Radiated emissions from cable out of enclosure....solve with shielded cable?...or 3 core cable with earth?

Hi,
We have SMPS’s in metal enclosures which are connected by a two core, unshielded cable (as attached). The cable is acting as an antenna and causing a radiated EMC failure. Which would be the best solution for radiated EMC?.(ie give less RF radiation)....

1….Making the cable a three core cable with the third wire being an earth connection, connecting earth from enclosure to enclosure.

2….Making the cable a shielded 3 core cable, with the shield being at earth potential, and the third wire again being an earth connection from enclosure to enclosure. The Shield is a metal foil surrounding all three cables.

Sorry to post again on the same general subject area. However, I find posting different questions in the same post never gets an answer. This is definitely a different question though. And could help others who will find this by web search…which will direct them to this edaboard I believe.

you say 2 core - do you have a part number?

i like twisted shielded pairs
the twist reduces the reception (or transmission)
the shield, when properly connected at the ends, reduces the reception or transmission

what frequency are you using?
how fast is an edge?

treez said:

Hi,
We have SMPS’s in metal enclosures which are connected by a two core, unshielded cable (as attached). The cable is acting as an antenna and causing a radiated EMC failure. Which would be the best solution for radiated EMC?.(ie give less RF radiation)....

1….Making the cable a three core cable with the third wire being an earth connection, connecting earth from enclosure to enclosure.

2….Making the cable a shielded 3 core cable, with the shield being at earth potential, and the third wire again being an earth connection from enclosure to enclosure. The Shield is a metal foil surrounding all three cables.

Sorry to post again on the same general subject area. However, I find posting different questions in the same post never gets an answer. This is definitely a different question though. And could help others who will find this by web search…which will direct them to this edaboard I believe.

Click to expand...

2. would be best for screening cables but doesn't sound like thats the issue here.

You have two metal boxes - These need to be bonded to the same ground plane locally at each box to prevent the boxes radiating. You can't really use a wire to Bond as at RF the wire will act as inductor which are high impedance to high frequencies but also as a perfect antenna to radiate from too. You shouldn't really connect a earth screen both sides as you will likely cause Ground loops which will radiate also.

Have you identifed whats causing the noise? If its broadband its likely conducted emissions from the SMPS either in the form of differential or common mode noise. It maybe better and easier to deal with the source through some filtering. What frequencies are you talking about? I have identified the points in your system you would require filtering to address radited emissions. Also note the potentially screening of the Buck. Really you would need to look at each area identifed and filter as appropriate to get the desired result. The arrows indicate things that when not filtered could radiate.

Thanks very much.

You shouldn't really connect a earth screen both sides as you will likely cause Ground loops which will radiate also.

Click to expand...

I noted in your kindly given picture there is a common earth plane to which both boxes are connected to with separate connections. unfortunately, we cannot afford this luxury. The earth connection to the second (RHS) box is via the third wire of the interconnecting cable, where it is connected to earth in the LHS box.

Also note the potentially screening of the Buck.

Click to expand...

A metal cover was put over the whole buck converter but this had no effect. It was a metal cover made up of a bent bit of cardboard with copper tape stuck all over it.....and had a soldered wire soldered to the copper tape and to the chassis.

treez said:

2….Making the cable a shielded 3 core cable

Click to expand...

this is better, i think

some "tricks" from my experience in emc-lab:

1 low power led-driver (flyback ) - putting 50-100pF capacitor at drain-source of power mosfet - lowering radiated and conducted emi
2 cable filter(a kind of zcat1325) at input and/or output cable - lowering radiated and conducted emi
3 common mode choke at output ( a lot of driver have it) - lowering radiated and conducted emi
4 led pcb connection to "+, -" or earth - may affect radiated and conducted emi (need to test)

treez said:

Thanks very much.

I noted in your kindly given picture there is a common earth plane to which both boxes are connected to with separate connections. unfortunately, we cannot afford this luxury. The earth connection to the second (RHS) box is via the third wire of the interconnecting cable, where it is connected to earth in the LHS box.

A metal cover was put over the whole buck converter but this had no effect. It was a metal cover made up of a bent bit of cardboard with copper tape stuck all over it.....and had a soldered wire soldered to the copper tape and to the chassis.

Click to expand...

Thats how EMC is normaly measured, both units are connected to the Ground plane, the bit that goes between the two boxes is an interconnecting bundle. Thats the only way i think you can achieve a valid EMI test for your setup.

I'm afraid A metal box only works for RF if you Bond it to Earth Locally otherwise it will conduct and radiate like a wire. A wired earth is normally only used for electrical safety not RF. If your issue is Common Mode then you need access to the Ground Plane via your Box to be able to filter it otherwise you won't get to a solution with your current approach. If its differential Noise then that can be filtered via a LC differential filter.