Sheilded Cables and Grounding

Discussion in 'General Electronics Chat' started by WarGoat, Jul 19, 2013.

  1. WarGoat

    Thread Starter New Member

    Jul 19, 2013
    10
    1
    Hi all

    I have two cables that, by necessity run next two each other. One carries power for an EL wire and the other carries analog sensor data.

    When the EL Wire is active it creates substantial noise for the analog sensor data.

    To combat this I expect I will need some form of shielded cable. My question is what can I use to ground this cable - given that my power source is a 12VDC adapter from a wall plug. Essentially I do not have access to the 'ground' in the wall.

    I look forward to any thoughts.
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    I would shield the sensor cable as the first priority, since it might receive noise from sources other than your EL power wire.

    The shield doesn't necessarily have to be grounded to block interference. But you could ground it to the sensor ground.
     
  3. WarGoat

    Thread Starter New Member

    Jul 19, 2013
    10
    1
    Thanks Wayneh, I'll be trying to shield it but I'm not sure what you mean by 'sensor ground'.

    The sensors are basically variable resistors (which resist based on what they sense) and aren't powered outside of a simple current passing through them.
     
  4. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    I'd try leaving the shield not connected, or connected to the low voltage side. See which works better.
     
  5. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,507
    2,367
  6. WarGoat

    Thread Starter New Member

    Jul 19, 2013
    10
    1
    Interesting manual Max but I'm not sure it's that applicable. All the examples it contains include a circuit ground (your objections are noted here) and/or a metal case to terminate on.

    My enclosure is entirely plastic so, assuming I acquire a cable shield (tube) with which covers my cable reasonably well what do I do either each end? If I leave it unattached isn't it the shield just going to act as an antenna?
     
  7. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,507
    2,367
    If you have service power inside the plastic enclosure or service (earth) ground available, you should terminate the service ground at a suitable terminal block or copper plate and terminate all shield and bonding conductors, if there is any metallic part to the equipment.
    Called star point grounding.
    Much depends on the equipment you are using the sensors on?
    Presumably there is more to the system than a wall-wart and a couple of sensors?
    Max.
     
  8. WarGoat

    Thread Starter New Member

    Jul 19, 2013
    10
    1
    Hi Max

    The enclosure itself contains raspberry pi unit which is powered off a custom board (from a 12DVC source), a multiplexor and several other components.

    It doesn't seem like a good idea to ground the shield a terminal of the 12VDC source (I have no basis for this other than it just feels very wrong and easy to short circuit).
     
  9. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,507
    2,367
    Short circuit to what?
    The Siemens PDF might seem a bit 'heavy' for what you are doing, but the principles involved in grounding and shielding are pretty much apply across the board.
    Anytime you have sensors tied to a unit, they usually go out into the 'real world' where noise etc can be imposed.
    If you don't feel comfortable earth grounding the 12vdc common, the shield should at least be grounded.
    Like I say, depends on the rest of the equipment.
    Max.
     
  10. WarGoat

    Thread Starter New Member

    Jul 19, 2013
    10
    1
    I'm not sure I follow your logic - you say the shield should be grounded - that's actually my original question - how do you do this? I have no earth terminal available. This is a series of components all powered of two 12DVC pins.
     
  11. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,507
    2,367
    You mentioned initially that you have two cables ran alongside each other and one has a adverse effect on the other, in order to try and combat that by grounding the shield, you require a earth ground from some point that you can attach the shield to in order to make the shield electrically neutral.
    In a case such as this, this is the first thing to attempt as a line of defense.
    If they both have the same power source, it may be necessary to connect the power common to earth ground, when faced with this kind of problem, it is often a case of elimination, starting with the obvious, which IMO, is first earth ground the shield.
    Max.
     
  12. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    FWIW, I've seen plenty of recommendations to NOT ground the shielding or connect it to anything. Seems counterintuitive but the advice was based on experience.
     
  13. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,507
    2,367
    Strange? from day one when low voltage logic systems were married to AC industrial controls, the stress was to implement grounded shielding on any low voltage signal sources.
    Initially for many years the stress was on only grounding one end of the shield, if you read chapter 6 of the Siemens PDF, the trend is now to equi-potential bond the system to earth ground and to ground the shield at both end, of course I am talking about machine control in general.
    If the shield is floating, then how does it effect the shielding process?
    Max.
     
  14. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    Actually found some data on shielding termination. Looks like grounding is good at low frequency (e.g. power line).

    I'll be very skeptical of the no-ground advice if I ever see it again.
     
  15. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,507
    2,367
    That appears to refer to GHZ transmission on RG223 Co-Ax?
    Probably out of the realm of this thread.:D
    Max.
     
  16. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,170
    395
    I would use twisted pairs with over all shield , shield connected only to neg. 12V. Finding = potential grounds is not always easy or possible. " once measured 3V of hashy 3 ph 60 HZ from building frame from first floor to 3 rd floor.
     
    strantor likes this.
  17. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    You're partly right, I missed the "MHz" label on the axis, thinking it was Hz. Oops. But the co-axial testing in Fig.3 included lower frequencies down to "0" MHz. No idea what their lowest frequencies actually were.
     
  18. strantor

    AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    Nothing wrong with connecting the shield to the 12V common. Your fears are unwarranted.
     
  19. Duane P Wetick

    Active Member

    Apr 23, 2009
    408
    19
    When shielding a cable, remember to only terminate one end of the shield. This provides a one-way path for any induced currents. The higher impedance wire will generally be the one that gets any induced interference. Decoupling capacitors may be introduced at one end to provide a path for AC currents, but will block DC current. It is good practice to separate signal wiring from HV circuits (ignition) on 12V equipment.

    Cheers, DPW [Everything has limitations...and I hate limitations.]
     
  20. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    That contradicts most of what I could find on the issue, which recommends terminating both ends of the shield. So-called ground loop problems from grounding both ends were myth-busted in one source I found.
     
Loading...