Clipping is the state of an amplifier when, being fed by a sine-wave input signal, it produces the output sinusoid with cut peaks of its upper and lower halves and whose form ultimately can be turned to square-wave. There are many circuits which register the clipping onset, they differ by their complexity and the principle of operation. The commonly used so-called clipping indicating circuits are rather amplifier output level monitors and the LED indication takes place here when this voltage level is expected to produce the amplifier clipping, for given power supply rails.
     The described below clipping indicator was specially designed for use in my low-voltage power amplifier running from a single 14V supply voltage and featuring very low distortion (0,003%) on a 4-Ohm load until reaching a 15W output power. The produced then clipped output voltage is soft, clean and strongly symmetrical in the whole audio range (see Fig.1), therefore subjectively the amplifier sounding doesn't seem impaired and unpleasant even in condition of deep clipping.

Fig.1. Low-voltage power amplifier clipping.

     This condition is a normal mode of this amplifier operation, the amplifier doesn't suffer overheating and in general proves to be extremely enduring. The clipping monitoring is necessary here for output power adjustment to be sure objectively that produced distortion is within the lowest limits.
     For high-power amplifiers (>100W) the use of a clipping indicator may be also actual and for safety reasons, particularly if their current protection isn't quite reliable. Another extreme case is when the protection is too anxious and can trigger the indicator before the output natural clipping occurs, forcedly changing this output. All depends on how the protection is adjusted, but it shouldn't mislead regarding the cause of registered clipping.
     The offered circuit compares the amplifier input and the reduced proportionally to its gain output, the difference in their shape (more than 2% output distortion) turning on the LED indication of constant 0,3sec duration even if the clipping occurs in a single period of the highest audio frequency. Circuit modifications for the inverting and non-inverting power amplifier configurations are designed and built and I would like to consider them in detail.
     Clipping indicator for the most widespread non-inverting amplifier is depicted in Fig.2. The amplifier elements are presented here too, their values are the same as in my VK-5 power amplifier.

Fig.2. Clipping indicator for the non-inverting power amplifier.

     The amplifier output signal is attenuated (R4, R5) according to the amplifier closed-loop gain which for this circuit is K=(R2+R3)/R3=34. The stage on transistor Q1 inverts the input signal phase and serves as a buffer to prevent the clipping indicator influence on the very sensitive amplifier input circuitry. For power amplifiers with THD<0,005% I even recommend to take the input signal from the emitter of the first differential pair transistor.
     The attenuated and inverted AC signals are summed at the base of transistor Q2 with the help of a network C4, R8, C5, R6, R7. The R7 adjustment allows to achieve an almost full AC signal subtraction at the Q2 collector when the power amplifier operates in a linear mode without reaching its output clipping. At highest audio frequencies a certain small AC signal is left at the collector of Q2 due to the phase difference between the amplifier input and output.
     The DC voltage at the Q2 collector (about +8V) is set slightly higher than the threshold of switching a NAND gate U1B, the U1B output having therefore a low potential, the output of the next NAND gate U1A being at the same time high.
     The amplifier output clipping breaks the established signal balance at the base of transistor Q2 and the AC voltage appearing at its collector changes the state of the U1B gate output just at the moments when the clipping takes place, the produced pulses being inverted by the U1A gate. The rising edge of the U1A output pulse triggers a monostable built on the CD4027 flip-flop (U2B) which then generates a high going output pulse providing via transistor Q4 a 0,3sec LED indication of the occurred clipping.

Fig.3. Transient analysis of the clipping indicator.

     The work of the described above circuit was simulated in the Multisim 10 program, its main processes are illustrated by the graph of Fig.3 where the blue trace shows the amplifier clipped output, green trace – the Q2 collector voltage, aqua trace – the U1A gate output and red trace – the U2B monostable output pulse whose duration is intentionally reduced from 0,3sec to 4,2msec (C7=12nF) to use a single time scale for all the traces. Distortion of the depicted clipped sinusoid is 4,3%, the minimum level of distortion the indicator can register is 2%.
     The chosen for simulation amplifier uses a ±15V power supply, while the clipping indicator circuit can run from a single +(20÷40)V supply voltage available in any power amplifier. Adapting the indicator to a concrete amplifier requires calculation of the resistors’ R4, R5 values to establish the amplifier output attenuation (R4+R5)/R5 exactly equal to the amplifier closed-loop gain K=(R2+R3)/R3.
     The circuit adjustment is carried out with the help of monitoring the state of the Q2 collector - its DC potential of +8V should be set by choosing the value of resistor R17 and a minimum AC signal level at this point can be reached by regulating the potentiometer R7, the procedure being performed in conditions of the amplifier nominal non-clipped 1kHz output.

Fig.4. Clipping indicator for the inverting power amplifier.

     The clipping indicator implementation for the inverting power amplifier is represented in Fig.4. In comparison with the previous circuit it doesn’t contain the input inverter because in the inverting amplifier its input and output are normally in anti-phase. As for the rest, the two circuits are schematically identical, it’s only necessary to bear in mind that for the inverting configuration the amplifier closed-loop gain is K=-R2/R3 and just this formula should be used when calculating the amplifier output attenuator.
                                                                                                                                                            pdf version here