This may have been answered before, but I couldn't find it by searching. Is it OK to run 24 VDC, 4-20mA and RS-485 in same cable? If so, should it be twisted pair individually shielded or one shield for all pairs shield?

As you can see, I am trying to cut down on the number of cables for this project. 24VDC will be drawing around 100mA.

Thanks

 

Yes,

it's OK. Twisted pair cable would be preferable. I have often mixed different signals like that without a problem.

Roy

 

Roy... although you have had success, it doesn't mean every installation will be successful:

See thread "1026211569" covering interference between 24Vdc circuits and data-communication circuits.

Regards, Phil Corso (cepsicon [at] aol.com)

 

I think it would be fine. Devicenet does it all the time. The 24VDC should ideally be twisted in a pair. The individual shields may be overkill, maybe only necessary if your 24VDC current is not clean. RS-485 is differential, meaning it will reject noise present in both conductors. The current loop should be fairly robust to voltage spikes.

Bill Sturm

 

Bill... the corrective action you recommend may be void if the installation violates NEC requirements. A common problem often overlooked by engineers!

The referenced thread in my previous post will tell you why?

Regards, Phil

 

I am not an expert in the NEC codes, but I have been in the controls industry for a while. I am certainly not aware of any requirements that disallow low voltage power cables to be run with communications cables. I stand corrected if they exist. I did review the other topics that you had mentioned. Does the issue have to do with the possibility of the 24VDC wires shorting against higher voltages in the field and then coupling with communications lines? I also understand that 24VDC can have high frequency transients, caused by switching loads and possibly switching power supplies. Some some noise and spike suppression might be a nice idea here. Maybe a good linear power supply also. How does Devicenet or POE get away with this? Do they have special noise filtering? Current limiting?

Bill Sturm

 

I am the original poster (after all these replies I decided to register for an account), and I find this topic very interesting. Thanks for all of the feedback. We (my company) is purchasing a handful of sensors that have 24VDC, 4-20 and RS-485 all in one cable going to the sensor.

The question in my mind began when I found out that the cables being delivered from the manufacturer were not going to be long enough. So I set out to find a cable capable of doing the job. The manufacturer wasn't much help when I asked for a recommended cable (I should have asked what cable they use!).

Also, another thing I am finding disturbing is that there is no RS-485 ground on the sensor. Just a "ground" for the 24VDC and shield. Maybe the 24VDC shield is used for the 485.

For what it's worth, I haven't made up my mind on the cabling yet.

 

As Phil pointed out "It won't work in every case" however if you are not talking great distances it should be fine.

I don't believe the ground wire in RS-485 is used as part of the signal, just for shielding.

If it's part of a large installation with Home run cable and field junction boxes I would use twisted pair instrument cable. If it's just part of a skid or machine you can probably find something suitable in the Belden catalog.

Roy

 

If the RS-485 transceivers get their power from the 24VDC supply or at least share a common ground point, then no RS-485 ground is needed. In fact, having one would create a ground loop. The 24VDC common will work for both circuits.

Bill Sturm

 

Bill... I never said power and communicatons circuits were disallowed in the same cable!

Certainly, your EMI abatement suggestion, as well as short-circuit consequences are covered by the "GEP" umbrella. I just pointed out that there are legal restrictions to consider... espscially in the litigious society in which we are immersed.

Of course if Boss's location doesn't use NEC, then my caveat is moot.

Phil Corso

 

That's a tough question to answer. In general, yes. To be safe you should use individually shielded pairs in case the RS-485 or 24VDC line is terribly noisy. If cost is a huge issue, you could try the overall shield, but you have to recognize that you might have some noise between the conductors. Having said that, both RS-485 and 4-20mA are fairly good with noise immunity.

KEJR

 

Phil said: "I never said power and communicatons circuits were disallowed in the same cable!"

Okay. What DID you say, exactly?

Phil again: "I just pointed out that there are legal restrictions to consider..."

Like Bill, I am not aware of any legal restrictions on using 24VDC power and communications in the same cable. As he points out, DeviceNet and Power over Ethernet both do this, and Profibus has hybrid cables that have power and comm. Heck, in AS-i and gigabit PoE the communications is on the same WIRE as the power! I'd feel perfectly comfortable using (low-voltage) power and comm in one cable, including, god-forbid, defending the decision in court.

-James Ingraham
Sage Automation, Inc.

 

James... thanks for making my point, which was that engineers often overlook a very common problem!

The problem I alluded to was that the insulation rating of all wiring must equal or exceed the maximum circuit voltage within the cable.

By now, some of you have already applied the Italian hand-salute acknowledging the fact!

Regards, Phil

 

I know about the rules regarding insulation voltage ratings in a raceway, and I agree it is commonly overlooked. It could easily be considered negligence if it caused any harm. I figured it wouldn't matter in this case, since the OP was planning to use all low voltages, 24 volts or less. In my experience, most cables are rated at least 250 or 300 volts.

Bill Sturm

 

If given a choice between 24VDC and 24VAC, which would be preferable in terms of noise reduction?

Is a 60Hz sine wave less noisy for a bus running at e.g. 9600Hz than is a switching supply at maybe 50kHz?

 

Pitchblende...

1) what is connstruction of the RS-485 cable.

2) What is conctruction of the 24V cable?

3) What is current-magnitude (ma or A) in the 24V cable?

Regards,
Phil Corso

 

Phil,

Thank you for your response and sorry for my delay.

1) I am considering a Draka TXOI(i) cable which has four separately foil-shielded twisted pairs of 0.75mm² conductors. The cable is then collectively shielded and braided.

2) I intend to put the power bus in the same cable, on a separate pair.

3) I expect the entire bus to draw around 1A day-to-day.

 

Pitchblend... your choice is an excellent one, BUT, two caveats:

1) Ground all foils and sheath at one end... preferably the source-end if the source, too, is grounded!

2) The 1-Amp current carried by power-pair presents no problem! However, if either conductor is intentionally earthed, and an arcing ground-fault occurs, then with certainty its foil as well as others will be involved! Whether or not it is problematic depends on current-magnitude (hence damage) and transient-duration!

Regards,
Phil

 

Phil

In that case I will opt for a DC supply with built-in current limiting and not earth the supply-ground.

I realize I need to ground the shield all the way but I also know that message can be hard to convey to the electricians. Guess I will put an image of what I mean inside each device.

Thank you for your help.

 

Hi,

>it's OK. Twisted pair cable would be preferable. I have
>often mixed different signals like that without a problem.

I have question.. it is ok to run a MSTP bacnet network with 24vac in same conduit?

Why I am having a 18v reading from shield to 24v and 14v from 24v to +/- of STP wire even the STP wire is floating from ground.