High Fidelity, Low Power: A New Audio Amplifier from Analog Devices

You can simplify and improve the performance of your audio circuits with amplifier ICs such as the SSM6322.

I think we are all well aware that most audio signal chains these days are not purely analog. There is no doubt that sound begins as an analog phenomenon, and at some point it must return to the analog world if the objective is to produce something that is compatible with the human ear. But the intermediate steps—storage, filtering, analysis—can take place in the digital realm.

A fundamental component in these mixed-signal audio circuits is the DAC. At some point the ones and zeros need to be converted into an analog voltage, and typically this analog voltage will need some amplification, because DACs are not designed for high output current. This brings us to one of the beneficial characteristics of the SSM6322: it is designed to interface directly with DACs.

Features

Diagram taken from the datasheet.

As you can see, the SSM6322 is a two-channel device, with each channel consisting primarily of an input amplifier and output amplifier. The input amplifier is differential, which makes sense because higher-quality DACs generally employ differential output stages. The output of each channel is a single-ended audio signal that can drive one of the two speakers in a pair of headphones.

ADI thinks that the SSM6322 is notable for its excellent audio performance combined with low power dissipation. The datasheet mentions a specialized amplifier configuration—a difference amplifier combined with a “common-mode loop”—that reduces power consumption by eliminating two amplifiers that would otherwise be necessary. The device also has two shutdown modes for further improvements in battery life.

The aforementioned common-mode loop is one of the characteristics that qualifies the SSM6322 for its status as a particularly “DAC-friendly” audio amplifier.

Diagram taken from the datasheet.

The section enclosed in the dotted line corresponds to the common-mode control circuit. The feedback action of that circuit results in a VP2/VN2 input that is a “DC signal set by the voltage at the REF2 pin” (page 18). This fixed common-mode input level actually helps the DAC to maintain optimal performance, and the amplifier also exhibits its best performance under these conditions.
If you’ve worked with audio circuitry, you know that “click-and-pop” can be an annoying problem. If indeed you have dealt with this issue, you’ll be glad to know that the SSM6322 incorporates click-and-pop suppression circuitry. I can’t vouch for its efficacy, but it’s there.

One other interesting feature that I’d like to mention is the SSM6322’s support for speaker-drive audio signals originating from another component. This is related to the shutdown modes mentioned above. When the device is in low-power mode, the output driver is placed in a high-impedance state that allows other devices to safely drive the same signal.

Capacitive Loads

If you’ve read this article, you know that capacitance on an amplifier’s output node can cause stability problems. Even if the capacitance doesn’t drive the amplifier into full-blown oscillation, it can still affect performance by creating a gain peak in the frequency response.
The following plot conveys the SSM6322’s ability to deal with capacitive loads.

Plot taken from the datasheet.

This doesn’t strike me as catastrophic, especially considering that the peaking occurs way outside the audio band. In other words, all of the frequencies that are relevant to audio playback will experience the same gain (namely, 0 dB).

The above plot includes load capacitance up to 200 pF. If you look at the individual curves, you can see that the general trend is more capacitance equals more peaking. From these two facts I draw the conclusion that you could start to run into trouble as load capacitance goes significantly above 200 pF. Hence, ADI provides a circuit that you can use if you need to drive a high-capacitance load.

Diagram taken from the datasheet.

Note the series resistor; this is a standard method of dealing with high load capacitance. This resistor modifies the frequency response such that stability is actually worse at the lower load capacitances: you can see that there is still significant peaking at 330 pF, but the response is flat at 2.2 nF.

Diagram taken from the datasheet.

---

Do you have any experience with common-mode control circuity? If you can offer any insights into the details of how this technique works, feel free to leave a comment.