Editor’s note: This question was asked in response to Interference Technology’s recent webinar by Keith Armstrong. To view the webinar, click here.
Question: If the mains lead positive and negative are balanced to ‘ground’ does this reduce CM noise?
I assume ‘positive’ and ‘negative’ mean live and neutral, or Phase 1 and Phase 2 in the case that the neutral is a centre-tap in the distribution transformer’s secondary winding so both wires are ‘live’ (e.g. in the USA 220V mains is supplied in domestic premises as ±110V).
Yes, balancing the live and neutral (or phase 1 and phase 2) to ground helps to reduce the CM noise emissions from the mains cable but only for the DM currents flowing in that cable. Essentially, it helps prevent DM to CM conversion in the cable.
But it’s not a trivial thing to achieve balanced impedances on the live and neutral, or three phase, leads, over a wide frequency range. Careful, regular twisting of the wires is good up to 1MHz or so, but mains cables aren’t made carefully enough in general for higher frequencies so we often find ourselves having to clip large ferrite CM chokes onto the cable at the equipment end. Sometimes it is even necessary to space several ferrite chokes along the length of the cable as well!
It is important to note that having live and neutral (or phase 1 and phase 2) wires that are well-balanced around the safety ground wire, will do nothing for any CM currents flowing in the mains cable.
These CM currents could be caused, for instance, by inadequate shielding of an electronic product that allowed stray currents to flow out of its PCBs or from its heatsinks into the floor, walls, ceiling of a room or into any people or other equipment in that room. These strays currents mostly return to the circuits that “lost” them via cables, especially the mains cable, flowing equally on all of the wires in these cables – for example on the live, neutral and safety ground wires in the mains cable).
If these cables are shielded, the stray CM return currents will just flow on the outsides of their shields. It is possible to reduce some frequencies to some degree by fitting CM chokes, which is why during the process of suppressing emissions that are over the limit, we often find ourselves with two (or more!) clip-on ferrite CM chokes on every cable that connects to the unit concerned!
Because no-one likes to buy products that have their own weight in ferrite chokes clipped to all their cables, a more practical solution to this specific example of a CM current problem, is to reduce the stray capacitance radiated emissions from the unit, by better circuit design, better PCB design, better internal cabling design, and better shielding. (To reduce costs, these days most manufacturers use PCB-mounted shielding instead of enclosure shielding.)
Notice that all of these solutions require more in-depth design iteration than just throwing filtering and shielding at a unit that fails its EMC tests – which is why it is very important indeed, to help avoid financial risks (this is the only kind of language our bosses understand) to design EMC in from the start of a project – and not leave it until a new products is failing its compliance tests!
There are many other possible causes of stray CM currents, e.g. inadequate filtering and/or shielding of signal/control/data cables, poor PCB layout, poor choice of ICs, poor design of heatsinks, etc., etc.
-Keith Armstrong
