspect many of us wonder about, but haven't had the opportunity to try. The risk/benefit ratio simply isn't there.

I think the big reason we use shielding is to lower the common mode noise. The less common mode noise you pick up in an instrument cable, the less noise you'll get when the signal hits your I/O card. In other words, if you already have a fairly high CMRR at the input of your I/O card you may not notice much difference compared to a shielded cable.

And then again, even if you did see a difference, the instrument itself may be so erratic and inprecise that you may not care.

Good Luck!

Jake Brodsky
mailto:[email protected]

Al Pawlowski

For 4/20 mA process signalling, twisted pair should always be prefered over any other type I know of.

Other types can be used, however, but the loop noise immunity will suffer. Just how much depends on how closely, and evenly, an individual loop's source and return line are spaced to each other.

Whether the reduction is too much will depend on the specific layout. For instance, I have used "zip" cord for multiple loops within control panel wire trays without encountering any problems. On the other hand, "zip" cord probably would not work well if you tried to run any significant distance alongside large relay, or motor, power lines.

The aluminized mylar shields typical of twisted pair cable are another matter. They can be useful for high speed, high impendance or milli-volt
wiring. However, I have found them more trouble than useful for 4/20mA process loops. They take more effort to "dress" nicely at terminations and
the shields usually get connected so as to cause otherwise avoidable ground loops.

Al Pawlowski, PE
[email protected]
dba ALMONT Engineering
Baton Rouge, LA USA

Glenn Sahlin

I hate to jump on the bandwagon, but the benefits of a twisted, shielded pair far outweigh the cons of the noise problems you can stumble into. Personally, having crossed paths with a few noise devils in the past, any analog signal in one of my systems is gonna be twisted, with braided under foil shield. Plain and simple, cable is cheap, your time and reputation are not.

[email protected]