What are effects of using shielding on motor cable?

I have some long runs for PWM 3con 14 for brushless dc motors 24DC typically 2-4 Amps max. I never use shielded cable on motors but in this case the runs are 80-150’ and running nears lots of data lines. I can prevent a lot of parallel tight proximity but there are lots of intersections. Does shielding to Gnd one end at the driver have any effect on the signals? With only a DSS scope and volt meter on hand is there any test to see the reduction of noise coming off the cables using shielding? How about put a scope on some cat 5 pairs near the motor lines at varying proximities to see the effect of crosstalk? I don’t even know if any visible noise from an unshielded cable indicates that shielding is warranted. Any advice welcome.

Comments

  • evanhevanh Posts: 8,390
    edited 2019-11-29 - 15:36:48
    First up, 24 Volts is probably not a big deal for requiring this. But for 230 Volt motors, shielding is vital when using "inverters". Basic rule is do it regardless.

    I've not actually tried measuring myself but a short length of hookup wire on the end of the scope probe should do it. The concern is radiated noise so, to compare, all you need is a defined repeatable distance from the cable of concern. I don't know what can give you a metric.

    Yes, shielding will help. Connected to mains earth. Shouldn't matter if one or both ends is earthed. If the drive has a specific terminal for shielding then use that too.

    Another help is adding a filter at the drive - since the drive is the source of the noise. This helps also for the capacitive losses. The longer cables will incur a lot of losses on the sharp square waveform without a filter. It literally heats the cable, even without a motor. I won't be much help on component selection though. I've always just installed what the drive suppliers provide.
    We have the vastness of the internet and yet billions of people decided to spend most of their time within a horribly designed, fake-news emporium of a website that sucks every possible piece of personal information out of you so it can sell it to others. And they see nothing wrong with that.
  • Thanks for those tips. I’m going to see what the scope will show. No idea what any noise will mean in the real world as far as if it’s a real problem but at least I can observe shielded vs unshielded.
  • So you are using 24 VDC / PWM / data lines, on 3-conductor #14 wire on lengths of over 100 feet. The 24 VDC at 4 Amps is going to have a significant voltage drop over those distances - depending on your setup you might loose up to half of your Power Supply voltage. There are reference & lookup tables for voltage drop, but try it and see. Any PWM or data signal is going to have cross talk. It depends on how the data cable is constructed, shielded, twisted wire, and impedance matching, data & baud rate, all contribute to noise. There are IC driver chips that will help in the design phase, and you can buy kits commercially. Unless you already have an installation in use and/or similar to the one you are now constructing - try not to do that.
  • MicksterMickster Posts: 1,525
    edited 2019-12-01 - 18:02:57
    You're within 10% volt drop and I doubt that you will have a noise issue.

    Edit: Oh that was with a 30m run. Didn't pick up on the 150ft.
    Failure is not an option...it's bundled with the software.
  • Had a 400+volt, 100HP Vector drive, once.
    Light was low inside the panel and when the axis went under load, there was like a plasma over the cable's insulation. How that didn't mess up the encoder, I do not know. I felt like Tesla :nerd:
    Failure is not an option...it's bundled with the software.
  • When I buy D.C. Brushless motors and driver amps the cables are pre-made and shielded. So shielded is a definite especially on. a 100ft.
  • I used to buy Emerson D.C. Brushless servo. I think it's now control techniques
  • The key thing T Chap is whether is your own site or not?

    If it is, do what you want until you see glitches appear on your data, then you'll know that you need to increase the separation

    If its a customer site, you really should shield the cables. Is it just cost of the cables you're worried about? Are they a big percentage of the job?

    I've been at factories where they have had all sorts of issues due to a shielded motor cable that didn't have its shield tied to ground. So the whole thing was a leaky radiator along its entire length, causing the more sensitive equipment around the factory to have errors in measurement
  • I think I’ll install shielded for the motors and clutches and other solenoid 24 stuff. Once installed there is no further access to the wiring.

    Part of the question is what are the effects of using shielded on motor wires. I have a stigma from many years ago where I put motor wires inside a metal conduit that was tied to ground and the motors stopped working. I never solved it, just ran the wires outside the conduit. Twilight zone experience that I never got to the bottom of, maybe too tired! Who knows, but was checking to see if any anomaly could occur with PWM wire inside a shield. This is more to avoid screwing up somebody else’s AV lines.


    Voltage drop can be dealt with with larger supply.



  • Too many odd phases in one conduit causes induction and overheating. If your running three phase power that requires double wiring (six or more conductors) you have to run another conduit. So only three phases per conduit. Steppers run up to four phases. You probably got some induction spikes in the line.
  • long lengths of shielded multiconductor motor cable causes a capacitive effect from each phase to shield.
    If you are using PWM to make sine waves this causes more load on the transistors, which is usually fine up to a point, usually there is a maximum cable length spec.

    No shield is better than a floating shield disconnected on both ends. The energy on the shield wants to go back to the drive that induced it. I cringe when I see the shield terminations that end right at the entry inside an enclosure, and then the 3 wires go many feet up to the drive... It's like NO, now your whole cabinet is resonating and radiating, but you see it everywhere. The newer 480V VFD's and servo drives have pluggable motor connectors that include terminals for the 3 phases, ground, and shield to make it idiot-resistant*

    *no such thing has idiot proof.
  • Why don't you just use DC brushless motors with intergrated controllers. Then you don't have to worry about pulses over your 100 ft. run.
  • The motor has no access once installed. I can’t put electronics at the motor as the odds of the motor failing are almost zero. The odds of the driver failing is a potential risk factor so it must accessible. I found 14g 4 conductor today with shield. I wonder if there is any test I can run with a scope to see the difference between the 3 signals going to the BLDC under some load. Compare both shielded and unshielded and note the difference. Also try to observe crosstalk to an adjacent cat5 and compare I’d mainly like to understand the difference.
  • Don't think I'd be too thrilled with that machine designer. What happens if a motor coupling fails?
    Failure is not an option...it's bundled with the software.
  • Mickster wrote: »
    Don't think I'd be too thrilled with that machine designer. What happens if a motor coupling fails?

    +1. Serviceability should be a major design consideration.
    In science there is no authority. There is only experiment.
    Life is unpredictable. Eat dessert first.
  • Do you know the pwm switching frequency of the drive? Many vfd's seem to be in the 4 to 15 kHz range
  • Tubular wrote: »
    Do you know the pwm switching frequency of the drive? Many vfd's seem to be in the 4 to 15 kHz range

    https://www.onsemi.com/pub/Collateral/MC33035-D.PDF

    That is the controller IC I use. I need to look at the schematic to see what timing resistor I put in that controls the frequency. I'll check this eve.

  • I took a 15’ 4con 18g shielded cable, Using 3 wires for the BLDC motor, not under any real load. Probably 1 amp if even that much. Along side the motor cable I wiretied a cat5 cable that was cut at each end. Took a scope and connected a lead at the controller side to a random wire(doesn’t make any difference). There is also the motor sensor cat5 in the bundle that handles 3 hall sensors 0-5v and 2 encoder signals 0-5v on this Cat5 to the motor.

    The signal cables to the motor are 15' unshielded Cat5, single ended. The encoder signals and hall sensors are terminated at the main board( Prop board) with a 100ohm and .047uf cap. See image for the termination method.

    Channel 1 Showing 1 volt per division , this shows the crosstalk of motor PWM on the unterminated. unconnected test wire.

    Channel 2 Shows 1 of 3 Hall sensors These only have a termination resistor 100ohm and .047uf cap to GND as termination.

    Channel 3 Shows 1 of 2 Encoder signals post Schmitt Trigger.

    Channel 4 Shows the same encoder signal pre Schmitt trigger.

    The images show the noise with the motor running, both with the motor cable shield connected to chassis ground and without. You can see the obvious noise difference on the non terminated test cat5. But the cap and resistor termination take care of most of it, and I have never had an issue with this design. The newer project includes up to 150' runs, will be using 14G stranded shielded wire 24V PWM, same motor and controller, but the encoder and hall sensors are over RS485 drivers and are very clean. This is more of a test to see what I might be doing to other peoples AV cables, there are tons nearby. There is no access to the cables once installed.

    One images shows my method of termination into the main box off the RJ45. Into the cap and res, then into the Scmitt trigger, then to prop.
  • This shows the scope connected to the test cat5.
  • but the encoder and hall sensors are over RS485 drivers

    :cool:
    Failure is not an option...it's bundled with the software.
Sign In or Register to comment.