John Linsley Hood - March 10, 2010
BJTs and MOSFETs are also available in small-signal and larger power versions, whereas FETs and MOSFETs are manufactured in both enhancement-mode and depletion-mode forms. Predictably, this allows the contemporary circuit designer very considerable scope for circuit innovation, by comparison with electronic engineers of the past, for whom there was only a very limited range of vacuum tube devices.
In addition, there is a very wide range of integrated circuits (ICs), which are complete functional modules in some (usually quite small) individual packages. These are designed both for general-purpose use, such as operational amplifiers, and for more specific applications, such as voltage regulator devices, current mirrors, current sources, phase-sensitive rectifiers, and an enormous variety of designs for digital applications, which mostly lie outside the scope of this book.
In the case of discrete devices, I think it is unnecessary for the purposes of audio amplifier design to understand the physical mechanisms by which the devices work, provided that their would-be user has a reasonable grasp of their operating characteristics and limitations and, above all, a knowledge of just what is available.
The fabrication techniques may be based on the use of a completely undoped (intrinsic) slice of silicon, into which carefully controlled quantities of impurities are diffused through an appropriate mask pattern from both sides of the slice. These are described in the manufacturers' literature as double diffused, triple diffused, and so on.
In a later technique, evolved by the Fairchild Instrument Corporation, all the diffusions were made from one side of the slice. These devices were called planar and had, normally, a better HF response and more precisely controlled characteristics than, for example, equivalent double-diffused devices.
In a further, more recent, technique, also due to Fairchild, the silicon slice will have been made to grow a surface layer of uniformly doped silicon on the exposed side (which will usually form the base region of a transistor) and a single diffusion was then made into this doped layer to form the emitter junction. This technique was called epitaxial and led to transistors with superior characteristics, especially at HF. Since this is the least expensive BJT fabrication process, it will normally be used wherever it is practicable, and if no process is specified it may reasonably be supposed to be a planar-epitaxial type.
In contrast to a thermionic valve, which is a voltage-controlled device, the BJT is a current operated one. So while a change in the base voltage will result in a change in the collector current, this has a very nonlinear relationship to the applied base voltage. In comparison to this, the collector current changes with the input current to the base in a relatively linear manner. Unfortunately, this linear relationship between Ic and Ib tends to deteriorate at higher base current levels, as shown in Figure 9.1.
Figure 9.1: BJT nonlinearity.
This relationship between base and collector currents is called the current gain, and for AC operation is given the term hfe, and its nonlinearity is an obvious source of distortion when the device is used as an amplifier. Alternatively, one could regard this lack of linearity as a change of hFE (this term is used to define the DC or LF characteristics of the device) as the base current is changed. A further problem of a similar kind is the change in hfe as a function of signal frequency, as shown in Figure 9.2.
Figure 9.2: Decrease in hfe with frequency.
However, as a current amplifier (which generally implies operation from a high impedance signal source) the behavior of a BJT is vastly more linear than when used as a voltage amplifying stage, for which the input voltage/output current relationships are shown for an NPN silicon transistor as line 'a' in Figure 9.3. (I have included, as line 'b' , for reference, the comparable characteristics for a germanium junction transistor, although this would normally be a PNP device with a negative base voltage, and a negative collector voltage supply line.) By comparison with, say, a triode valve, whose anode current/grid voltage relationships are also shown as line 'c' in Figure 9.3, the BJT is a grossly nonlinear amplifying device, even if some input (positive in the case of an NPN device) DC bias voltage has been chosen so that the transistor operates on a part of the curve away from the nonconducting initial region.
Figure 9.3: Comparative characteristics of valve, germanium, and silicon based BJTs.
Control of Operating Bias
9.2 Control of Operating Bias
There are three basic ways of providing a DC quiescent voltage bias to a BJT, which is shown in Figure 9.4. In the first of these methods, shown in Figure 9.4(a), an arrangement that is fortunately seldom used, the method adopted is simply to connect an input resistor, R1, between the base of the transistor and some suitable voltage source. This voltage can then be adjusted so that the collector current of the transistor is of the right order to place the collector potential near its desired operating voltage.
Figure 9.4: Biasing circuits.
The snag with this scheme is that transistors vary quite a lot from one to another of nominally the same type, so this would require to be set anew for each individual device. Also, if the operating temperature changes, the current gain of the device (which is temperature sensitive) will be altered and, with it, the collector current of Q1 and its working potential.
The arrangement shown in Figure 9.4(b) is somewhat preferable in that a high current gain transistor, or one working at a higher temperature, will pass more current, and this will lower the collector voltage of Q1, which will, in turn, reduce the bias current flowing through R1. However, this also provides NFB and will limit the stage gain to a value somewhat less than R1/Zin.
The method almost invariably used in competently designed circuitry is that shown in Figure 9.4(c), or some equivalent layout. In this, a potential divider (R1,R2) having an output impedance low in relation to the base impedance of Q1 is used to provide a fixed DC base potential. Since the emitter will, by emitter"follower action, sit at a potential, depending on emitter current, which is about 0.6 V below that of the base, the value of R4 will then determine the emitter and collector currents, and the operating conditions so provided will hold good for almost any broadly similar device used in this position. Since the emitter resistor would cause a significant reduction in stage gain, as seen in the equivalent analysis of valve cathode bias systems, it is customary to bypass this resistor with a capacitor, C2, which is chosen to have an impedance low in relation to R4 and R3.
9.3 Stage Gain
The stage gain of a BJT, used as a simple amplifier, can be determined from the relationship:
Vout / Vin = hfeRL / (RS + ri)
where RS is the source resistance, RL is the collector load resistor, hfe is the small-signal (AC) current gain, and ri is the internal emitter-base resistance of the transistor. An alternative and somewhat simpler approach is similar to that used for a pentode valve gain stage in which
Vout / Vin = gmRL
where the gm of a typical modern planar epitaxial silicon transistor will be in the range of 25"40 mS/mA of collector current. Because the gm of the junction transistor is so high, high stage gains can be obtained with a relatively low value of load resistor.
For example, a small-signal transistor with a supply voltage of 15 V, a 4k7 collector load resistor, and a collector current of 2 mA will have a low frequency stage gain, for a relatively low source resistance, of some 300. If some way can be found for increasing the load impedance, without also increasing the voltage drop across the load, very high gains indeed can be achieved - up to 2500 with a junction FET acting as a high impedance constant current load.1
A predictable, but interesting aspect of stage gain is that the higher the gain, which can be obtained from a circuit module, the lower the distortion in this which will be due to the input device. This is so because if increasingly small segments are taken from any curve, they will progressively approach more closely to a straight line in their form. This allows a very low THD figure, much less than 0.01% at 2V rms output, over the frequency range 10 Hz"20 kHz, to be obtained from the simple NPN/PNP feedback pair shown in Figure 9.5 , which would have an open loop gain of several thousand.
Figure 9.5: NPN/PNP feedback pair.
The distortion contributed by Q2 will be relatively low because of the high effective source resistance seen by the Q2 base. A similar low level of distortion is given by the amplifier layout (bipolar transistor with constant current load) described earlier because of the very high stage gain of the amplifying transistor and the consequent utilization of only a very small portion of its Ic/Vb curve.
Basic Junction Transistor Circuit Configurations
9.4 Basic Junction Transistor Circuit Configurations
As in the case of the thermionic valve, there are a number of layouts, in addition to the simple single transistor amplifier shown in Figure 9.4 or the two-stage amplifier of Figure 9.5, that can be used to provide a voltage gain or to perform an impedance transformation function. There is, for example, the grounded base layout of Figure 9.6, which has a very low input impedance, a high output impedance, and a very good HF response.
Figure 9.6: Grounded base stage.
This circuit is far from being only of academic interest in the audio field in that it can provide, for example, a very effective low input impedance amplifier circuit for a moving coil pick-up cartridge. I showed a circuit of this type, dating from about 1980, in an earlier book (Audio Electronics, Newnes, 1995, p. 133).
The cascode layout is also used very widely as a voltage amplifier stage, using a circuit arrangement of the kind shown in Figure 9.7(a). As in the case of the valve amplifier stage, this circuit gives very good input/output isolation and an excellent HF performance due to its freedom from capacitative feedback from output to input. It can also be rearranged, as shown in Figure 9.7(b), so that the input stage acts as an emitter"follower, which gives a very high input impedance.
Figure 9.7: Cascode layouts.
The long-tailed pair layout, shown in its simplest form in Figure 9.8(a), gives a very good input/output isolation; also, because it is of its nature a push"pull layout, it gives a measure of reduction in even-order harmonic distortion. Its principal advantage, and the reason why this layout is normally used, is that it allows, if the tail resistor (R1) is returned to a -ve supply rail, both of the input signal ports to be referenced to the 0-V line - a feature that is enormously valuable in DC amplifying systems.
Figure 9.8: Long-tailed pair layouts.
The designer may sometimes seek to improve the performance of the circuit block by using a high impedance (active) tail in place of a simple resistor, as shown in Figure 9.8(b). This will lessen the likelihood of unwanted signal breakthrough from the "ve supply rail, as well as ensuring a greater degree of dynamic balance, and signal transfer, between the two halves.
Although like all solid-state amplifying systems it is free from the bugbears of hum and noise intrusion from the heater supply of a valve amplifying stage - likely in any valve amplifier where there is a high impedance between cathode and ground - it is less good from the point of view of thermal noise than a similar single stage amplifier, partly because there is an additional device in the signal line and partly because the gain of a long-tailed pair layout will only be half that of a comparable single device gain stage. This arises because if a voltage increment is applied to the base of Q1, then the Q1 emitter will only rise half of that amount due to the constraint from Q2, which will also see, but in opposite phase and halved in size, the same voltage increment. This allows, as in the case of the valve phase splitter, a very close similarity, but in opposite phase, of the output currents at Q1 and Q2 collectors.
9.5 Emitter"Follower Systems
These are the solid-state equivalent of the valve cathode follower layout, although offering superior performance and greater versatility. In the simple circuit shown in Figure 9.9 (the case shown is for an NPN transistor, but a virtually identical circuit, but with negative supply rails, could be made with a similar PNP transistor), the emitter will sit at a quiescent potential about 0.6 V more negative than that of the base, and this will follow, quite accurately, any signal voltage excursions applied to the base. (There are some caveats in respect of capacitative loads; these potential problems will be explored under the heading of slew rate limiting.)
Figure 9.9: Emitter"follower.
The output impedance of this circuit is low because this is approximately equal to 1/gm, and the gm of a typical small-signal, silicon BJT is of the order of 35 mA/V (35 mS) per mA of emitter current. So, if Q1 is operated at 5 mA, the expected output impedance, at low frequencies, will be 1/(5.35) kilohms, or 5.7 ohms, a value that is sufficiently smaller than any likely value for R1, that the presence of this resistor will not greatly affect the output impedance of the circuit.
The output impedance of a simple emitter"follower can be reduced still further by the circuit elaboration shown in Figure 9.10, known as a compound emitter"follower. In this, the output impedance is lowered in proportion to the effective current gain of Q2 in that, by analogy with the output impedance of an operational amplifier with overall NFB, any change in the potential of the Q1 emitter, brought about by an externally impressed voltage, will result in an opposing change in the collector current of Q2.
Figure 9.10: Compound emitter"follower.
This layout gives a comparable result to that of the Darlington pair, of two transistors, in cascade, connected as emitter"followers, shown in Figure 9.11, except that the arrangement of Figure 9.10 will only have an input/output DC offset equivalent to a single emitter-base forward voltage drop, whereas the layout of Figure 9.11 will have two, giving a combined quiescent voltage offset of the order of 1.3"1.5 V. Nevertheless, in commercial terms, the popularity of power transistors, connected internally as a Darlington pair, mainly for use in the output stages of audio amplifiers, is great enough for a range of single chip Darlington devices to be offered by the semiconductor manufacturers.
Figure 9.11: Darlington pair.
Thermal Dissipation Limits
9.6 Thermal Dissipation Limits
Unlike a thermionic valve, the active area of a BJT is very small, in the range of 0.5 mm for a small signal device to 4 mm or more for a power transistor. Because the physical area of the component is so small - this is a quite deliberate choice on the part of the manufacturer because it reduces the individual component cost by allowing a very large number of components to be fabricated on a single monocrystalline silicon slice - the slice thickness must also be kept as small as possible - values of 0.15"0.5 mm are typical - in order to assist the conduction of any heat evolved by the transistor action away from the collector junction to the metallic header on which the device is mounted.
Whereas in a valve, in which the internal electrode structure is quite massive and heat is lost by a combination of radiation and convection, the problem of overheating is usually the unwanted release of gases trapped in its internal metalwork, the problem in a BJT is the phenomenon known as thermal runaway. This can happen because the potential barrier of a P-N junction (that voltage that must be exceeded before current will flow in the forward direction) is temperature dependent and decreases with temperature.
Because there will be unavoidable nonuniformities in the doping levels across the junction, this will lead to nonuniform current flow through the junction sandwich, with the greatest flow taking place through the hottest region. If the ability of the device to conduct heat away from the junction is inadequate to prevent the junction temperature rising above permissible levels, this process can become cumulative. This will result in the total current flow through the device being funneled through some very small area of the junction, which may permanently damage the transistor. This malfunction is termed secondary breakdown, and the operating limits imposed by the need to avoid this failure mechanism are shown in Figure 9.12. Field effect devices do not suffer from this type of failure.
Figure 9.12 : Bipolar breakdown limits.
Coming up in Part 2: JFETs, MOSFETs, power BJTs vs. power MOSFETs as amplifier output devices, and other circuit layouts.
Printed with permission from Newnes. A division of Elsevier. Copyright 2009. "Audio Engineering: Know It All" by Douglas Self et al. For more information about this book and other similar titles, please visit www.newnespress.com.
1. Linsley Hood, J., Wireless World, 437 "441, September, 1971.
2. Linsley Hood, J., 'Low noise systems', Electronics Today International, 42 "46, 1992.
Distortion in power amplifiers, Part II: the input stage | Part III: the voltage-amplifier stage
Analog Circuits: World Class Design, Part 2 | Part 3
MOSFET or bipolar, which should you use?