Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Ground Loop Problem

  1. Sep 19, 2015 #1
    Hello, i have a really annoiyng problem , i have a robot which is equipped with a video camera(PAL) so nothing too fancy, the information is transmitted through a coaxial cable about 200 m of cables,the attenuation of high frequency is not to big to cause any problems,also the robot has some motors ,this motors take about 3 amps and the cable has a total resistance of about (2 ohms/200m),the problem is that when i turn on the motor the video starts to become brighter until reachs a saturation level.I know that the problem is ground loop which produces a voltage on shield of coax cable , but i don't know how to remove this dc offset in ground of video.I tried a common-mode choke but without success.
     
    Last edited: Sep 19, 2015
  2. jcsd
  3. Sep 19, 2015 #2
    Are you intentionally using the coax as a power cord? Or is this some sort of unintentional effect?
     
  4. Sep 19, 2015 #3
    The robot is powered up by another two conductors which have their own resistance, the idea is that the ground coax is going in the same general ground so i can't break the ground loop because will still exist some resistance in conductor that will compromise my composite signal.I'm thinking about a isoltation transformer or a differential transmitter.I don't know how good is that type of transformer and where i can find one at a international distributor like Farnel or TME.
     
  5. Sep 19, 2015 #4

    dlgoff

    User Avatar
    Science Advisor
    Gold Member

    Best way/place to eliminate noise is at the source. Have you tried, say a 0.1μf to 1.0μf capacitor across the motor leads? Noise could be causing an offset.
     
  6. Sep 19, 2015 #5
    I will try, i forget about EMI caused by motor.i tried to decouple ground using a 10nF it's ok but the dc bias of composite signal is gone so i have some issues with syncronization of vertical an horizontal.
     
  7. Sep 19, 2015 #6

    nsaspook

    User Avatar
    Science Advisor

    The best solution is to use differential signals so the shield ground on the (new) twisted pair video transmission cable can be galvanically isolated from the power loop. You can also try a DC-restoration circuit to recover the proper sync levels. Depending on the signal level sometimes only a simple signal diode (in a clamp circuit) after the isolation capacitors on the input and ground will restore sync-levels to something usable.
    https://www.maximintegrated.com/en/app-notes/index.mvp/id/3303
     
  8. Sep 19, 2015 #7
    Thank you for proper answear.
    Do i need a shielded twisted pair ?
     
  9. Sep 19, 2015 #8

    nsaspook

    User Avatar
    Science Advisor

    Yes, for a 200M run with lots of electrical noise. It's common to use shielded/screened Cat5E cable for CCTV installations for long runs. The impedance is usually close to 100 ohms instead of the standard 75 ohm for coax so use the correct termination resistor on the differential receiver to reduce signal reflections.
    http://www.farnell.com/datasheets/313294.pdf
     
  10. Sep 19, 2015 #9
    I understood, thank you for help.I am not too good about transformers and stuff like this, i want to know how this galvanic isolation works if i don't have any type of AC signal on cable like DC voltage which supplies the power for device.If you would like to explaine me will be helpful.
     
  11. Sep 19, 2015 #10

    nsaspook

    User Avatar
    Science Advisor

    It's normal to see this problem when you have the power and signal grounds for a remote device with common connections on both ends of a long cable with a small wire resistance that develops a small voltage across the ends as current flows in every parallel ground wire back to the voltage source. It might be possible to totally isolate (if it's all low DC voltage) the camera power/signal ground and supply in the robot if they are don't share a common wires in the robot tether and only connect the camera ground to the common power ground point at the controller. Galvanic isolation works by removing the direct electron path in a circuit. The signal energy is then isolated from the total end to end circuit loop current by space into two loops and is coupled between those loops by the magnetic field of the isolation transformer or the electric field of the capacitor for any AC signal.
     
  12. Sep 20, 2015 #11
    Would you like to explain me how ground loop works ? From my knowlegde i know that the ground loop is produced when are more paths to ground with different resistances and this two(three) resistances behave like resistor in parallel.But some people advice me to add a series resistor on shield of coax with a value between 10 to 100 ohms to "isolate" the current.But i know that the resistors in parallel have the same voltage across them.Thank you in advance !
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Ground Loop Problem
  1. Ground loop question (Replies: 11)

Loading...