Audio metering example using LEDs to monitor audio quality

Almost any device that plays or records audio has audio metering capabilities, from cell phones that display a bar graph of audio output level to home stereos with flashing LEDs to live broadcasts.

There are various standards for audio metering, involving the levels being monitored and the time constants used to calculate the average. The behavior of volume unit (VU) meters is defined in ANSI C16.5-1942, British Standard BS 6840, and IEC 60268-17. These standards define 0 VU = +4d Bm with an integration time of 300 ms. Traditionally, a moving pointer meter is used, meaning that from zero signal to 0 VU, the pointer should reach 0 VU in 300 milliseconds (with the same decay). This makes the traditional VU meter great for getting an idea of ​​loudness, but a poor way to monitor transients. This was fine in traditional broadcast and recording systems (primarily tube and tape based), both of which tended to saturate gracefully.

In today's digital systems, clipping inputs or outputs sound terrible, so Peak Program Meters (PPMs) are used. PPMs are similar to VUs, but have much faster integration times (usually around 10 ms). Note that there are multiple standards bodies that define signal levels and precise integration times. Display technology is an important part of these standards. Getting the ballistics right to move the needle 90 degrees (or more) with 90% accuracy is much harder and slower than lighting up a few LEDs!

With LEDs, the quality of audio monitoring changes dramatically. For example, a simple voice recorder product would require a different metering system than a micro system that uses LEDs as visuals. In a live sound or broadcast environment, accuracy and seeing the difference between average loudness and the absolute peak of a signal is extremely important.

Most instruments can be divided into two parts: the front-end circuit and the comparator circuit (Figure 1).

Figure 1. Two-part block diagram including the front end and comparator.

Signal Conditioning

Signal conditioning happens on the front end. There are three things you should do here, plus one optional item:

Audio level gain/attenuation to configure what voltage should equal the highest LED/meter value

Integration time (e.g. 300 ms for VU, <10 ms for PPM, etc.)

Signal Rectification

Optional: Peak hold and direct signal circuit

Signal rectification is primarily needed in comparator type systems. Since not all signals are symmetrical, full signal rectification is required. The sound output of some instruments (and some voices) puts more pressure on the microphone in one direction. For accurate monitoring, the signal should be fully rectified. Make sure to compensate for the typical 0.7 V diode drop.

Figure 2 shows a full-wave rectifier circuit. This circuit can be implemented in a very small space using a BAV99 dual diode and 100 kOhm resistor network in a SOT23 package. For small voltage systems (3.3V), a quad low-voltage rail-to-rail output operational amplifier (op amp) such as the LMV324 can be used in a single-supply system, but the ground should be replaced with Vcc/2. The dual-output op amp used in this example is the LF353 because of its low cost, high impedance input, and wide voltage range support (±15V).

Figure 2. Full-wave rectifier circuit.

FIG. 3 is an example schematic diagram of signal rectification and a VU integrator.

Figure 3. Full-wave rectifier with VU integrator

For systems that need to monitor a peak signal versus the direct signal (and its integration time), a peak hold circuit (or code) can be used. In software, this can be accomplished by storing the maximum peak value of the ADC in a separate register and holding it for a specific number of samples. Whenever a new larger sample is received, simply overwrite the peak value and reset the counter. When the counter eventually reaches zero, reset the peak value to zero.

In analog circuitry, the rectified output can be run through the circuit in Figure 4. D1 biases the output by one diode drop (~0.7V), which is then dropped by 0.7V by D2. D2 ensures that audio content can only flow into C1 and R1, and not back through the feedback circuit. The values ​​of C1 and R1 set the decay time of the peak hold. The impedance of the comparator array ADC also acts as an impedance to ground.

Figure 4. Peak hold circuit example