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: Planes need to be grounded to something?
Answer: It depends on what is meant by the word “grounded,” because there is an awful lot of general confusion associated with issues of grounding/earthing.
As my slides said, metal planes are valuable as “image planes” – providing shielding benefits even if electrically isolated from the circuits they are protecting.
And – as my slides said – using them as current return paths for DM currents provides additional EMC benefits, which usually (but not always) means they are connected to the 0V DC power rail.
And also – as my slides said – using them as return paths for CM currents helps provide low-impedance local paths for these strays, again providing EMC benefits.
All of the above EMC benefits are only obtained when the metal plane is much closer to the circuits concerned than one-tenth of the wavelength at the highest frequency that we wish to control for emissions or immunity. Preferably closer than one-hundredth of the wavelength, e.g. < 3mm for up to 1GHz.
Notice that none of the above have any requirement for “grounding” to a safety ground electrode. I can’t answer any better than the above because the question does not say what it means by the phrase “grounded to something”.
For many decades there have been many huge myths surrounding the word “ground”, both in circuit design and EMC design, which I tried to dispel in my webinar.
When an electrical safety engineer talks about grounding he or she means a low-series-resistance (usually < 1 Ohm) connection to the electrodes in the ground – the soil on which the building or railway line sits – which is also connected to the neutral of the secondary of the mains power supply HV distribution transformer that supplies the building or railway line.
The aim is for any insulation failures in the mains wiring to result in a current flow into the ground that is so high (because of the low-resistance grounding network) that fuses or circuit-breakers in the mains supply’s phases open very quickly to reduce exposure of personnel to electric shock risks, and also so that the energy released in the fault is insufficient to cause a fire or explosion (or, in high-power systems, to limit the energy released in the fault to that which can be safely contained by the enclosure).
When a lightning protection engineer talks about grounding, he or she means a low-series-impedance (usually < 1 Ohm) connection to the electrodes in the ground (i.e. the soil).
Notice that this is not the same as the electrical safety engineer’s grounding requirements. The electrical safety engineer is only concerned with 60Hz (or 50Hz in some countries, or 400Hz in some parts of aircraft, or 16.67Hz in some railway systems) – and so he or she only needs low resistance, which can be achieved by long wires if they have sufficient cross sectional area.
But the lightning protection engineer has to deal with lighting surge currents, the energy spectrum of which is generally considered to peak at around 10kHz and extend up to about 10MHz. Any wires or braid straps longer than about 500mm have too much inductance to achieve the low-impedance required for surge suppression, so he or she must use other metal structures, usually interconnected in some form of mesh, to be able to control lighting surges as their currents are routed back to the soil from which they originally came (in the form of lightning bolts).
(And yes, I did mean to say soil, and not clouds. The charge that accumulates on clouds with respect to the soil, that causes lightning strikes to the ground, comes originally from the soil.)
When a static-control engineer talks about grounding, he or she means a reliable conductive connection of any sort to the safety grounding structure of a building – and sometimes they will prefer ‘grounding’ connections that have very high series resistance, such as 1 Megohm.
When a circuit design engineer talks about grounding, he or she means a very low-impedance DC power supply rail that is common to one or more circuits.
This is usually the 0V rail or plane, and it does not need to be electrically connected to the safety ground for the circuits to work fully to specification (otherwise, how could cellphones, iPods, laptop PCs, cars, aircraft, etc. possibly work?).
When an EMC engineer talks about grounding, he or she means providing a low-impedance path for a stray current to quickly and easily find its way back to its source, so that it doesn’t cause excessive fields that cause interference with other circuits. This is usually provided by a chassis or enclosure metal structure, which generally needs connections to all of the DC power rails used by the circuits that have low impedances over the frequency range to be controlled.
Note that “EMC grounding” does not need to be directly electrically connected to any DC rails (power or 0V), and does not need to be connected to the safety ground at all, to function fully.
The confusion and myths over the word “grounding” has arisen because it is common for all of the above five quite different “grounding” requirements to be met by the same metal structure – the chassis or enclosure metalwork that is connected to the electrical safety “grounding” network that is meshed to provide the lightning protection’s “grounding” network, and which is also connected to the electronic’s 0V DC power rail.
This fact has unfortunately led to people imagining that the connection to the safety ground electrodes in the soil are somehow important for a circuit to work, or for EMC mitigation techniques to work.
It has also unfortunately led to people imagining that – like the DC currents that flow from the DC rail to the 0V rail – AC currents (including stray RF noise currents) also flow ‘downhill’ to the 0V rail.
In fact, all AC currents at whatever frequency – and whether they are DM or CM – always flow preferentially in whatever paths they can find that have the lowest impedance, even if that means through stray capacitance in the air, or stray capacitance through an insulator!
But the functions of all these five different types of “grounds” are very different, so it obviously causes confusion to use the same term for each. This very confusion has cost many manufacturers very dearly, over several decades, in unnecessarily high development costs and time-to-market delays, and unnecessarily high costs-of-manufacture.
So I always strongly recommend never using the word “ground” (or “grounding” or “grounded”) at all, except when concerned with the electrical safety of mains-powered equipment.
-Keith Armstrong
