electromagnetic shielding

Discussion in 'The Projects Forum' started by yassser, Jan 21, 2013.

  1. yassser

    Thread Starter Member

    Jul 25, 2011
    91
    0
    hello,

    I'm making a controller to controller a milling machine , the machine has a box where the control buttons should be arranged , I have the electronic control circuits , and I also have some buttons that control some 220v ac contactors , the problem is that the magnetic field caused by the currents in the contactors wires cause the electronics to misbehave , sometimes completely halt , unfortunately it would be impractical to seperate the contactors buttons from the electronic control buttons .

    The electronic circuit is enclosed in a metal box , my question is to what should I connect the metal box to have a shielding effect , I connected it to the machine body and there was no effect , I connected it to the neutral line , and there was no effect either , there is no difference in effect between when the metal box is closed , and when it is open and the circuits are exposed .

    Is there something I'm doing wrong ?

    I forgot to mention that these malfunctions happens only when I change the state of the contactor , particularly when I switch a contactor off , I thought it must be that the current in the coil is stopped suddenly , so the magnetic filed rate of change got high and induced high voltages in the control circuits , I don't know if I'm right about that.
     
    Last edited: Jan 21, 2013
  2. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    4,771
    971
    Well without a schematic and diagrams of component/wiring locations we cannot do anything but guess. Anything thats metallic that may become energized in a fault situation or is to be used for shielding must be grounded. Have you verified the machine body is grounded?

    And what makes you think its the magnetic field that is causing a problem?
    And define "misbehave"?
    Could it be the voltage spike/flyback caused by the switching of the contactor instead? (much more likely)
     
  3. yassser

    Thread Starter Member

    Jul 25, 2011
    91
    0
    can I ask what does grounded means ? does that mean that the machine body is connected to metal bar which is inserted in the ground ? or just connected to the neutral line ?

    misbehave means : that the lcd starts writing strange characters and symbols in strange places , sometimes the microcontroller halts and stops receiving orders.

    I forgot to mention that these malfunctions happens only when I change the state of the contactor , particularly when I switch a contactor off , I thought it must be that the current in the coil is stopped suddenly , so the magnetic filed rate of change got high and induced high voltages in the control circuits , I don't know if I'm right about that.

    and I don't know about the voltage flywheel spikes , the contactors circuits and the control circuit are completely isolated electrically , how could the voltage spike affect the control circuit?

    thanks very much
     
  4. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    4,771
    971
    schematic?
     
  5. yassser

    Thread Starter Member

    Jul 25, 2011
    91
    0
    [​IMG][​IMG]

    the problem happens when the electronic control unit is placed near the green lines(which control the contactors through the buttons , and it mainly happens when a contactor is switched off.
     
  6. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    That is not a schematic, it is even hardly a block diagram. You know, the details like how is the contactor driven, with what voltage, etc. actually matter.
     
  7. yassser

    Thread Starter Member

    Jul 25, 2011
    91
    0
    [​IMG]

    I'm sorry , I was just trying to simplify it , tell me if there anything else I should explain
     
  8. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    You need to explain a LOT more. What is the powersupply to the ECU? Where is the physical position of the wires with respect to the contactor wires? How exactly does the ECU control the "power electronics" part?
    Really a complete schematic would be much better than asking a thousand questions.
     
  9. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    This may be a lot of things. In such applications it is common to optocouplers to isolate the control unit from the more high power parts. And also using separate power supplies for the two
     
  10. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
  11. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    When opening a contact that carries current there will always be sparks. These emit RF, often strong enough to influence nearby electronic circuits.

    Try to put a capacitor over the contact that you believe to be the culprit. I suggest something like a few nanofarad, high voltage capacitor, e.g. a common value is 2kV.

    I experienced this problem several times and this solved it.

    If your sensible electronics was really good shielded against RF with a metal box then the only way it can be disturbed is through the wires that go in and out. They act as antennas.

    By the way, you cannot efficiently shield a strong disturbing magnetic field with a metal enclosure alone. The material would need to be ferromagnetic and have a very high permeability like Mu-metal, even better having several layers that do not touch. It does not shield but concentrate flux away from where you don't want it.

    If you want to shield something against RF make sure you don't have holes in your enclosure and ground it. In and outputs need to be properly filtered/terminated.
     
  12. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    If your relay is DC controlled. You will also need a flyback diode(use Google)
     
  13. gootee

    Senior Member

    Apr 24, 2007
    447
    50
    Also, examine the wiring itself. Every pair of wires that form a circuit should have NO SPACE between them. Otherwise they will make an ANTENNA, for time-varying magnetic fields, which will induce time-varying currents in the wires. So it's best to use shielded twisted pair, for each pair, with the shield connected at one end only, to the chassis. Next-best would probably be to just tightly twist each pair of wires, all the way to each end, probably with at least 4 turns per inch.

    And keep any low-level signal wires away from wires with large or fast-changing currents. If they must approach each other, they should be at a 90-degree angle to each other.

    And if there is a ground-return wire, to the power supply, from anywhere, make sure that low-level signal grounds do not share the same ground-return conductor as any large or fast-changing ground-return currents. Run separate wires for any low-level or reference grounds.
     
  14. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,067
    3,836
    You are likely getting a decent reverse voltage surge kickback from your relays (contactors) when they turn off. You need to add reversed biased diodes across your contacts and possible other places. Make a. Hand-drawn sketch of relays and power supply for your controller.
     
  15. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488

    The high voltage would appear across the coil, not the contacts. The contacts will cause sparks, a diode wouldn't do much across the contacts, a (small) capacitor does.
     
Loading...