With all this in mind, the phantom powered electret microphone circuit in Figure 20.2 makes a good starting point for the development of a homemade condenser mic suitable for use in recording interesting sounds with an individual character and a unique twist. As with all the major circuits throughout this book, possible breadboard and stripboard layouts are provided in Appendix D, but as always, have a go at your own layouts before resorting to the suggestions provided there. In this instance it might be useful to aim for a long thin stripboard layout so as to fit the finished circuit inside a slender enclosure, to make a good looking microphone.

Figure 20.3 provides the necessary details for the transistor used in this circuit. Others can be substituted following the guidelines given in Chapter 17. As always, if an alternative transistor is to be employed, be sure to check the pinout for the particular part in question. Most crucially, notice here that the device used in this particular circuit is of the less common PNP type (with the arrow on the emitter pointing in, and the emitter itself operating at a positive potential relative to the collector).

Phantom powering is a commonly encountered method of providing power to an audio circuit such as a microphone or DI box. Notice in Figure 20.2 that no power supply connections are indicated, and the only external connections to the circuit are the three wires of a standard balanced audio connection (most typically implemented in the form of an XLR connector). In phantom powering the power is delivered to the circuit over the same wires across which the audio emerges (Figure 20.4). Pins two and three both present +48V delivered through 6k8Ω, with pin one supplying the common ground path necessary in order to complete the circuit and allow current to flow. A phantom power supply can only deliver a fairly modest amount of current into the circuit being powered (limited by the 6k8 resistors in the supply), so care must be taken in the design of such circuits.

In drafting the schematic layout in Figure 20.2 the clarity of the circuit was better served by a neater layout, at the expense of the ‘positive at the top, negative at the bottom’ schematic diagram convention. Thus the ground connection, representing the most negative point in this circuit, is located in the middle of the diagram. This point attaches to pin one of the XLR.

The common connection point running along the bottom of the diagram is connected to this ground point through a 12V Zener diode (D1). This holds the collectors on the transistors, and all the other components connected here, at a potential of approximately +12V. The two emitters see the phantom power supply directly through the 6k8 resistors at the far end of the XLR. They are held at 48V minus whatever is dropped across these resistors. The microphone capsule itself drives the two bases through C1 and C2, providing the two anti-phase signals which ultimately drive the hot and cold outputs.

Often a capacitor will be placed between the amplifier output and the loudspeaker in a circuit design. The fundamental requirement which such a capacitor is intended to meet is that there be no DC component to the signal present across the loudspeaker’s terminals. This is important for two reasons. Firstly, a DC component will result in the loudspeaker cone adopting an offset resting position, resulting in elevated distortion in the audio it produces. Secondly, a DC voltage will result in a corresponding DC current constantly flowing through the loudspeaker coil. This will result in an excessive heating effect, especially when no AC signal is present to keeping the coil moving back and forth, cooling it down. The inevitable result is a burned out speaker coil.

This observation leads to a perhaps unexpected conclusion for the matching of amplifiers and loudspeakers. It is obvious that a low power handling speaker should not be connected to a more powerful amplifier, as the amp is likely to blow the speaker when turned up. It may not be so obvious that the opposite arrangement – low power amp and high power speaker – also has the potential to blow the speaker. The amp, being lower power, is more likely to be overdriven and start clipping badly. Clipping sends repeated short bursts of high current through the loudspeaker coil, without any heat dissipating movement accompanying it, and so burnout can once again ensue.

When considering whether or not an output capacitor is required between amp and speaker, the quiescent voltage level at the output of the amp needs to be compared to the voltage level at the other speaker terminal. If they are the same, no capacitor is needed. Consider the two circuit configurations illustrated in Figure 20.7. In each case the minus terminal of the loudspeaker is connected to ground. Thus, when no signal is present at the input of the amplifier, the plus terminal of the loudspeaker must also sit at ground potential.

An audio amplifier needs to be able to amplify the positive and negative portions of an audio signal equally well. It can only produce outputs in the range of voltages available from its power supply, and therefore with no input signal the output must sit at a voltage half way between its positive and negative power rails. In Figure 20.7a a single rail supply is utilised. The midpoint in this case is +6V, and so this is where the output will sit. With the minus terminal of the speaker connected to ground (0V), clearly a DC blocking capacitor is needed between the amp and the plus terminal of the speaker. On the other hand, in Figure 20.7b a split supply is shown powering the amp. With its power rails at +12V and −12V, its output will sit at 0V, thus matching the ground potential on the minus speaker terminal, and so no capacitor is required.

Notice that the capacitor indicated in Figure 20.7a is shown as a polar type. The capacitor will act as a high pass filter, and since it is attached to a relatively low impedance in the form of the loudspeaker, it needs to be quite large in order not to block too much low frequency content from the audio signal. Large capacitors are most often polar, which is why this device is marked as such. The cutoff frequency of this high pass filter can be calculated using Eq. 13.13. A smaller value can be used to restrict lower frequencies, which might be desired when driving a small, low power handling driver such as the little half watt unit illustrated in Figure 20.5.

There are two situations in particular in which the coupling capacitors discussed above are not needed: split rail and bridge mode operation. The split rail case has been addressed. Figure 20.8 illustrates the second situation, that of a bridge mode amplifier configuration. The requirement, as previously stated, is that no DC component exists across the terminals of the loudspeaker. The bridge mode configuration is unusual in that the minus terminal of the loudspeaker is not connected to ground. Instead the speaker floats around the quiescent output level of the amplifiers attached to the two terminals. Clearly these two amps must have matched outputs, so as to maintain the required equal quiescent levels, and avoid any unwanted DC current flow.

Figure 20.8a illustrates the basic bridge mode configuration: two differential amplifiers with the same signal fed to the noninverting input of one and the inverting input of the other. The other inputs are grounded, and the speaker is bridged across the two outputs (hence the name). In this case the amplifier powering scheme can be single or split rail, without the need for coupling capacitors in either case.

Figure 20.8b shows a simple concrete implementation of the bridge mode configuration using a pair of 386 low power audio amplifier ICs. Pins two and three are the inputs. The input signal is fed into pin two on IC1 and pin three on IC2, with the other input pins grounded. Pin five is the output on each chip, connected here to either terminal of the loudspeaker.

With absolutely no other components required apart from input, output, and power, this must be a contender for the simplest amp ever (alongside the bare-bones amp in Figure 18.8a, which uses just one 386, but does require an output capacitor in addition). In this very simple circuit pins one, seven, and eight are left unconnected. Obviously there are many enhancements which could be applied to this basic design, but it does operate quite effectively even as shown.

One further point must be given careful attention when examining a bridge mode configuration, that of amplifier loading. As previously stated, an amplifier will have a minimum load impedance, perhaps 4Ω. In the bridged configuration it is important to note that each of the two amplifiers involved sees only half of the load connected between them. Recall that load impedance is measured to ground, and note that there is a virtual earth point half way along the speaker’s voice coil, with the terminal voltages seesawing up and down either side of it. Therefore, the speaker (or combination of speakers) connected across the amplifier outputs must have an impedance of at least twice the amplifiers’ minimum load impedance.

One other circuit element which is very often seen in conjunction with a standard moving coil loudspeaker is shown in Figure 20.9. This is what is usually referred to as a zobel network. In simple circuits, the network usually takes the form of a small resistor (usually 10Ω) and a capacitor (often around 47nF) in series. A lot of work and a lot of maths can go into modelling and designing ideal solutions, but the simple approach serves well for all but the most demanding applications.

Electrically, a moving coil loudspeaker can be thought of as a resistor and an inductor in series. It is made from a coil of wire, so this should come as no surprise. Inductors have a frequency dependant impedance (Eq. 13.2). As such, if an amplifier has just a loudspeaker attached to its output, it will see a load which varies widely with frequency. A zobel network has an impedance which varies in the opposite direction – the speaker has inductance while the zobel network has capacitance.

The commonest moving coil loudspeakers have a quotes resistance of 8Ω, so the 10Ω resistor in the network matches this closely enough. The net result of the speaker and zobel in parallel is a load with a much more consistent impedance across the operating frequency range of the amplifier. This steadier loading tend to make an amplifier behave better.

As mentioned above, loudspeakers have as one of their primary figures of merit, a characteristic impedance, the commonest being 8Ω, and likewise amplifiers have a minimum load impedance. It is a common situation to find multiple loudspeaker drivers wired together and connected to a single amplifier output (think of a guitar amp driving a two or four speaker cab). In such a case it is important to know how to determine the total impedance which a particular speaker combination will represent.

In actual fact it is very straightforward in most cases. The familiar resistor series and parallel rules are all that is required. Indeed, for almost all wiring configurations likely to be encountered the situation is even easier. It is usually the case (and usually the only advisable course) that only identical drivers are wired in combination like this (LF/HF combinations involving a crossover are a different matter, and are not covered here.)

Revising the series and parallel rules from Chapter 12, it quickly become apparent that two equal resistances in series result in a total resistance double the size, while two equal resistances in parallel results in a total half the size. So for instance, two 8 drivers in series looks like 16Ω, and two 8Ω drivers in parallel looks like 4Ω, see Figure 20.10.

The section on microphones earlier in this chapter focused on the commonly available electret condenser microphone capsule. Condenser microphones represent one of the two key microphone types found in audio work. The second type is the dynamic microphone. In many ways the dynamic mic is a more straightforward device than the condenser. Fundamentally it shares exactly the same working parts as the moving coil loudspeaker – in fact, so much so that it turns out to be a simple matter to use a moving coil loudspeaker as a microphone.

The key differences between the two are size and mass. Generally the detection element of a microphone (the diaphragm) is small and light so that sound can move it around easily. The equivalent part of a loudspeaker (the cone) tends to be much more bulky, so it can move a lot of air. So for a speaker to work effectively as a mic, it should be relatively light and the sound it is to be used to detect should be relatively energetic (i.e. loud and low frequency). Spend any time in a recording studio and you are bound to come across the NS-10 kick drum microphone or its commercialised variants. Basically its just a mid sized driver used as a microphone on loud, low frequency sources.

To make a humbucker requires two single coil pickups whose magnets are mounted in opposite orientations. (In fact humbuckers often use a single magnet installed so as to direct its two poles one at each of the pickup windings.) The two pickups in a humbucker are placed side by side but are wired together out of phase, as in Figure 20.14. Each pickup generate both the wanted and unwanted signals described previously. The out-of-phase wiring means that the unwanted signal (which is not affected by the magnet orientations) cancels between the two pickups. The wanted signals however get flipped twice, once by the out-of-phase wiring and again by the reverse oriented magnets. This means that the wanted signals from the two pickups end up back in the same polarity and thus adds, actually giving a stronger signal.

With two binary options (wiring phase and magnet orientation), there are four possible ways in which two single coil pickups can be wired together, and four corresponding types of behaviour. These four possible variations are laid out in Table 20.1. While many factors combine to make up the final tone emanating from an electric guitar, these basic wiring options can certainly prove a useful tool in tailoring the sound of an instrument.

Other considerations not discussed here include: series versus parallel pickup wiring, choice of magnet type (there are several), pickup positioning along the guitar body, and proximity of pickup to strings. Some pickups even have per-string magnet pole-pieces which can be individually adjusted to modify their response. And then there are many non pickup related factors, from the strings, to the guitar body, to the amp, all of which play a role – interesting topics but not for the current discussion.

To round off this section on coil pickups, a simple electric guitar wiring diagram is presented (Figure 20.15). The two pickups indicated may be single coil variants, or they may represent complete humbucker pickups, or indeed other nonstandard wiring pickup pairs. If single pickups are assumed then it is certainly worth checking the magnet orientation of the pair. If they are opposite (or if one can easily be converted, as discussed above), then it might be advantageous to wire them out of phase to form a humbucker pair, even if they are not mounted right beside each other. The dot at the top of each coil indicates that these two are wired in phase. Swapping the two connections to one of the coils is all that is needed to place the two coils out of phase.

Pickup selection switches can be far more elaborate than the one shown here. In this case the three switch positions simply select one, the other, or both pickups in parallel. The symbol indicates an dpdt on-on-on device, and ‘NC’ stands for ‘no connection’. A standard dp3t switch could also be used here, with jumper wires between the appropriate two pairs of throws.

The tone stack is a fairly standard arrangement for an electric guitar. With the pot turned up, the large resistance means little signal goes down that way, and the signal from the pickups passes un-affected. As the pot is turned down, the capacitor continues to block low frequencies from shunting to ground, but high frequencies can make their way through. Thus the tone control works by bleeding off more or less of the high frequencies from the incoming signal, depending on the position of the pot control.

Finally, the volume pot is absolutely standard, simply tapping off its output signal somewhere between full and zero, depending on the position of the pot knob, and the guitar output is provided via a standard TS jack.

Figure 20.16b shown a less commonly encountered variant on the basic piezo element. In this type the top silvered contact area is separated into two unconnected zones – the main drive contact, and a smaller feedback region used to sense when the element has been deformed. These three terminal types are called self drive piezos, and are most commonly used to implement very simple oscillator sounders, as examined next in Learning by Doing 20.6. Regular piezos can also be used to reproduce a full audio signal (typically at a low fairly quality). Refer back to Learning by Doing 14.2 (p. 255) for an example of how to drive a piezo sounder with an audio signal.