The state of the art in class D amplification has progressed rapidly in the past few years, most noticeably for lower-power applications that require less than 50 W per channel. Class D is inherently more efficient than the traditional class AB amplifier because the output stages are always on or off, with no intermediate bias stage necessary.
This efficiency advantage has never held widespread appeal for designers, since the disadvantages of higher parts cost, poor audio performance (compared with class AB), and the need for output filtering often outweighed it. However, two major factors recently reversed this trend:
- Market need: Two rapidly growing end-equipment segments benefit from class D amplification in different ways. The first is cell phones, where speakerphone and push-to-talk (PTT) modes both benefit from higher efficiency, providing longer battery life. Also, the growth of LCD flat-panel displays necessitates "cool-running" electronics because the display color contrast suffers under elevated operating temperatures. Again, class D efficiency means less power has to be dissipated in the drive electronics, hence cooler product operation and better looking pictures.
- Available technology: Driven by market need, improvements in class D technology are now available. Specifically, several manufacturers offer cost equality with class AB, improved audio performance on a par with class AB, and some novel output-modulation schemes that ease the electromagnetic-interference (EMI) burden in many applications.
Some of the newer components, derived from older PWM-style (pulse width modulation) architectures, incorporate sophisticated modulation techniques that achieve "filter-less" operation for lower-power systems. Efficiency claims can be verified on the bench, but some designers suspect that products based on these new techniques will be rife with electromagnetic compatibility/RF interference (EMC/RFI) problems. In reality, effective pc-board layouts and short runs of speaker cable can ensure sufficiently low radiated EMI to pass the applicable FCC or CE standards.
In some applications, the physical layout necessitates long speaker cables. RF emissions must be more tightly controlled in these instances because the speaker cables act as antennas. The longer the speaker cable, the lower the frequency at which it acts efficiently as an antenna. Similarly, some applications have requirements for EMI emissions below that of CE/FCC regulations, perhaps to meet automotive specifications, or where interference with other circuitry at lower frequencies must be avoided.
Employing stereo class D amplifiers in a flat-panel TV is an obvious example. With speakers typically arrayed at the outer edges of the device, long speaker cables are hard to avoid. If analog video signals are present, simply meeting FCC or CE RF emissions may not suffice. (Measurement of these limits is specified from 30 MHz upwards.)
Therefore, suppressing the switching fundamental may be necessary to avoid interference effects with the video signal. If you need to use the traditional LC filters that operate well with older PWM amplifiers, you should analyze them to ensure they're effective in suppressing the high-frequency switching transients produced by the latest amplifiers.
PWM-BASED CLASS D AMPS
Traditional class D amplifiers are usually based on the principle of pulse-width modulation. Their outputs can be configured either as single-ended or as a fully differential bridge-tied load (BTL).
Figure 1 shows the output waveforms typical of a BTL, PWM-based, class D amplifier. The fast switching times and nearly rail-to-rail swings make this type of amplifier very efficient. Conversely, the wide output spectrum implied by those same parameters can lead to high-frequency RF emissions and interference. Output filters typically are included to suppress such unwanted effects.
As Figure 1 shows, the mirror-image waveforms assert very little common-mode (CM) signal on the speaker or cables (lower trace), provided the waveforms of the inverting and noninverting output devices are well matched. Note that a 50% duty cycle represents a zero input signal (idle). You then can design a differential low-pass filter that attenuates high-frequency content in the waveforms (due to the rapid switching) but preserves the low frequencies intended for the loudspeaker.
NEW MODULATION TECHNIQUES
Due to the rising interest in class D amplifiers, several manufacturers have introduced modulation schemes that provide independent control of the two halves of the H-bridge. These schemes offer two key advantages:
- For very low audio signals and at idle, virtually no differential switching is present across the load. This improves quiescent-current consumption over PWM designs.
- The minimal-pulse CM switching helps to reduce the turn-on and turn-off transients. The dc idle level (post filtering) on each of the BTL "legs" is close to GND, which minimizes any mismatch from filter components or stray capacitance (that can give rise to audible click-and-pop when the amp is enabled or disabled).
There clearly are some advantages to the new techniques, but the amplifier outputs no longer are mirror images of each other (Fig. 2). The waveforms shown (representing the MAX9704 stereo class D amplifier) have a high level of CM content. Output-filter requirements are, therefore, different from those of an amplifier with the traditional differential and complementary PWM outputs. Compared with PWM, the new modulation scheme includes a high level of CM signal, and any output-filter design should take that into account. A traditional differential filter topology may provide poor results, as the following example shows.
Figure 3a depicts a traditional LC, PWM class D output filter implemented with ideal values. For simplification, the speaker load is represented as a pure 8-½ resistance, and the inductor's dc resistance is assumed to be negligible. Some straightforward Spice simulations can highlight the problem.
Figure 3b shows the response of the Figure 3a filter driven by a differential input signal. Each output node (FILT1, FILT2) is plotted with respect to GND. The values chosen create a second-order slope above 30 kHz and a well-controlled transition. Group delay is flat across the audio band at approximately 4 µs.
Figure 3c shows the same filter output driven with a common-mode signal. Again, each output is plotted with respect to GND. The result (note the shifted Y axis!) is heavily peaked and obviously very underdamped. That's easily understood, considering how the filter appears to a CM signal. Because the simulation provides ideally matched inductors and capacitors, the differential signal across the resistive load is zero. Subsequently, it has no damping effect on the LC components.
L1 interacts with C1 (as does L2 with C3) to provide the peaked response. In the time domain, this condition would indicate heavy overshoot and ringing. C2 contributes zero when driven common-mode, meaning the filter's cutoff frequency (or more accurately in this case, its resonant frequency) is higher than that of the differential case.
It probably isn't a problem if the output spectrum has zero CM energy at that frequency. If the peaking frequency coincides with the class D switching frequency, large voltage-output excursions can appear at the speaker and the cabling.
Further, the MAX9704 in its spread-spectrum mode (SSM) may excite the underdamped filter by producing appreciable noise energy above the audio band. SSM is a pin-selectable option in which the high-frequency switching energy is "whitened" and lowered in amplitude by randomizing the switching period on a cycle-to-cycle basis, also easing EMI compliance in a filter-less design.
One solution is to preserve the basic architecture of Figure 3a, but add damping elements that suppress the highly resonant common modes. Take a look at the addition of two series RC elements from each output node to GND in Figure 4a. If efficiency isn't important, you can simply add resistors to GND. But capacitors C4 and C5 help to minimize excessive power dissipation in R1 and R2.
The values of C4 and C5 impose a tradeoff. They must be large enough to allow R1 and R2 to damp out the peaking, but small enough to minimize the power loss at high audio frequencies (usually up to 20 kHz). This tradeoff is made easier if the CM cutoff frequency is much higher than the differential-mode frequency, a condition implemented by increasing the ratio of C2 to C1 and C3.
By increasing the CM cutoff frequency, C4 and C5 can be made smaller and R1 and R2 larger, minimizing the audio-frequency power loss into R1 and R2. Pushing the CM cutoff frequency too high, however, allows more common-mode on the cables, so you must determine a reasonable limit in the ratio between the differential and CM –3-dB points. For this filter, we've adopted 1:5 for that ratio.
Figure 4b shows the filter of Figure 4a driven differentially, and Figure 4c shows the response when driven commonmode. Note the higher-frequency, CM cutoff in Figure 4c (–3 dB at around 110 kHz, versus 28 kHz for the differential case), with gentle but well-controlled peaking. This cutoff is well above the highest audio frequencies (and below the class D switching-frequency fundamental), so it should be of little consequence.
Some products with low switching frequencies (200 to 300 kHz) won't work well with this approach. For those products, then, you may have to adopt other methods and topologies. The MAX9704 supplies the best results when set for fixed-frequency mode (FFM) operation at 940 kHz (FS1 = low, FS2 = high) (Fig. 5). For this particular part, FFM sets the switching period at a constant value—pin-selectable between three values—to suit a given application.
Figures 6 and 7 show time-domain performance for the Figure 4 filter when driven by the MAX9704 class D amplifier. A resistive 8-½ load was used in both cases. Figure 6 shows the FILT1 and FILT2 nodes overlaid (top traces), and the resulting 1-kHz differential output waveform (lower trace). Noise on the upper traces is the residual of the output switching after filtering (supply voltage is 15 V).
Figure 7 shows a detail on the trace of Figure 6. Note that the ripple, mostly from the 940-kHz switching frequency, appears as common-mode on both channels. Note also the absence of higher harmonics, which shows the effective suppression of EMI frequencies. (Radiated-EMI measurements usually start above 30 MHz.)
The filter designs in this article all assume a resistive load of 8 Ω. Voice-coil inductance causes the impedance of most wide-range moving-coil loudspeakers to rise above 20 kHz or so. That property makes efficient " filterless" operation possible, but you must account for the rising impedance when optimizing the component values for any additional EMI output filtering.
For audio designers, the need for filtering also arises in the lab when attempting to evaluate and characterize class D amplifier performance. Even if the end product can pass EMC tests without filters, the amplifier selection and evaluation pose problems. Many audio analyzers intended for measuring total harmonic distortion plus noise (THD+N) or amplitude response from conventional audio amplifiers can provide false results when driven by a "filterless" class D amplifier.