• Any exposed metal (conducting) part (e.g., the chassis or output cables) is capable of causing an electrical shock to the user. To prevent a shock, such parts must be earthed and/or isolated from the high-voltage parts of the power supply in some way.

• No single point failure anywhere in the equipment should lead the user to be exposed to an electrical shock. There should be two levels of protection, so if one gives way, there is still one level of protection available. That is the basic premise underlying safety: the probability of two unconnected/independent faults occurring at the same point and at the same instant is considered negligible.

• Levels of protection that are considered essentially equivalent are (a) earthing of the exposed metal surface, (b) physical separation (typically up to 4 mm) between any exposed metal and parts of the circuit containing high voltage, and (c) a layer of approved insulator (or dielectric) between any exposed metal and the high voltage. Note that in (c) this (single-level protection) insulator/dielectric must have a minimum dielectric withstand capability of 1,500 VAC or 2,121 VDC.

• To qualify the above slightly — connecting the metal enclosure of the equipment to Earth is not always considered an acceptable level of safety and protection, simply because earthing may still not be guaranteed. For example, the wiring of older houses may not even contain an Earth wire inside the common household power outlet. However, assuming for now that earthing is acceptable, we then need one more level of protection as per the safety concept. This could be “4 mm” of separation for example (IEC 60950-1 provides tables for knowing exactly how many millimeters are required, and where). But consider the case of a high-voltage MOSFET (switch) mounted directly on the (earthed) metal enclosure (say, to provide better heatsinking). We obviously can’t provide “4 mm” of physical separation in this case (between the MOSFET and the earthed enclosure). So, now we need to place one layer of approved insulator between them. In this position, the insulator is said to serve as “basic insulation.”

• If earthing is not considered acceptable to be recognized as a valid level of safety, or if we are dealing with equipment with a two-wire AC cord (no Earth wire), then, besides the layer of basic insulation, we need an additional level of protection between exposed metal (e.g., output leads) and high voltage (AC). This could be another insulating layer, with identical dielectric withstand capability. It is then “supplementary insulation.” Together, the two layers (basic+supplementary) are said to constitute “double insulation.” We could also use a single layer of insulation, with dielectric withstand properties equivalent to double insulation (i.e., 3,000 VAC or 4,242 VDC). That would then be called “reinforced insulation.”

• Why do we connect the enclosure to Earth in the first place, if that is not always acceptable from a safety standpoint? The main reason for earthing a metal enclosure is that we want to prevent radiation from inside the equipment from spilling out. Without a metal enclosure, there is very little chance that a typical off-line switching power supply can comply with radiated-mode (and perhaps even possibly conducted-mode) emission limits. By earthing, the enclosure is held at a fixed potential and therefore forms an excellent shield around the power supply for meeting radiated emission limits. But the metal enclosure is also, rather expectedly, eyed by engineers as an excellent and fortuitous heatsink. So, in practice, power semiconductors are often going to be mounted on the enclosure directly (with appropriate insulation ratings as indicated above). However, by doing this, we also create leakage paths (resistive/capacitive) from the internal subsystems/circuitry to the metal chassis. And even if we ensure that these leakage currents (CM noise currents) are small enough not to constitute a safety hazard, they can present a major EMI headache. We realize that if these leakage currents are not “drained out” in some way, the enclosure will charge up to some unpredictable/indeterminate voltage, and will ultimately start radiating (a dipole, or an electric field source). That would clearly be contrary to the very purpose of using a metal enclosure. So, we really need to connect the enclosure to Earth, other than for safety reasons. We note that even if we didn’t have power devices mounted on the enclosure, there could be other leakage paths present from inside circuitry to the enclosure. And besides that, an unearthed enclosure would also inductively pick up and re-radiate the strong internal electric/magnetic fields.

• To help in diverting CM currents away, capacitors are often connected between various parts of the power supply and the earthed enclosure (or earth terminal). These are called “Y-capacitors” or Y-caps. Similarly, to reduce DM currents, high-voltage capacitors are often placed between L and N of the incoming AC mains. These line-to-line components are called “X-capacitors” or X-caps. All Y-capacitors on the Primary side of an AC–DC power supply must be rated for a certain minimum voltage as discussed later. And, of the Y-caps on the Primary side, those that are before the bridge rectifier (i.e., on the line side) also carry low-frequency AC, and therefore safety agencies in effect regulate their total capacitance value. We will discuss this later.

• The bottom-line is that (a) providing a good metal enclosure and (b) properly connecting it to Earth are the most effective methods of preventing radiated EMI. However, by creating this galvanic connection (to Earth), we also now provide a “freeway” for conducted (CM) noise to flow “merrily” into the wiring of the building. So, in trying to suppress radiated EMI, we might end up increasing the conducted EMI, and vice versa. We thus realize that to be able to stay within the applicable conducted emission limits, we now need to provide a conducted EMI (CM noise) filter somewhere. Such are the reasons why, in Chapter 15, when we first introduced the topic of EMI, we had warned that EMI is a veritable balloon — push it in on one side only to find it bulging out on the other.

• Generally speaking, if the equipment is designed not to have any Earth connection (e.g., a two-wire AC cord), there will usually be no metal enclosure present either. Ignoring the problem of meeting radiation limits for now, the good news here is that no significant CM noise can be created either — simply because CM noise needs an Earth connection by definition. Therefore, a CM filter need not be present in this case. However, we must remember that conducted noise limits include not only CM noise, but DM noise too. So, irrespective of the type of enclosure and earthing scheme, DM filters are always required.

• In plastic enclosures, to comply with radiation limits, the inside of the plastic box may be shielded, say by an insulated wrap-around metal foil, or by depositing metal on its insides.

(a) Vacuum deposition: A metal (e.g., aluminum) is melted in a vacuum chamber and its droplets sputtered onto the inner surface of the enclosure, gradually building up a continuous metallic layer. This process is ideal when very thin coatings are required on relatively detailed moldings. It is less effective if used on an enclosure with poorly fitting joints because the thin metal layer does not provide gap-filling properties, making electrical continuity difficult to maintain. Therefore, RF leakage around joints and seams is a real problem. Tooling costs are also high because only very precise masks and fixtures will ensure that metal is applied to only the inner surfaces of the enclosure.

(b) Loaded paints: This is the most cost-effective method. A silicone rubber shields the external surfaces, and the metallic-loaded paints are applied using a wet spray process using either an automated or a manual setup. A finished coating thickness of 50–75 μm gives good coverage without obliterating fine details of the molding. Copper- or silver-loaded paint is suitable for most applications requiring commercial levels of attenuation.

(c) Zinc arc spray: This results in a relatively thick layer over the molding surface. It is very effective in applications where a high magnetic field is expected, because of the direct relationship between material thickness and signal absorption. However, this process involves elevated temperatures, so it is most suitable for use with polycarbonate enclosures. The thickness of the coating can also obliterate fine details of enclosure molding.

(d) Electroless plating: This produces the best performance, but it is very expensive and not particularly suitable for high-volume applications. Masking is very difficult, and the screening material can easily get deposited on both the internal and the external surfaces. Further, a secondary finishing operation is required to remove the excess material and perhaps to apply a final painted-over finish. But since colors are intrinsic to the materials used in plastic enclosures, the secondary paint operation is a major disadvantage. All this gets complicated and costly. Therefore, this method is justified only in applications where high attenuation levels are the overriding concern.

• Both the CM and DM stages are symmetrical (balanced). From the viewpoint of the noise emerging from the bridge rectifier and flowing towards the LISN, there are in effect two LC filters in cascade (both for DM and CM noise). This filter configuration can provide good high-frequency attenuation (roll-off). We know that any LC filter provides 40-dB/decade attenuation, so two in series can, in principle, provide up to 80 dB/decade.

• Typical practical values for the inductance of a CM choke in medium-power converters range from 10 mH to 50 mH (per leg). The large values result because we are limited by safety considerations in the amount of total Y-capacitance we can use. On the other hand, there being no similar concern for DM noise, the DM choke is always much smaller (in inductance, but not necessarily in size as we will see). Typical values for the DM choke are 500 μH to 1 mH.

• The leakage inductance of the CM choke L_lk acts as a DM choke, albeit one with a rather small inductance. The leakage inductance of the CM choke is roughly 1–3% of L_cm, depending on its construction. That can serve as an unintentional, but effective DM choke.

• However, since the capacitance associated with a DM filter can be very large, and is not restricted by safety concerns, low-power flybacks (up to 70–80 W) rarely even use a discrete DM choke. Instead they usually have two CM chokes in cascade, relying on the leakage of each to provide DM noise filtering.

• If the chosen CM choke is toroidal in shape rather than a U-core or E-core, it will have almost no leakage inductance. In that case, we would most likely need to provide discrete DM chokes to create a DM filter.

• In Figure 16.1, we have shown both the CM and the DM filter stages as being symmetrical (balanced). For example, we have placed identical DM chokes on each of the L and N lines. However, the DM choke is also a part of the CM equivalent circuit as we can see from the CM equivalent circuit. As mentioned in Chapter 15, since line impedance imbalance can cause CM noise to get converted into DM noise, it is generally advised to keep both the CM and DM stages symmetrical (balanced).

• Occasionally, unbalanced DM filters provide acceptable overall EMI scans — for example, a single DM choke placed on one line only. It is worth trying out for sure. Sometimes, in very low-power applications, or in the case of DC–DC converters/bricks, a plain decoupling capacitor (e.g., C1) may suffice too. Tuned filter stages are also sometimes seen in commercial off-line power supplies (e.g., from Weir Lambda, UK). But there are some anecdotal industry experiences that suggest that under severe line transients or under input surge waveforms, as those typically used for immunity testing, tuned filters can display unexpected oscillations (resonances), ultimately provoking failure of the power supply itself. Therefore, tuned filters are usually avoided in commercial designs.

• One obvious way to maintain equal CM inductances in both lines is to wind them on the same core (e.g., a toroid). That automatically assures a good inductance match (assuming of course that there are an equal number of windings per leg). Note that if we are winding the CM choke ourselves (as during prototyping), we must note the relative direction of the windings, as indicated in Figure 16.1 (see the embedded third and fourth sample CM choke pictures from the left). With such a winding arrangement, the magnetic field inside the core will cancel out completely (in principle) for DM noise. For the same reason, the flux due to the operating AC line current will also cancel out (that too being differential in nature). Therefore, the CM choke will be basically “visible” only to the CM noise component. This choke does not need to be “large” based on DC-bias considerations, but in reality, is in fact quite bulky — due to the high inductance required, and because of the associated large number of turns (all rated for the AC operating current).

• While winding toroidal CM chokes, we need to keep other safety requirements in mind too. We cannot simply wind the two windings carelessly overlapping each other — we need to maintain a specified physical separation between the windings of each AC phase. Nor can we just use a bare ferrite toroid to wind them on because ferrite can be a rather good electrical conductor. Neither should we rely on the enamel coating of a typical copper magnet wire as it is typically not a safety-approved coating. What we really need is a safety-approved coating for the toroid and/or a suitable bobbin.

Note: The reader is cautioned that there are several widely used but confusing symbols for the CM choke existing in schematics found in related literature. Whatever the symbol, as long as it is meant to serve as a CM choke, the direction of the windings must be as in the embedded toroid pictures in Figure 16.1. The correct winding polarity is also indicated by dots in the same figure.

• We should also keep in mind that in theory, the AC operating current flux cancels out in the CM choke. In reality, on account of small imbalances in symmetry of the windings the choke could “topple over” (i.e., progressively flux-staircase over toward one side). This would expectedly degrade the EMI performance, but in extreme cases, the core may even saturate. Though that is not a catastrophic event, it still needs to be fixed since the filter is likely no longer EMI-compliant. Sometimes we may need to oversize the CM choke, just to give it a higher-than-zero DC-bias rating. We may also need to provide some air gap. This may be an actual air gap (between split halves), or it may be a distributed gap, as in powdered iron cores. See Chapter 5 for a better understanding of air gaps and why they help smoothen out production tolerances in core materials and magnetic assemblies. But all such steps do lead to an increase in size of the choke.

• The inter-winding capacitance of a choke affects its characteristics significantly at high frequencies. This can be intuitively visualized as an AC path providing an easy detour for noise to flow past the windings rather than through them. To minimize the end-to-end capacitance of a toroidal winding, it is recommended that the winding be single layer. Coming to U-core type of CM chokes, in Figure 16.1, the sample CM choke picture second from the left is better than the one to its left, in terms of minimizing end-to-end capacitance. That is because of the split introduced in each winding section by the special bobbin used. The split also helps increase the leakage inductance (which helps reduce DM noise). CM choke bobbins with multiple splits are also available, at a higher price of course.

• If we reverse the current (or winding) direction in one of the windings of a CM choke, then it becomes a DM choke (for both lines). However, now it is also subject to the flux produced by the AC line input current (no flux cancellation occurs). Therefore, DM chokes, in general, should always be put through a “core-saturation check.” We may see that DM chokes may need to be quite large, just to avoid core saturation — despite the fact that their inductance is usually much less than that of CM chokes. See Chapter 14 for the peak currents in the input lines. That should help ensure the DM choke is really effective.

• We can consider spending some more money and avail of magnetic materials like “amorphous” cores or “Kool Mu^®” if we want to achieve higher inductance (with higher saturation flux densities) in a smaller size.

• Note that the EMI filter stage is usually placed before the input bridge (i.e., toward the incoming AC line input) — because in that position it also suppresses any noise originating from the bridge diodes. Diodes are known to produce a significant amount of medium- to high-frequency noise, especially at the moment they are just turning OFF. Small RC snubbers (or sometimes just a “C”) are therefore often placed across each diode of the input bridge. Though sometimes, we can get away simply by choosing diodes with softer recovery characteristics.

• Note that input rectifier bridges using ultrafast diodes are often touted as offering a significant reduction in EMI. Opinions about them remain mixed. Some people claim that with these ultrafast diode packs, it is possible to remove the small ceramic capacitors often placed across the four diodes of a typical bridge rectifier. However, in actual tests conducted by the author, the ultrafast diodes didn’t seem to make much difference. If anything, the conducted EMI spectrum actually worsened. We know that very fast diodes can have very “snappy” characteristics too, producing rather sharp spikes of reverse current and forward voltage at turn-off and turn-on, respectively. And that could explain the rather poor results with them.

• Line-to-line capacitors (the X-capacitors or “X-caps”), placed before the input bridge, must be safety approved. After the bridge (i.e., on the rectified side), it’s basically a “don’t care” situation from the safety point of view. At that position, we can use any suitably rated cap really. Note that since an approved X-cap is essentially a front-end component, it can see rather huge voltage spikes coming in from the AC mains. For example, the voltage spikes could be from motorized equipment connected to the same wiring, turning ON and OFF repeatedly. That is why an approved X-cap, even though rated “250 VAC,” is actually 100% impulse-tested up to 2.5-kV peak.

• Line-to-earth capacitors, or “Y-capacitors,” as discussed previously, used anywhere on the Primary side (in an off-line application) must always be safety approved. Any Y-cap placed as a bridging component from anywhere on the Primary side to anywhere on the Secondary side must be safety approved. Since Y-caps are critical in terms of having the potential to cause electrocution if they fail, approved Y-caps are typically impulse-tested up to 5-kV peak. They should also be guaranteed to fail open, not shorted. Depending on the equipment’s expected region of operation, some safety agencies (e.g., in Scandinavian regions) may require us to assume no earthing is present, and therefore we will need to create two levels of protection, by placing two Y-caps in series (basically corresponding to double insulation).

• Note that Y-caps placed between the Secondary side and the Earth/enclosure are often just standard (non-approved) 0.1-μF/50-VDC ceramic capacitors. Sometimes, the low-voltage (<60-V) outputs can even be directly connected to the enclosure. Safety agencies don’t care. However, if the power supply is intended to provide the 48-V rail in PoE (Power over Ethernet) applications, by IEEE requirements, we need to provide 1,500-VRMS (~2,500-VDC) functional isolation between Earth/enclosure and the 48-V output rail plus its return. The reason is that severe voltage spikes may be picked up by the cables carrying data and power across the building. And if we connect those directly to the enclosure, we can cause the user an unpleasant (though not fatal) shock. So, in off-line power supplies meant for telecom applications in general, we cannot use low-voltage Y-caps on the outputs anymore. To support PoE, we therefore use either approved or nonapproved 2.5-kV-rated Y-capacitors on the outputs.

Note: What were traditionally called X- and Y-capacitors are more accurately “X2” and “Y2” capacitors respectively. Historically, X2 and Y2 caps were meant for single-phase equipment, whereas X1 and Y1 caps were meant for three-phase equipment. From the viewpoint of safety regulations (like impulse voltage rating and so on), X1 and Y1 caps are considered equivalent to two X2 and Y2 capacitors in series respectively. Therefore, for example, Y1 caps are typically 100% impulse-tested to 8 kV (Y2 caps to 5 kV). We could use Y1 caps for meeting Scandinavian safety deviations.

• Traditionally, off-line X-caps were of special metallized film+paper construction, whereas Y-caps were a specially constructed disc ceramic type. However, we can also find X-caps that are ceramic, as we can find Y-caps which are film type. It’s a choice dictated by cost, performance, and stability concerns. Film capacitors are known to always provide much better stability over temperature, voltage, time, and so on — than most ceramics. In addition, if they are of “metallized” construction, they also possess self-healing properties. Note that ceramic capacitors, in general, do not have any inherent self-healing property. However, ceramic caps used as approved Y-caps are specifically constructed in such a way that they will not fail shorted under any condition.

• If for any reason (e.g., filter bandwidth or cost), ceramic is preferred for a Y-cap position, then we need to carefully account for its basic tolerance, the variation of capacitance with respect to temperature and applied voltage, and all other long-term variations and drifts. That is because we need a certain filtering efficacy over the life of the product. But at the same time we can’t increase the leakage current into the chassis or we will violate ground leakage current limits. We should therefore keep in mind that the capacitance value stated in the datasheet is not just a nominal (or typical) value, but in fact can be a rather misleading value. For example, the fine print may reveal that the test voltage at which the capacitance is stated is close to, or equal to, zero volts. So, the actual capacitance value the capacitor presents in a real working circuit may be very different from its declared (nominal) value. This is, in general, especially true for ceramic capacitors which use a high dielectric constant (“high-K”) material (e.g., Z5U, Y5V, etc.). We should also know that ceramic capacitors age slowly, except for COG/NP0 types. But COG/NP0 capacitors are very expensive and bulkier too. A typical X7R capacitor ages 1% for every decade of time (in hours). So, its capacitance after 1,000 h will be 1% less than what it was after 100 h, and so on. Higher dielectric constant ceramics like Z5U can age 4–6% for every decade of time. So, in effect our filter stage too gets less effective with time. And we need to account for this upfront, in the initial design.

• Theoretical filter performance is based on the assumption that we are using “ideal” components. However, real-life inductors are always accompanied by some winding resistance (DCR) and some inter-winding capacitance. Similarly, real capacitors have an equivalent series resistance (ESR) and an equivalent series inductance (ESL). At high frequencies, the inductance will start to dominate, and so a capacitor will basically no longer be functioning as one (from the signal point of view). However, capacitors with smaller capacitances generally remain capacitive up to much higher frequencies than do larger capacitances. See Table 16.1 for some typical self-resonant frequencies (the point above which, capacitors start becoming inductive). Therefore, quite often, a smaller Y-cap may help in situations where a larger Y-cap is not yielding the required results. We can also consider paralleling a larger value Y-cap with a small Y-cap.

• Surface mount (“SMD”) versions of off-line safety capacitors are available — for example from Wima in Germany and Syfer in UK. But especially for SMD caps, we must keep in mind that it is not enough for a manufacturer to spontaneously declare their capacitor “complies” with a certain safety standard — the capacitor should actually be approved. Only after being tested by various safety agencies, is it allowed to carry their respective certification marks. From the electrical point of view, one of the great advantages of SMD caps is their very low ESL. This improves their high-frequency performance in any filter application.

• Note that in the relentless drive toward reducing parasitics, we may forget that some ESR or DC winding resistance (“DCR”) is often useful in helping damp out oscillations. Without any resistance to burn up the energy, oscillations will last forever. That is one of the reasons why engineers sometimes pass one or both of the leads of a standard through-hole Y-cap through a small ferrite bead (preferably of a material with lossy characteristics, like Ni–Zn). This can often help suppress a particular high-frequency resonance involving the Y-cap that is perhaps showing up in the conducted EMI scan. But we must be careful that in doing so, we do not land up with a radiation problem instead.

• Designers of low-voltage, low-power DC–DC converters may find the “X2Y” patented product range available from Syfer (and from the company X2Y itself) useful if they need to miniaturize and lower the component count. This is a three-terminal integrated SMD capacitor-based EMI filter that simultaneously provides line-to-line decoupling and also line-to-ground decoupling. From X2Y’s website quoted verbatim: “The patented X2Y^® Technology consists of proprietary electrode arrangements that are embedded in passive components. Components with X2Y technology can be manufactured in a variety of dielectric materials including ceramic, metal oxide varistor (MOV), and ferrite. The main embodiment of X2Y Technology is in a multilayer ceramic capacitor, this allows capacitor manufacturers to license X2Y. End-users can then purchase X2Y^® components just as they would any other passive component. Currently, there are six licensed manufacturers who make and sell X2Y components.”

• Picor (a subsidiary of Vicor) sells active input EMI filter stages for standard 48-V bricks and off-line power supplies. These are expensive (typically ~$50 per unit), but may be a viable choice if board space is at a premium.

• We note that a Y-cap is always tested to higher safety standards than an X-cap. So, we can always use a Y-cap at an X-cap position, but not vice versa. For example, we can consider placing a ceramic Y-cap in parallel with a film X-cap to improve the DM filter bandwidth. However, nowadays, to keep things simple, manufacturers are selling safety approved caps designated for use either as an X-cap or a Y-cap.

• We often run into the problem of needing to improve filter performance in the low-frequency region of the CISPR-22 Class B limits. We do this by trying to increase the “LC” product as much as practically possible (in the process, lowering the resonant frequency). Given a choice, we would prefer to realize the large target value of the LC product, by using larger capacitances instead of impractically sized inductors. But as we know, the total Y-capacitance (before the bridge rectifier) is limited by safety considerations. Note that X-caps too seemed to be limited for many years to a maximum value of 0.22 μF, or occasionally 0.47 μF. But that was actually based on availability and component technology limitations. Nowadays, we can get approved X-caps up to 10 μF. We should, however, be conscious that large input capacitances can cause undesirably high inrush surge currents at power-up. This may also cause eventual failure of the X-cap, especially if it is the very first component after the AC input inlet. The X-cap may not fail outright since we know that film caps can self-heal from such events. However, eventually, the capacitance will degrade over time. Therefore, despite EMI concerns, we should try to place X-caps after any input surge protection element, for example, after the NTC (negative temperature coefficient) thermistor, or a wirewound resistor, perhaps even after a front-end choke.

CM filter: We see that the discrete DM choke L_dm acts as a CM filter element too. The CM filter stage consists of two LC stages. One stage has a large LC product (low-frequency pole) corresponding to the product 2×L_cm×C4. The other provides high-frequency filtering since its LC product is relatively small: L_dm×C2.

DM filter: We see that the CM capacitor (C4) acts as a DM filter element too. The DM filter stage consists of two LC stages. One stage has a large LC product (low-frequency pole) corresponding to the product 2×L_dm×C3. The other provides high-frequency filtering since its LC product is small: L_lk. Note that if we use a toroidal CM choke, this high-frequency DM filter stage may barely exist since L_lk will be almost zero.

(a) As the frequency increases, the LC filter becomes more and more effective, at the rate of 40 dB/decade. So, we get an additional advantage of 40 dB/decade over what the limits allow.

(b) The LISN sensitivity actually increases with frequency (see Figure 16.6). That will give higher readings as the frequency increases. It is, in effect, a disadvantage of 3.5–10 dB/decade depending on frequency.