hd2.gif - 14Kb

8·2 Op-Amps as Amplifier Stages
The obvious use of Op-Amp ICs is as signal amplifiers. In general terms these are of four main types. Inverting, Non-Inverting, Differential, and Buffers. We have already discussed some of these in previous sections but figure 8·2 shows a comparison of the required circuits, all using the same basic Op-Amp. It is conventional to call the two inputs Inverting and Non-Inverting depending upon the sign of the resulting output and the gain when a signal is fed to the relevant input whilst the other is connected to zero volts (shown as an Earth symbol). In figure 8·1 was shown entering the non-inverting input and entering the inverting input. It is also conventional to label these inputs with a plus sign for the non-inverting input and a minus sign for the inverting input to indicate the sign of the gain that will be applied to the signal entering via that connection.

Fig2.gif - 27Kb

In each case the behaviour of the amplifier is controlled by the feedback from the output to the inverting input (i.e. the input where is shown to be applied in figure 8·1). Given an Op-Amp with a very high open-loop gain we can expect the voltages at the two inputs to always be very similar when the output is a modest voltage. e.g. If the amplifier has an open loop differential voltage gain of (quite a common value) then when the output is, say, 10 Volts, the two inputs will only differ in voltage by 0·1microvolts.

In a circuit like the Inverting arrangement the non-inverting input is connected to Earth (nominally zero volts) directly. Hence the output always adjusts to keep the inverting input to the Op-Amp at almost zero volts (give or take a few microvolts). Hence the input to an inverting input connection sees a resistance ( in these examples) whose other end is connected to a Virtual Earth, so sees an effective input resistance of . Signals presented to a non-inverting input in the other arrangements see an input which the Op-Amp tends to adjust to almost equal the input. Hence the non-inverting arrangement has a very high input resistance. The output impedance of all the arrangements is very low provided we don’t ask for more current than the Op-Amp can supply since the feedback tries always to assert the output voltage required.

8.3 Filters, Tone Controls, and EQ
Although often used to amplify signals, Op-Amps have many other uses. We have already seen in part 3 how an amplifier with a differential input can be used as part of a active filter. In principle, these active filters are just feedback amplifiers with highly frequency-specific feedback networks that manipulate the closed-loop gain as a controlled function of the signal frequency. in addition to the high/low/bandpass (or band reject) filters outlined in part 3 there are a number of other frequency-dependent functions which Op-Amps can be used to perform. Here we can take the example of Tone Controls circuits sometimes used in audio systems.

Although Tone Controls are now rarely provided in domestic Hi-Fi equipment, they still appear in some professional items and can be very useful in improving less-than-perfect source material. Their main task is to adjust the frequency response to obtain a more natural result. However they can also be useful for special purposes such as deliberately manipulating the sound. In scientific areas beyond Hi-Fi some form of adjustment of the frequency response can be very useful in ‘pre whitening’ the spectrum of a signal. This means boosting some frequencies and attenuating others in order to obtain a spectrum which has a more uniform power spectral density. It allows recording or transmission systems to be used more efficiently and provide an optimum signal to noise performance.

Fig3.gif - 18Kb

The classic form of tone control in Hi-Fi is the Baxandall arrangement shown in figure 8·3. This arrangement is called a Baxandall tone control, named after its inventor. We can understand how it works by noticing that it is actually a development of the non-inverting amplifier arrangement shown in figure 8·2. However the normal pair of feedback resistors have been replaced by quite complicated arrangements of resistance and capacitance. The circuit is laid out in a symmetric manner. In this case the impedance between the signal input and the inverting input of the amplifier is and the impedance between the output and the inverting input is . The voltage gain is therefore now


where both and may be complex and have frequency-dependent values. However if we set both potentiometers to their central positions we find that despite being individually frequency dependent we get at all frequencies. Hence when the pots are centered the frequency response is nominally flat and has a gain of . However if we move either potentiometer setting away from its central position we imbalance the system and produce a value of which varies with frequency.

Consider first the effect of the upper pot (the one). At high frequencies the pair of 47 nF capacitors act as a short circuit and clamp the three wires of this 100k pot together. Hence adjusting the 100k pot does not alter the high frequency behaviour of the circuit. However at lower frequencies the impedance of the capacitors rises and the pot has some effect. Hence the 100k pot acts as a Bass Control and allows us to boost or cut the relative gain for low frequency signals. Now consider the lower pot (the one). Here the effect of the associated capacitors is reversed. The 10 nF capacitors mean that the arm of the circuit which contains the 47k pot essentially loses contact with the input and output at low frequencies. Hence the 47k pot has not effect upon low frequency signals, but it does upon high frequencies. It therefore acts as a Treble Control and can be used to boost or cut the relative gain at high frequencies.

Various forms of Tone adjustment can be applied. For example, the Graphic Equaliser circuits which are popular in studios and PA systems use a bank of bandpass filters to break the signal’s frequency range into chunks. Each frequency section is then amplified and the results added (or subtracted) back together with various controlled gains to rebuild the overall signal. By altering the relative gains of the bandpass filters specific tonal bands can be boosted or cut to alter the sound. These complex circuits do reveal one of the main potential problems of Tonal adjustments. Any slight unwanted imbalances mean that it can be almost impossible to get a flat response should it be desired! For this reason, professional or high quality system use close tolerance components and usually have a ‘defeat’ switch that allows the signal to bypass the entire system when tonal adjustments are not required. Given the good quality of signals that are often available these days, tonal adjustment is usually only of value for special purposes or for reducing the severity of problems with historic recordings, or ones made incorrectly. There is therefore something of a ‘theological’ debate in Hi-Fi as to whether people should have, or use, such systems at all. Purists say not. Realists find them useful. As with most engineering this is a case of “Yer pays yer money and yer takes yer choice”!

Content and pages maintained by: Jim Lesurf (jcgl@st-and.ac.uk)
using TechWriter Pro and HTMLEdit on a StrongARM powered RISCOS machine.
University of St. Andrews, St Andrews, Fife KY16 9SS, Scotland.