Audio compression is a signal processing operation that reduces the volume of loud sounds or amplifies quiet sounds thus reducing or compressing an audio signal's dynamic range. Compression is commonly used in sound recording and reproduction, broadcasting, live sound reinforcement and in some instrument amplifiers.
Here, we present a simple compressor circuit which is able to compress the dynamic range of any analog audio signal. The specific circuit does not amplify quiet sounds but it just compresses loud sounds.
The simple compressor-limiter is an analog circuit and can be used to process only analog audio. It can be adjusted to change the way it affect sounds through two potentiometers. It uses soft compression for middle-loudness sounds and hard-compression for very loud sounds, thus acting as a limiter for very loud sounds. Basically, when acting as a limiter, the circuit uses high ratio compression and fast attack time.
The simple compressor – limiter, was actually designed to be used in FM radio broadcasting, where fast attack time is crucial because the modulation index of any FM transmitter must be kept below a certain limit, otherwise some interference may occur to neighbor radio channels.
The electronic circuit
The simple compressor – limiter uses TL082, which is a high quality dual operation amplifier IC. One op- amp of the TL082 is used to measure the level of the input signal and the other op-amp is used as a variable-gain amplifier. The measured signal level provides feedback, and above a certain threshold, it reduces the level of the variable – gain amplifier.
IC1-A is used as a typical non-inverting amplifier. Its gain is equal to 1+[R4/(R3+ rDS)], where rDS is the drain – source resistance of the T1 FET. The FET acts as a voltage controlled resistor (VCR) which is controlled by the voltage potential applied to its gate.
IC1-B is a typical inverting amplifier which takes at its input the output signal of IC1-A, and amplifies it furthermore. The gain of the inverting amplifier is defined by the ratio (R10+R11)/R12.
D1 and D2 form a rectifier. Together, with C8 and R7, are used as an RMS detector which produces a negative voltage which is roughly approximate to the audio sound level. The time constant C8*(R7+R16) defines the speed of detection, thus also defines the attack and release times of the compressor.
The negative voltage which is produced from the RMS detector is applied to the gate of T1 and controls its drain – source resistance, which in turn, controls the gain of the variable-gain amplifier. As the sound level increases, the negative voltage produced by the RMS detector increases. As this negative voltage increases, the drain – source resistance of the T1 also increases and lowers the gain of the variable – gain amplifier. This operation reduces the volume of the loud sounds and compresses the dynamic range of the audio signal.
While the time constant C8*(R7+R16) defines the attack and release times of the compressor, the threshold of the limiter is defined by the gain of the non-inverting amplifier which in turn is defined by R11. Thus, R7 can be used to adjust the speed of the compressor and R11 can be used to adjust its threshold.
JFET as a voltage- controlled resistor
For a junction field-effect transistor (JFET) under certain operating conditions, the resistance of the drain-source channel is a function of the gate-source voltage alone and the JFET will behave as an almost pure ohmic resistor. Maximum drain-source current, IDSS, and minimum resistance rDS(on), will exist when the gate-source voltage is equal to zero volts (VGS = 0). If the gate voltage is increased (negatively for n-channel JFETs and positively for p-channel), the resistance will also increase. When the drain current is reduced to a point where the FET is no longer conductive, the maximum resistance is reached. The voltage at this point is referred to as the pinchoff or cutoff voltage and is symbolized by VGS = VGS(off). Thus the device functions as a voltage- controlled resistor.
Figure 2 details typical operating characteristics of an n-channel JFET. Most amplification or switching operations of FETs occur in the constant-current (saturated) region, shown as Region 2. A close inspection of Region 1 (the unsaturated or pre-pinchoff area) reveals that the effective slope indicative of conductance across the channel from drain-to-source is different for each value of gate-source bias voltage. The slope is relatively constant over a range of applied drain voltages, so long as the gate voltage is also constant and the drain voltage is low.
The last observation is an essential detail which is taken into account in our compressor design in order to avoid any signal distortion:
It is obvious (from figure 2) that bias lines bend down as VDS increases in a positive direction toward the pinch-off voltage of the FET. The bending of the bias lines results in a change in rDS, and hence the distortion encountered in VCR circuits.
There are two ways to avoid distortion when using a FET as VCR:
- By keeping VDS at extremely low levels
- By using some negative feedback from drain to gate
We use both of the above techniques in our simple compressor- limiter circuit.
R1 and R2 form a voltage divider which acts as an attenuator. The attenuator reduces the input signal level by 33db, in order to keep the drain - source voltage of T1 within safe limits. The effect of the -33db attenuation of the R1 - R2 network on the output, is canceled by the normal-gain of the non-inverting amplifier (which is about +33db).
C6 and R5 form a negative feedback network which applies a part of drain voltage to the gate. During positive signal cycles, this causes the channel depletion layer to decrease, with a corresponding increase in drain current. Increasing the drain current for a given drain voltage tends to linearize the VGS bias curves. On the negative half-cycle, a small negative voltage is also coupled to the gate to reduce the amount of drain-gate forward bias. This in turn reduces the drain current and linearizes the bias lines. This way, the channel resistance becomes dependent on the dc gate control voltage and not on the drain signal, unless the VDS = VGS – VGS(off) locus is approached.
Typical performance curves
The performance of the simple compressor limiter was measured with the use of a dual channel oscilloscope and a signal generator. Measurements are shown in detail in Figure 3. Different color curves correspond to different threshold settings, as set from R11 potentiometer. The performance was typically the same from 30 Hz to 50 KHz; however the curves shown in the following figure were taken at 1 KHz.
How to build the circuit
For making things easy, we have designed an appropriate printed circuit board. The board has copper on one side, and you may easily etch and drill it. Copper on solder-side is shown on figure 4:
You must place and solder all the components on the PCB, according to the assembly guide of figure 5. After building and testing the circuit, you may place it on an amplifier’s front panel or you may use it as an autonomous device.
The simple compressor – limiter has also an analog output, to be used to drive a small uA – meter or a small voltmeter. The meter connected in that output will provide a visual indication of the actual compression ratio in real time.
Since the circuit is monophonic (supports a single audio channel), you may wish to build two identical circuits, to use them for a stereo system. In that case, it would be normal to use highest quality and low tolerance components to ensure the same performance on both audio channels.
The simple compressor – limiter requires power from 2 voltage sources; 1 positive voltage of +15V and 1 negative voltage of -15V. You may use a simple symmetrical power supply unit. Current supply does not exceed 5 mA.