Analog, 24v dc, and 115ac in the same conduit

olias

Member
Join Date
Apr 2002
Location
North Jersey
Posts
185
Our company is building a new machine. Our mechanical eng. only gave me a 2 inch conduit across the machine. I need to run 115ac from the main panel to a sub panel 48 inches away. Also in this conduit I need to bring a shielded multiconductor cable for 24 volt dc I/O and a shielded twisted pair for a 4-20ma for speed control of a drive. The machine is already built. Anyone have any ideas to protect the analog signal from noise?
I wish some one would talk to me before they design and build things.

Thanks for any great ideas.
 
If you use twisted pair shielded wire for the analog and ground the shield well at one end you might get away with it in a short run. If not, most codes allow running low voltage wire exposed outside conduit. You could do that, and wire tie or securely fasten the signal wire to the outside of the conduit, on the bottom or as protected from physical damage as possible.

Why not just run another conduit? If this is a special machine, I can't believe that is the only re-engineering required. If it is a production machine they need to make the change anyway.
 
I'm looking into what change would be most cost effective. Drilling holes in 1" plate that just came out of the paint shop is ****ing the assembly crew off. I may be able to get rid of the 115ac which would solve my problem. I just like to have a plan B.
 
Have you looked into running Armored Teck cable instead. You may be able to run this along you suport structure, This way you are just looking at drilling and taping 1/4" holes every 3 feet. You could use tie wraps to attach it though.
 
;) In response to your dilemna, I would convey to you that our company has run analog, shielded DC, and 115 VAC together for years and do not have a problem with it as long as your shielding is properly grounded AT ONE END ONLY. This eliminates a ground loop and keep in mind that the insulation on your low voltage cables is still rated for 600 V.
Good Luck, Glenn
 
Thats true

I am not sure what the american code says, but the canadian code says you can run low voltage ac and dc in the same conduit as long an both voltages are from the same source. However as a prefered practice I still like to seperate them.
 
I would recommend keeping them separate. I understand that it will be difficult at best to do this however.
Running wires of different voltage together caused one of the worst machine startup I have had.
The machine manufacture lost 2 VFD's while testing at their location and never bothered to figure out why they were destroyed. They shipped the machine anyway (This was the last unit of 5 similar machines ordered). Which left it in my court to figure out. I also fried 2 drives and when my supplier delivered the third drive (That I had same day shipped) We worked on the problem together (for quite some time, didn't want to cook drive #3) and came to the conclusion that the 120 VAC control wiring was inducing a higher voltage on the 24 volt DC control wires for the drive, that were in the same conduit destroying the control card within the drives.
We change the drive control circuits to 120 Vac and use relay's local to the drive to switch the 24 volt DC.
The machine manufacture never documented a voltage difference between the wires on the drawings and never recommended using a different conduit.
Just my two cents for what it's worth
 
olias,

Just to muddy the water a little more, if you look at it from a theroetical viewpoint (not my favorite view as you old-timers know!), then according to Faraday's Law, a shielded cable provides just as much protection from an induced voltage as a separate conduit does!

A shield is a Faraday cage and is still only a shield, whether it be a conduit or a wrap-around bunch of wires.

You did say that your 24 vdc I/O was inside a shielded cable, and so is your 4-20 mA analog signal cable. I have seen the day when I would have loved to have as good noise protection as you have got here, separate conduit or not!

Notice that noise protection and induced voltage protection are totally separate issues from the big SAFETY issue. Is it as safe to run low voltage and high voltage cables in the same conduit? That is a question of a different color.
 
While it is certainly bad practice and against code in some places, the reality is that it is done all the time with fairly consistent success.

I don't like it and always recommend against it, but I don't seem to be able to stop it.

I will, however, refuse to warranty-certify a drive with wiring this far off from recommendations.
 
I too have run ino this problem all to often. Had GE Factory reps in on several problems, and in a couple of instances, all dc lines had .01 to .001 disc caps placed on the dc lines. We had one project where we had RF getting into the wiring (Bird wattmeter coupled to one shielded pair read 3 watts). Suggest MOV's on AC to try to hold spikes down and prevent a high on a control line from starting a machine at an opportune time. While not pretty, I have ty-wrapped lines to coax. Some Plant Managers and Engineers object because they want "NEAT" installations, on some I have strapped "Smurf Tube" (flexible plastic conduit, locally only available in Smurf Blue), which neatens things up a little.

4-20 line should be okay, but as previously stated by others, BUT DO NOT ground it at both ends, unless you are really adventuresome. I have spent many hours troubleshooting machines that were connected that way for years or had the insulation wear thro and short the shield to ground. You can get some really wild happenings.

regards.....casey
 
Dick DV,

I can argue that it may NOT be such "bad practice" after all to used shielded cables in the same conduit as opposed to separate conduits. As I started before, the little electrons skipping along those wires do not know the difference between conduits as shields and wire as shields. They are perfectly happy with shielded cables.

Here is what I see in the field: Where the installer runs separate conduits for signal and power cables, he then argues (successfully) that his signal wires do not need to be shielded. Therefore, they come out of the conduits into a control cabinet and twist and wind their way THROUGH a bunch of high voltage and mixed voltage wiring, maybe for 3 feet, maybe for 10 feet, finally arriving at a terminal block, maybe with some extra induced baggage voltage.

On the other hand, if shielded cables are use for all signal wiring, and run in one conduit, then the shields can be extended right up to a grounded terminal next to the signal terminal. This way, the unshielded cable length is cut from 3-10 feet down to perhaps 2 inches. I will choose a 2-inch unshielded wire versus separate conduits any old day.

For this reason, shieled cables actually are more effective in reducing noise than separate conduits. Now the safety question is a different issue. Some local codes allow some mixed voltages, others do not. Me thinks they are tilitng at windmills, as the cables are all mixed together at the ends.
 
I worked for a company for 20 years. When I started with them there were a lot of problems with instruments going haywire and or failing. What we found was when the plant was built, the engineer had all the signal wires ran in the same conduits as the control wires, and they were all mixed up in the control panels. What we had to do to cure these problmes was to run seperate conduits for all the low voltage analog wires and seperate conduits for the 110 volt control wires. AND, seperate the wires in the panels to their own voltage level wireways. Sometimes you can get by with it and sometimes not. The only way to be sure the system would work the first time correctly was to seperate all the wiring. I do use twisted sheilded wire for analog signals but I put them in a seperate conduit from control wires, I'be had too much trouble in the past.
 
micah123,

Were your problems in the past caused by not having separate conduits, or by not properly shielding or separating the wires INSIDE the contorl panels? Based on all my 30 years in the field, I vote for "not separating them in the contorl panels". It does no good at all to run separtae conduits for 200 feet, then mix the wires all back together the last 3 to 10 feet inside the panels. Guys, I say again, if you use shielded cables for ALL signal wiring, then you will not have to worry about using separate conduits. Separate conduits becomes a non-issue if your signals are shielded. Without complete shielding, it is impossible to keep the voltages adequately separated in the panels, as you will find that high voltage wires and signal wiring will run to the same motor starter (or other device).

Using separate conduits is a cop-out. It is a cheaper method that has been promoted by electrical contractors as being better, but it is not as far as noise reduction is concerned. Yes, I said cheaper. Think about it. It is a lot quicker and cheaper for the electricians to use separate conduits, than to pay for shielded cables, and buy extra space in the panel for more terminal blocks for all those grounded shields, and then take the extra time to terminate all those shields. Whereas, on running separate conduits, once you have to install hangers for one conduit, the second and third ones are freebies--just make your trapeze hangers a little longer, no additional supports required. The only extra cost is the conduit material cost, which for a large bundle of cables, will be small in comparison to the labor cost of terminating all those shielded cables. In case you haven't noticed it is a new world where labor is much more expensive than materials, versus 30 years ago when materials were high compared to labor costs.

Your thermocouple wires should be shielded also, if you want reliable steady temperature readings.
 
I only have some minor disagreements with your post, Lancie. Most have to do with the in panel wiring. First of all, it is impossible to completely separate analog wiring in a panel, if for no other reason than you have to connect to I/O module terminals only inches apart. It isn't practical to use shielded cable on internal panel wiring for 4-20 mA loop powered devices, for example, because of the conections to the power supply.

In our panels we use separate wire duct for analog and 120 VAC wiring, but the separation is, reallisticly, minimal. We don't use shielded wire on the internal analog connections from the field terminal strip to the I/O card, or from the power supply to the field terminal strip. We haven't had problems with noise, EMI, or RFI in the panels ever, and that's with nearly 20 years of doing it this way.

I don't have a theoretical basis for this, but intuitively it makes sense that length of exposure to the EMI/RFI source will increase the induced voltage and current. I can't imagine that a parallel run of 1" of non-shielded analog 4-20 mA cable and a 120 VAC 20 Amp power conductor will have the same effect on signal integrity as a 5,000 ft. parallel run of these same two conductors.

Again, I can't cite references, but I don't agree with you about separate conduit not being effective. This is partly based on field experience and anecdotal evidence. Besides, EMR/RFI signal strength follows an inverse square law. Using separate conduit is certainly going to impose some physical separation between conductors.

EMI/RFI is a peculiar thing. I'm reminded of a story a crusty old veteran told me many years ago. He built a HeathKit radio (tube based, of course) and it wouldn't give him decent sound. He verified his assembly and wiring, and finally called their technical support. The technician told him "It sounds to me like the problem is, you are too neat. If all of your wiring is in neat bundles and the wires are all parallel you will pick up a lot of interference. Take your bundle apart, twist things around and have them criss cross, and you will be OK." My friend did this, and got many years of top quality sound out of that radio.
 
Last edited:
I was on a job installing an updated turbine control panel. During my cleaning up of the wiring mess that the previous contractor had done, I discovered that the Thermocouple wire was run on top of the 13KV generator cabling inside the high voltage vault. I went through all the trouble of verifying all the shields coming into my panel and separating all ac from low voltage DC inside my panels and now I discover the really high ac voltage and the really low DC voltage are mingling together. We decided since the discovery was last minute and there was no previous problem that we would lift the TC cables off the 13KV and hope for the best. It’s been running for two years now with no problem. Of course the TC cable is twisted. The twisting cuts back on inductance since the twists are shorter then the wave length of 60hz.

Also I’ve had to install ice cube interface relays upstream of a PLC to prevent 40 vac of inductance from triggering the input. The relay acts as a terminator that bleeds the induced voltage. This was on a conduit run that was ½ mile long.

As far as safety goes, why would it be unsafe to run 110 ac with the analogs? Is it on a GFCI protected circuit? Are the other wires 600 vac rated?

My experiences with ground loops have been with different ground potentials between the field and the control room. So that if both ends of the shield are grounded current flows on the shield and the analog is elevated and erratic. It makes no difference that there are only analogs in the conduit when this happens. I doubt there is much difference in the ground potential of your machine and controls that are only 10 feet apart. This is assuming proper equipment grounds of course. Even if the shield was grounded on both ends in your case; how would that be different then a conduit grounded on both ends?

These are some things I’ve run across over the years. I’m no engineer so keep that in mind.

Andy
 

Similar Topics

Hey guys were having difficulties with a flow meter that takes 24VDC PSU but sends 4-20ma output as we verified. The Compact Logix analog module...
Replies
17
Views
10,445
Hi guys, I've a field instrument (actuator) using 24V AC, 0-10V. The terminal connection as below: Terminal 1 = 24VAC Terminal 2 = 24VAC Com...
Replies
3
Views
3,921
Hello every body, I have an Analog I/O card with the input signal range of 0 to 5 V, 1 to 5 V, 0 to 10 V, –10 to 10 V I want to use it with 24v...
Replies
4
Views
3,542
New at this so here it goes, Could anybody give me a basic overview of what this type of FLEX I/O does, its an '8 input analog module...
Replies
2
Views
4,822
Hi We're having issues with an FX1N-1DA-BD Analog output module. We're using an metallic contamination sensor MCS 1000 to detect particles. This...
Replies
1
Views
60
Back
Top Bottom