1
\$\begingroup\$

BNC cables are used as inputs to a data-acquisition hardware. If long BNC cables are picking up the 50/60Hz mains hum through the air like antenna, how can I filter this? Would connecting the BNC conenctors' GNDs to the earth help? Or using a capacitor?

\$\endgroup\$
  • \$\begingroup\$ looks like a duplicate: electronics.stackexchange.com/q/200193 \$\endgroup\$ – Nick Alexeev Nov 12 '15 at 0:33
  • \$\begingroup\$ they are different questions \$\endgroup\$ – user16307 Nov 12 '15 at 0:33
  • \$\begingroup\$ You bnc cables are likely not the problem. RG58 has about 45 db of shielding. More likely it is your device. Coax is ground shielded \$\endgroup\$ – johnnymopo Nov 12 '15 at 0:48
  • \$\begingroup\$ what do you mean by 45 db of shielding? BNC would not pick up EMI? \$\endgroup\$ – user16307 Nov 12 '15 at 0:50
  • \$\begingroup\$ All coax will have shielding from external sources. RG58 or other typical bnc coax doesn't have a lot compared to speciality cables. Coax is meant to be used as an unbalanced guide. The shield and connectors should be grounded \$\endgroup\$ – johnnymopo Nov 12 '15 at 1:04
0
\$\begingroup\$

Co-ax and twisted pairs are usually good at removing common mode noise.

Ground loops in coax installations can be a problem as they cause unbalanced noise. This becomes more of a problem if the co-ax shield carries any stray or intended 50/60Hz current.

Make sure signal co-ax screen is not used or involved in current paths that include conductors other than internal core. Running the Co-ax parallel with unbalanced high current 50/60Hz conductors could ealisy cause stray currents to flow in the screen if it is grounded at both ends. Any current on the screen that is not on the core will show up as noise.

Also not that with co-ax capacitive pickup should be minimal if the sensor is shielded. Electromagnetic pickup should be cancelled out except at the terminal points.

You should make sure the grounding at the aquisition end is as good as possible and the co-ax screen is isolated from ground at the sensor end. Grounding the sensor housing locally is usually still a good idea (unless it is small and can be left to float with at co-ax screen potential) but having the co-ax shield grounded to two unequal grounds is always a bad idea.

\$\endgroup\$
  • \$\begingroup\$ How can I test if signal co-ax screen is not used or involved in current paths that include conductors other than internal core? \$\endgroup\$ – user16307 Nov 12 '15 at 9:11
  • \$\begingroup\$ If you make a break in a section of cable screen you can see if sensors fail to work or by inserting an ammeter see if a current is flowing or investigating the sensor circuitry to see that is is using an independent ground return wire for power (and other things) and not sharing the co-ax screen. \$\endgroup\$ – KalleMP Nov 14 '15 at 0:00
1
\$\begingroup\$

There are a few things you can do (any one of which may or may not work, and I would try them in this order - don't remove previous steps, just keep adding on):

  1. Ground everything to a common ground
  2. ferrite chokes
  3. isolation transformers
  4. invest in higher quality cables if shielding is inadequate
  5. use a 55Hz 20Hz-wide notch filter to remove the hum (not preferable, but as a last resort)

Also, see this website: ARRL RFI. I am a member of this organization and they know what they are talking about.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.