Paul White demonstrates that you don't need to be a technical wizard to get the better of op-amps and that they can be a very cheap way of solving level matching problems.

The invention of operational amplifiers revolutionised the design of simple pre-amp circuits and phase inverters almost overnight. At a cost of literally pence, one of these little beasts will amplify or change the polarity of a signal for you and all you need are a few additional components and a couple of 9v batteries. The maths is never more complicated than simple multiplication or division and anyone who can solder at least passably can build something useful.

But before going any further, what does one of these mystical beasties look like? I've selected the popular TL071 as an example simply because it is cheap, common and does the job reasonably well. There are also other op-amps with the same physical layout and similar electrical properties that can be substituted instead such as the famous but now dated 741. Figure 1 shows a top view of the IC (integrated circuit) which comprises the TL071 and you will note that the manufacturers have put a notch at one end so we know which way round to put it. Starting at the notch end with pin 1, the numbering proceeds anticlockwise around the device until we end up back at the other side of the notch. The legs of this chip are arranged on a 0.1 inch matrix and may be soldered directly to the other components or preferably to a piece of Veroboard. This latter is a commercially available general purpose printed circuit board consisting of parallel copper strips bonded to a paxolin board. The board and strips are drilled on a 0.1 inch matrix and the end of a suitable drill bit may be used to break the copper tracks at any desired point. Though this isn't ideal for production jobs, it is very handy for one off projects and is easy to use so long as you take care to make the breaks clean and to avoid getting solder between the strips.

Schematically Speaking

You may have noticed that on circuit diagrams there are occasionally little triangles; this is the engineer's way of drawing an operational amplifier. The signal goes in at the wide end and comes out of the pointed end and Figure 1b shows the pin connections for our TL071. Pins two and three are inputs whilst pin six is the output. Pins four and seven carry the two power supplies, in this case batteries, and the other pins are unused. You may notice that the two inputs bear the legends + and - which stand for non-inverting and inverting. I don't want to get bogged down in theory because this article is supposed to be practical rather than academic so lets just say that the way these are connected up determines whether the output of our amplifier will be in or out of phase with the input. The easiest version to look at is a thing called the inverting amplifier.

Inverting Amplifier

Figure 2 shows our op-amp chip with a few components added, but don't panic. This simple circuit can be contrived to give any reasonable gain from zero to a 100 or more and, as its name implies, the output is out of phase with the input. This is suitable as a simple pre-amp and in this application the phase inversion won't cause any problems. If you read Ben Duncan's articles on phase, you'll know that problems only normally arise when a phase shifted signal is added to a non-shifted signal.

There are only three resistors and two of these are the same value: RAΩ. In this configuration, the input impedance is near enough the same as the value of RA, so let's make this 10KΩ as it is a sensible value for line level work and it makes calculations easy.

The gain of the amplifier is set by the ratio of RB to RA. If RB is bigger than RA, then the amplifier has a gain of more than one, and if RB is smaller than RA, then the gain is less than one. So if we want a gain often times, RB must be 10KΩ times 10 or 100KΩ, right? It's so easy that I can see you don't trust me. I'll do it again with my sleeves rolled up.

So what are these other bits for? C1, C2 and C3 are capacitors and they each do a different jobs. Fortunately you don't have to worry much about the values as I'm going to provide general purpose figures that will do the job in most cases.

C1 simply prevents any DC level getting into the input, which would be amplified and appear at the output 100 times bigger. C1 must however pass low audio frequencies without worry so let's choose a large value such as 100μF which will serve for lots of different values of RA and RB. Likewise, C2 does a similar job and this can again be made 100μF for the sake of a quiet life so that leaves us with C3. This little fellow is there to prevent the op-amp from getting over enthusiastic at high frequencies and turning into an oscillator; it effectively reduces the gain at high frequencies. Make it 22pF and all should be well. Before moving on, I must mention that C1 and C2 are what are known as electrolytic capacitors. These have a positive and a negative end which is marked on the case. They should be connected so that the highest DC voltage is on the positive end as they are in our diagram. In fact there would be no real problem if they were connected the wrong way round in this instance as the voltage across them is small. However, in other situations the correct observation of polarity is critical and wrong connection will destroy the capacitor - sometimes spectacularly. Also, electrolytic capacitors have a maximum working voltage so choose ones that are rated at higher than the battery voltage to be on the safe side.

So, already you can build yourself a little pre-amp to make up that missing gain. Remember that a gain of ten is the same as 20dBs when we are talking about signal voltages. You could even replace RB by a variable resistor or pot and have yourself a fully variable gain pre-amp. When setting the gain though, remember that the output signal can never be greater than the supply voltages and in practice it will be a volt or two less. Any attempt to exceed this will result in a clipped signal which will sound distorted. You can increase the range a little by upping both the battery voltages but don't go higher than 15v or your op-amp may take early retirement.

A benefit of this type of circuit is that the output impedance is fairly low by audio standards so it should drive any following pieces of equipment without any problems.

Non Inverting Amplifier

Figure 3 shows the layout of the non-inverting amplifier and this time you can make C1, C2 and C3 all equal to 100μF. C1 and C2 do the same jobs as before but C3 is there to reduce the low frequency gain of the amplifier in order to prevent DC offsets at the output. As I said, this is practical, not theoretical, so don't let it worry you.

This time the gain is the ratio of RB to RA plus one, the outcome of which is that the gain can never be less than unity; you can't make it zero as in the case of the inverting amp.

What are the other differences? Well it doesn't invert (as is obvious from the title), so it's safe to use in situations where phase might be a sensitive issue but it also has a higher input impedance than the the other configuration. If you make RC 1MΩ, then the input impedance won't be far off that value. It's high enough to take the output from an electric guitar without spoiling the tone, and the output impedance is still low so you can drive loads as low as a few hundred ohms without problem. You could make RA or RB variable to adjust the gain but remember that it won't go to zero, only down to one.

Buffer

And finally a simple circuit that needs no components other than the op-amp (Figure 4a), though a couple of capacitors and a resistor wouldn't go amiss, just to protect it from the indelicacies of the outside world. A practical implementation is shown in Figure 4b. This little circuit is a unity gain buffer which means that the output signal is exactly the same as the input except that the input impedance is high whilst the output impedance is low. This buffer circuit is often used in matching circuits where the input signal on its own would not be beefy enough to drive the next piece of equipment in the chain. Again, it could be used to feed an electric guitar into a mixing desk.

I've kept this article more simple than even I thought possible, so I'm fully expecting irate letters from Ben Duncan and Simon Bateson pointing out all the technicalities I've skipped over. But don't worry, just buy the bits and have a go. The chances are that you'll still get change from a pound and within no time, you'll be building useful bits and pieces in old treacle tins and wondering why you ever found level matching a problem.