Audio Amplifier With BJT

Transistors are the key components in many different kinds of audio preamplifiers, amplifiers, and tone-control circuits. Recent articles in this series have discussed the operation principles and applications for discrete bipolar junction transistors (BJT). Earlier articles have covered such subjects as low-power amplifier circuits, multivibrators, and oscillators.

Audio Amplifier Basics:
A modern stereo amplifier system has two closely matched high-fidelity audio amplifier channels. Typically each of those channels offers switch-selectable inputs for such signal sources as a tuner, tape-player, CD-player, TV, MTS, etc. Each also provides a single output signal to a high-power loudspeaker. To analyze one of those systems, it is useful to divide the system into three functional circuit blocks, as shown in Fig. 1.
The first of these blocks is the selector/preamplifier. It permits the system listener to select the desired input signal source, and it automatically applies an appropriate amplification level and frequency correction to the signal to condition it for the second circuit block, tone/volume control.

The tone/volume-control block permits the listener to adjust the frequency characteristics and the amplitude of the audible output to suit his individual taste. This block might also contain additional filter circuits including one specifically designed to screen out scratch and rumble.

The last section of the amplifier system is the power amplifier. It might be able to produce power levels from a few hundred milliwatts to hundreds of watts. Audio power amplifiers are designed to cover the audio frequency range with minimal distortion. Most quality products today include automatic overload and thermal-runaway protection.

The three sections of the audio amplifier system are all powered from a single built-in power supply. All three sections include individual power supply decoupling networks to prevent unwanted signal interference. The first two amplifier blocks will be discussed here.

Simple Preamplifiers:
The audio preamplifier circuit modifies the signal characteristics so that it will have a steady frequency response and the nominal 100-millivolt output amplitude necessary for driving the tone/volume control section.
If the input signal is derived from a radio tuner or a tape player, the signal characteristics are usually in a form that can be fed directly to the tone/control section, bypassing the preamplifier. However, if the input is obtained from a micro-phone or other audio input device, it will probably need preamplifier conditioning.
Two basic kinds of transducers are found in micro-phones and audio pickups: magnetic or piezoelectric ceramic/crystal. Magnetic transducers typically offer low output impedance and a low signal sensitivity of about 2 millivolts. Their outputs must be fed to a high-impedance preamplifier stage with near-unity voltage gain.

Most microphones have a near flat frequency response, so they can be matched to simple, flat-response preamplifier stages. Figure 2 shows a unity-gain preamplifier circuit that will work with most high-impedance ceramic or crystal microphones. It is an emitter-follower (common-collector) amplifier with an input network bootstrapped by C2 and R3. It has a typical input impedance of about 2 megohms. The combination of C5 and R5 decouples the amplifier from the DC power supply.
Figures 3 and 4 show alternative preamplifier circuits that will match magnetic microphones. The single-stage circuit of Fig. 3 gives 46dB (x200) of voltage gain, and will work with most magnetic microphones. The two-stage circuit of Fig. 4, however, gives 76dB of voltage gain, and it is intended for preamplification of the output of very-low-sensitivity magnetic microphones.



RIAA Preamplifier Circuits:
The replay of a constant-amplitude 20Hz to 20KHz variable-frequency signal that has been recorded on a phonograph disc with conventional stereo recording equipment will generate the nonlinear frequency response curve shown in Fig. 5. Here, the dotted line shows the idealized shape of this curve, and the solid line shows an actual shape.
Examination of the idealized (dotted) version of the curve in Fig. 5 will show that the response is flat between 500 and 2120 Hz. However, it rises at a rate of 6dB/octave (20 dB/decade above 2120 Hz), and falls at a 6dB/octave rate between 500 Hz and 50 Hz. The response then flattens at frequencies below 50Hz.

There are good--but difficult to explain--reasons why the precise Fig. 5 recording curves are used. However, all you really need to know is that they make it possible to produce disc recordings with excellent signal-to-noise ratios and wide dynamic ranges. The curves were applied during record pressing.
The important point to be made here is that when a disc is replayed, the output of the pickup device must be passed to the power amplifier through a preamplifier whose frequency equalization curve is the mirror image (exact inverse) of the one used to make the original recording. As a result, a linear overall record-to-replay response is obtained.
Figure 6 shows the RIAA equalization curve. RIAA is an abbreviation for the Recording Industry Association of America, the organization that standardized the precise specification of the curve for the equalization of phonograph records. When long-playing phonograph (record-player) records were the primary source of recorded music and audio entertainment, circuit designers had to include filter networks that corrected the input from the record to conform to the RIAA equalization curve.
The relatively recent(1994) world-wide conversion to compact discs (CDs) as the primary source of recorded music and entertainment has diminished the importance of the RIAA curve. Equalization is not required for linear signal sources such as CDs.

Nevertheless, a preamplifier with an RIAA equalization network is still needed if you want to play any of the pressed long-playing and 45 rpm records. This equalization can be obtained by wiring frequency-dependent, resistive capacitive feedback networks into a preamplifier. This circuitry causes the gain to fall as the frequency rises. One network will control the 50 to 500 Hz response, and the other will control the 2120 Hz to 20 kHz response.
Figure 7 is the schematic for an amplifier with those networks that will work with any magnetic phono cartridge. It gives a 1-volt output from a 6-millivolt input at 1KHz, and provides equalization that is within 1 dB of the RIAA standard between 40 Hz and 12KHz.


The preamplifier circuit is designed around transistors Q1 and Q2, with C2 and R5, and C3 and R6 forming the feedback resistor capacitor equalization network. The output of the emitter-follower buffer stage, transistor Q3, can be controlled by volume control potentiometer R10.
The quality of reproduction of ceramic or crystal phono cartridges is generally lower than that of magnetic cartridges, but they produce far higher amplitude output signals. Ceramic and crystal phone cartridges will work with simple equalization preamplifiers--one reason why those cartridges were installed in so many low-cost record players.

Figure 8 and 9 show alternative phone cartridge preamplifier/equalization circuits that can function with wither ceramic or crystal phono cartridges. Both circuits are designed around transistorized emitter-follower output stages Q1 and Q2. The output of the circuit in Fig. 8 can be controlled by volume control potentiometer R4, and that of Fig. 9 is controlled by R5.

The preamplifier/equalizer in Fig. 8 will work with any phone cartridge whose capacitance is between 1000 and 10,000pF. Two-stage equalization is provided by the resistance-capacitance network made up of C1, C2, R2, and R3. Preamplification/equalization for this circuit is typically within 1.6 dB of the RIAA standard between 40 Hz and 12KHz.
The alternative preamplifier/equalizer shown in Fig. 9 will work only with phono cartridges whose capacitance value are between 5000 and 10,000pF because this capacitance is part of the circuit's frequency response network. The other part of the network is formed by C1 and R3. At 50 Hz, this circuit has a high input impedance of about 600 kilohms, which causes only slight cartridge loading. However, as frequency increases, input impedance decreases sharply, increasing cartridge loading and effectively reducing circuit gain. The equalization curve approximates the RIAA standard, and circuit performance is adequate for most practical applications.

A Universal Preamplifier:
Most audio amplifier systems must have preamplifiers with many different characteristics. These include high-gain linear response for magnetic microphones, low-gain linear response for tuners, and high-gain RIAA equalization for magnetic phone cartridges.
To meet this broad requirement, most amplifier designers include a single universal preamplifier circuit such as the one shown in Fig. 10. Basically a high-gain linear amplifier, its characteristics can be altered by switching alternative resistor filter networks into its feedback system.
For example, when the selector switch is set to the Mag phono position, alternative input sources can be selected by S1-a, and appropriate linear-response gain control feedback resistors R8, R9, and R10 are now selected by S1-b. Those feedback resistor values are selected between 10 kilo ohms and 10 megohms to suit individual listener tastes. Circuit gain will be proportional, to the feedback resistor value.


Volume Control:
The Volume control circuitry of an audio amplifier system is normally located between the output of the preamplifier stage and the input of the tone-control circuit. It is usually only a potentiometer within the circuit, as shown in Figs. 7, 8, and 9. However, the catch here is that rapid rotation of the potentiometer knob can apply DC voltage to the next circuit for brief intervals. That voltage could upset circuit bias and cause severe signal distortion.
The block diagram in Fig. 11 shows the ideal topology and location for a volume control. It is fully DC-isolated from the output of the preamplifier by capacitor C1, and from the input of the tone-control circuit by C2. As a result, variation of the wiper of control potentiometer R1 has no effect on the DC bias levels of either circuit. Potentiometer R1 should have a logarithmic taper, that is, its output should be logarithmic function rather than linear.


Passive Tone Control:
A tone-control network permits the listener to change th system amplifier's frequency response to suit his own mood or taste. He can, for example, boost or reduce the low-frequency (treble) sections of a musical selection to emphasize the sounds of specific sections of the orchestra.
Tone-control networks typically consists of simple resistive-capacitive filters through which the signals are passed. Because these networks are passive, they cause some signal attenuation. Tone control networks can, if desired, be wired into the the feedback loops of simple transistor amplifiers to give the system an overall signal gain. Those are known as active tone control circuits.
Fig. 12-a shows a typical passive bass tone-control network, and Fig. 12-b through Fig. 12-d show the equivalent of this circuit when control potentiometer R3 is set to its maximum boost, maximum cut, and flat positions, respectively. Capacitors C1 and C2 are effectively open circuited when the frequency is at its lowest bass value. It can be seen from Fig. 12-b that the boost circuit is equivalent to a voltage divider formed by dividing 10 kilohms by 101 kilohms. This arrangement results in a low resistive value of about 100 ohms that only slightly attenuates bass signals.
The Fig. 12-c cut circuit, by contrast, has a voltage divider equal to 100 kilohms divided by a 1 kilohm which gives a signal attenuation of about 40 dB. Finally, in Fig. 12-d when potentiometer R3 is set to the flat position, it will have 90 kilohms of resistance above the wiper and 10 kilohms below it.
This circuit resistance value is equal to 100 kilohms divided by 11 kilohms.

It gives a signal attenuation of about 20 dB at all frequencies. As a result, the circuit gives a maximum bass boost of about 20 dB or cut relative to the flat signals.

Fig. 13 shows a typical passive treble tone-control network together with its equivalent circuits under maximum boost, maximum cut, and flat operating conditions. This circuit also provides about 20 dB of signal attenuation when potentiometer R3 is in the flat position, and it gives maximum treble boost or cut values of about 20 dB relative to its flat performance.

Finally, Fig. 14 shows how the Fig. 12-a and 13-a schematics can be combined to make a complete bass and treble tone-control network The 10-kilohm resistor R5 has been added to minimize unwanted interaction between the two connected circuit sections. The input to this network can be taken from the circuit's volume control, and its output can be fed to the input of the power amplifier.

Active Tone Controls:
A tone-control network can be included in the feedback path of a transistor amplifier so that the system will have an overall signal gain (rather than attenuation) when its controls are in the flat position. These networks can be simplifier versions of the basic circuit shown in Fig. 14. Fig. 15 is the schematic for an active tone-control circuit.

A comparison of Figs. 15 and 12 will reveal that the bass control section of Fig. 15 is a simplified version of Fig. 12-a. It can be seen that the two capacitors C1 and C2 of Fig. 12-a have been replaced by the single 0.039uF capacitor C2 of Fig. 15. Similarly, the treble version of Fig. 13-a, with resistors R1 and R2 eliminated. Resistors R3 and R4 balance the performance of the two section of the Fig. 15 control circuit.



An Audio Mixer:
A multichannel audio mixer is an attractive modification that can be added to the volume/tone-control section of an audio amplifier. This mixer permits several different audio signals to be mixed together to form a single composite output signal. This modification will be of value if, for example, you want to hear the front-door buzzer or the sounds of a baby crying in a child's room while you listening to music.

Figure 16 is the schematic for a three-channel audio mixer that will provide an overall gain of one between the output and each input channel. Each input channel includes a single 0.1 uF capacitor and a 100-kilohms resistor, to provide an output impedance of 100 kilohms. The number of input channels to this audio mixer can be increased by adding more capacitors and resistors with the same values as C1 and R1.

The mixer should be located between the output of the tone-control circuitry and the input to the power amplifier. One input should be taken from the output of the tone-control circuit, and the other inputs should either be grounded or taken from the desired source.




Ricardo A Monroy B
C.I 17646658
EES

---

Power Amplifier



An audio power amplifier can boost weak signals from a tuner, CD player, or tape deck to fill a room with sound. This article focuses on the operating principles and circuitry of low-frequency power amplifiers based on the bipolar junction transistor (BJT). Other articles in this series have discussed multivibrators, oscillators, audio preamplifiers, and tone-control circuits, all based on the BJT.

Power Amplifier Basics:
A transistorized audio power amplifier converts the medium-level, medium-impedance AC signal into a high-level, amplified signal that can drive a low-impedance audio transducer such as a speaker. A properly designed power amplifier will do this with minimal signal distortion.
Audio can be amplified with one or more power transistors in either of three configurations: Class A, Class B, and Class AB. Figure 1-a shows a single BJT Class A amplifier in a common-emitter configuration with a speaker as its collector load. A Class A amplifier can be identified by the way its input base is biased.
Fig. 1-a shows that BJT Q1's collector current has a quiescent value that is about halfway between the zero bias and cutoff positions. (The quiescent value is that value of transistor bias at which the negative- and positive-going AC input signals are zero.) This bias permits the positive and negative swings of the output collector AC current to reach their highest values without distortion. If the AC and DC impedances of the speaker load are equal, the collector voltage will assume a quiescent value that is about half the supply voltage.

The Class A circuit amplifies audio output with minimum distortion, but transistor Q1 consumes current continuously--even in the quiescent state--giving it low efficiency. Amplifier efficiency is defined as the ratio of AC power input to the load divided by the DC power consumed by the circuit.
At maximum output power, the efficiency of a typical Class A amplifier is only 40%, about 10% less than its theoretical 50% maximum. However, its efficiency falls to about 4% at one-tenth of its maximum output power level.
A typical Class B amplifier is shown in Fig. 2-a. It has a pair of BJTs, Q1 and Q2, operating 180° out-of-phase driving a common output load, in this example another speaker. In this topology, the BJTs operated as common-emitter amplifiers drive the speaker through push-pull transformer T2. A phase-splitting transformer T1, provides the input drives for Q1 and Q2 180° out-of-phase.
The outstanding characteristic of any Class B amplifier is that both transistors are biased off under quiescent conditions because they are operated without base bias. As a result, the amplifier draws almost no quiescent current. This gives it an efficiency that approaches 79% under all operating conditions. In Fig. 2-b, neither Q1 nor Q2 conducts until the input drive signal exceeds the base emitter zero-crossing voltage of the transistor. This occurs at about 600 millivolts for a typical power transistor.
The major disadvantage of the Class B amplifier is that its output signal is seriously distorted. THis can be seen from its dynamic transfer curve, also shown in Fig. 2-b.

Class AB Fundamentals:
Audio distortion caused by the crossover between two out-of-phase transistors is annoying. To overcome this defect, the Class B amplifier is modified into the third category called Class AB for most high-fidelity audio equipment. Fortunately, Class B distortion can usually be eliminated by slight forward bias to the base of each transistor, as shown in Fig. 3-a. This modification sharply reduces the quiescent current of a Class B amplifier and converts it into a Class AB amplifier.
Many early transistorized power amplifiers were Class AB, as shown in Fig. 3-a, but that circuit is rarely seen today. That circuit requires one transformer for input phase-splitting and another for driving the speaker, both costly electronics components.
In addition, electrical characteristics of both Q1 and Q2 must be closely matched. The amplification of each transistor will be unequal if they are not, and it will be impossible to minimize output distortion. Figure 3a shows a dynamic transfer characteristic for a Class AB power amplifier.
The Class AB amplifier shown in Fig. 4 avoids both transformers and the need to match transistors. A complementary pair of transistors (Q1 and NPN and Q2 a PNP) is connected as an emitter follower. Powered by a split (dual) supply, the circuit's two emitter followers are biased through R1 and R2 so that their outputs are at zero volts; no current flows in the speaker under quiescent conditions.

Nevertheless, a slight forward bias can be applied with trimmer potentiometer R3 so that Q1 and Q2 pass modest quiescent currents to prevent crossover distortion. Identical input signals are applied through C1 and C2 to the base of the emitter followers, which avoid a split-phase drive.
When an input signal is applied to the Fig. 4 circuit, the positive swing drives PNP Q2 off while driving NPN Q1 on. Transistor Q1 acts as current source with a very low output (emitter) impedance if feeds a faithful unity-gain copy of the input voltage signal to the speaker. The transistor characteristics have little or no effect on this response.
Similarly, negative swings of the input signal drive Q1 off and Q2 on. Because Q2 is a PNP BJT, it becomes a current sink with minimal input (emitter) impedance. It also produces a faithful unity-gain copy of the voltage signal to the speaker, again with Q2's characteristics having little or no effect on the circuit's response.
As a result, the Fig. 4 circuit does not require that Q1 be matched to Q2, and neither input nor output transformers are required. Modification of this circuit, as shown in Figs. 5-a and b, work from single ended power supplies. In Fig. 5-a, one side of the speaker is connected to the amplifier through high-value blocking capacitor C3 and, and the other end is connected to ground; in Fig. 5-b, one side is connected to C3 and the other side is connected to the positive supply. All three circuits are popular in modern high-fidelity audio power amplifiers based on integrated circuitry.

Class AB Variations:
The circuit in Figs. 4-a is a unity-voltage gain amplifier so one obvious improvement is to add a voltage-amplifying driver stage, as shown in Figs. 6. Transistor Q1, configured as a common-emitter amplifier, drives two emitter followers, Q2 and Q3, through its collector load resistor R1.
Note that Q1's base bias is derived from the circuit's output through resistors R2 and R3. This configuration provides DC feedback to stabilize the circuit's operating points and AC feedback to minimize signal distortion.
The Fig. 6 circuit illustrates how a form of auto-bias can be applied to Q2 and Q3 through the silicon diodes D1 and D2. If the simple voltage-divider biasing method in Fig. 4 is used in the Fig. 6 circuit, its quiescent current will increase as ambient temperature rises and decrease as it fall. (This is caused by the thermal characteristics of a transistor's base-emitter junction.)
The biasing in Fig. 6 is derived from the forward voltage drop of series diodes D1 and D2 whose thermal characteristics are closely matched to those of the base-emitter junctions of Q2 and Q3. Consequently, this circuit offers excellent thermal compensation.
Practical amplifiers include a pre-set trimmer potentiometer in series with D1 and D2. This component makes it possible to adjust biased voltage over a limited range. Low-value resistors R4 and R5 in series with the emitters of Q2 and Q3 provide some negative DC feedback.
The impedance of the Fig. 4 circuit equals the product of the speaker load impedance and the current gain of either Q1 or Q2. The circuit can be improved by replacing transistors Q1 and Q2 with Darlingron pairs which will significantly increase the circuit's input impedance and increase the amplifier's collector load capacity.
Figures 7 to 9 show three different ways of modifying the Fig. 6 circuit by replacing individual transistors with Darlington pairs. For example, in Fig. 7, transistors Q2 and Q3 form a Darlingron NPN pair, and Q4 and Q5 form a darlington PNP pair. There are four base-emitter junctions between the bases of Q2 and Q4, and the output circuit is biased with a string of four silicon diodes, D1 and D4, in series to compensate for the Darlingron pairs.
Figure 8, Q2 and Q3 are a Darlington NPN pair, but Q4 and Q5 are a complementary pair of common-emitter amplifiers. They operate with 100% negative feedback, and provide unity-voltage gain and very high input impedance. Thisquasi-complementary output stage is probably the most popular Class AB power amplifier topology today. Notice the three silicon biasing diodes, D1, D2, and D3.
Finally, in Figure 9, both pairs Q2 and Q3 and Q4 and Q5 are complementary pair of unity-gain, common-emitter amplifiers with 100% negative feedback. Because the pairs produce outputs that are mirror images of each other, the circuit has a complementary output stage. Notice that this circuit has only two silicon biasing diodes, D1 and D2.

Amplified Diodes:
The circuits in Figs. 6 to 9 include strings of two to four silicon biasing diodes. Each of those strings can be replaced by single transistor and two resistors configured as an amplified diode, as shown in Figs. 10.
The output voltage of the circuit, V_out can be calculated from the formula: V_out = V_BE x R1 + R2/R2
If resistor R1 is replaced by a short circuit, the circuit's output will be equal to the base-emitter junction "diode" voltage of Q1 (V_BE). The circuit will then have the thermal characteristics of a discrete diode.

      If resistor R1 equals R2, the circuit will act like two series-connected diodes, and if R1 equals three times R2, the circuit will act like four series-connected diodes, and so on. Therefore, the circuit in Figs. 10 can be made to simulate any desired whole or fractional number of series-connected diodes, depending on how the R1/R2 ratios are adjusted.
Figure 11 shows how the circuit in Fig. 10 can be modified to act as a fully adjustable "amplifier diode", with an output variable from 1 to 5.7 times the base-emitter junction voltage (V_BE)

   Bootstrapping:
The main purpose of the Q1 driver stage in Fig. 6, the base complementary amplifier, is to give the amplifier significant voltage gain. At any given value of Q1 collector current, this voltage gain is directly proportional to the effective Q1 collector load value. It follows that the value of resistor R1 should be as large as possible to maximize voltage gain. However, there are several reasons why this does not work.
First, the effective or AC value of R1 equals the actual R1 value shunted by the input impedance of the Q2-Q3 power amplifier stage. Therefore, if R1 has a higher value, the power amplifier input impedance must be even greater. That can usually be done by replacing Q2 and Q3 with high-gain transistor pairs, as was done in Figs. 7 to 9.
The second reason is that Q1 in Fig. 6 must be biased so that its collector assumes a quiescent half-supply voltage value to provide maximum output signal swings; this condition is set by the Q1's collector current and resistor R1's value.
The true value of R1 is predetermined by biasing requirements. To achieve high voltage gain, a way must be found to make the AC impedance of R1 much greater than its DC value. This is accomplished with he bootstrapping technique shown in Figs. 12 & 13.
In Fig. 12, Q1's collector load consists of R1 and R2 in series. The circuit's output signal, which also appears across SPKR1, is fed back to the R1-R2 junction through C2. This output signal is a near unity-voltage-gain copy of the signal appearing on Q1's collector.
If resistor R1 has a value of 1 kilohm, the Q2-Q3 stage provides a voltage gain of 0.9. As a result, an undefined signal voltage appears at the low end of resistor R2, and 0.9 times that undefined voltage appears at the top of R2. In other words, only one-tenth of the unknown signal voltage is developed across R2. Therefore, it passes one-tenth of the signal current that would be expected from a 1-kilohm resistor.
This means that the AC signal impedance value of R2 is ten times greater (10-kilohms) than its DC value, and the signal voltage gain is increased correspondingly. In practical circuits, "bootstrapping" permits the effective voltage gain and collector load impedance of Q1 to be increased by the factor of about twenty.
Fig. 13 is the schematic for an alternative version of Fig. 12 without one resistor and one capacitor. In this circuit. SPKR1 is part of Q1's collector load, and it is bootstrapped through capacitor C2.
As an alternative to bootstrapping, the load resistor can be replaced with a simple transistor constant-current generator. This design is found in many integrated circuit audio power amplifiers.

Alternative Drivers:
Returning once again to Fig. 6, notice that parallel DC and AC voltage form the R1-R2 divider network is fed back to the Q1 driver stage. This is a simple and stable circuit, but its gain and input impedance are low. Moreover, it will work only over a limited power supply voltage range.
Figure 14 is a variation of the Fig. 6 circuit intended to function as a driver stage. Current feedback through resistors R1 and R2 allows the circuit to work over a wide supply voltage range. The feedback resistors can be AC decoupled (as shown) through C2 to increase the gain and input impedance, but at the expense of increased signal distortion. Transistor Q1 can be replaced with a Darlington pair if very high input impedance is desired.

Another alternative driver stage, Fig. 15, depends on series DC and AC feedback to give it more gain and higher input impedance than can be obtained from the Fig. 6 circuit. In this circuit, PNP transistor Q1 is directly coupled to NPN transistor Q2.
Finally, Fig. 16 is the schematic for a driver circuit specifically intended for use in amplifiers with dual or split power supplies that have direct-coupled input and output stages referenced to ground. The input stage of this driver stage is a long-tailed pair. Both the input and output will be centered on DC ground if the values of resistors R1 and R4 are equal. This circuit is found in many integrated circuit power amplifiers.
An IC power amplifier:
Improvements in the power-handling capabilities of monolithic integrated circuits have permitted power amplifier to be integrated on a single silicon substrate or chip. The techniques for designing integrated circuit power amplifiers are similar to those for discrete device circuits. It turns out that the similarities between discrete and IC power amplifier designs are closer than for most other linear circuits.
Figure 17 is a simplified circuit diagram for the LM380, an IC power amplifier, drawn in the manufacturer's data book style. The LM380 was developed by National Semiconductor Corporation for consumer applications. It features an internally fixed gain of 50 (34 dB) and an output that automatically centers itself at one-half of the supply voltage.
An unusual input stage permits inputs to be referenced to the ground or AC coupled, as required. The output stage of the LM380 is protected with both short-circuit current limiting and thermal-shutdown circuitry.
The LM380 has two input terminals. Both Q1 and Q2 are connected as PNP emitter followers that drive the Q3 and Q4 differential amplifier transistor pairs. The PNP inputs reference the input to gro8und, thus permitting direct coupling of the input transducer.


The output is biased to half the supply voltage by resistor ratio R1/R2 (resistor R1 is formed by two 25-kilohm resistors and R2 has a value of 25-kilohms). Negative DC feedback, through resistor R2, balances the differential stage with the output at half supply, because R1 = R2.
The output of the differential amplifier stage is direct coupled into the base of Q12, which is a common-emitter, voltage-gain amplifier with a constant current-source load provide by Q11. Internal compensation is provided by the pole-splitting capacitor C'. Pole-splitting compensation permits wide power bandwidth (100 KHz at 2 watts, 8 ohms).
The collector signal of Q12 is fed to output pin 8 of the IC through the combination of emitter-coupled Q7 and the quasi-complementary pair emitter followers Q8 and Q9. The short-circuit current is typical 1.3 amperes.

Ricardo A Monroy B
C.I 17646658
EES

---

An Audio Amp:
This particular project involved injecting the audio from a TV receiver into a stereo system. (These days even the cheapest TV has that feature, including MTS stereo inputs for digital accessories). Anyways, the audio-output portion of the TV-audio receiver was abandoned because of its poor frequency response and high distortion. Instead, we wanted to come right off the detector into a quality audio amplifier and speaker. So, after picking off the audio at a convenient point in the set (in this case, from a potentiometer), we wanted to feed it to the auxiliary input of the stereo amplifier.

The amplifier we used required an input of 1 volt RMS, but a quick check with an AC VTVM indicated that out picked-off audio signal was only 0.1-volt RMS. Obviously, an amplifier with a gain of 10 was needed.

Scanning the literature on transistor amplifiers reveled that a common-emitter amplifier with a voltage-divider bias circuit would solve our problem nicely. Such a circuit is shown in Fig. 1. Some of that circuit's characteristics include: moderate input impedance, moderate voltage gain, inverted output, and input/output impedance and gain that depend only slight on transistor beta.
There are, of course, several rules that must be followed in using a common-emitter amplifier, including:

• With a positive supply use an NPN transistor.
• With a negative supply use an PNP transistor.
• The supply voltage must not exceed the transistor's Vce rating.
• The power-dissipation rating of the transistor must not be exceeded.
  
• The beta of the transistor should be 100 or higher.

In our example the following facts are known:

• Our amplifier had a single-ended 12-volt power supply.
• We need a voltage gain of 100.
• The input impedance of the amplifier should be about 15K, the same as the potentiometer from which the audio was taken.
• The impedance of the stereo amplifier's auxiliary input is about 50K.

As is the case in most circuit designs, a few facts are known, and the rest must be calculated or picked using a a few "rules of thumb". We will learn how to make the calculations next.

Doing the Math:
For maximum undistorted output swing, we will make the quiescent collector voltage 1/2 the supply voltage. See Fig. 2. The drop across R_c must therefore be 6 volts.
The value of R_c, the collector load resistance, is chosen considering output impedance, gain, and collector current. If possible, the output impedance should be lower than the impedance of the circuit we are feeding by a factor of 10 or more. Doing so will avoid circuit loading. So let's make R_c equal to 4700 ohms, which is about 50K/10. Collector current I_c, is equal to 0.5V_cc/R_c, or 6/4700 = 1.28 mA. That current is certainly low enough that we will not exceed any collector-current ratings, so let's go on.
To achieve maximum stability, the emitter resistor should be in the range of 40 to 1000 ohms. Voltage gain (A_v) = R_c/R_e, so R_e = R_c/A_v. In our case R_e equals 4700/10, or 470 ohms. That falls within the range of acceptable values.
The current through the emitter resistor consists of the collector current plus the base current. The base current here is significantly smaller than the collector current, so it can be ignored for the next calculation.

The voltage drop across the emitter resistor = I_c X R_e, or 1.28 mA x 470 ohms = 0.602 volts. The base voltage must exceed the emitter voltage by 0.6 volts for a silicon transistor and by 0.2 volts for a germanium transistor. We'll use a silicon transistor (most if not all germanium types are obsolete) in our circuit, so the base voltage must be 0.6 + 0.602 = 1.202 volts.
The input impedance of the circuit equals R2 in parallel with the emitter resistor times beta; input impedance will vary with the transistor's beta. FOr our example, assume we are using a transistor with a beta of 100. We want the input impedance to be about 15000 ohms. Solving for R2, we find:

    Zin = (R2 X Re X beta)/[R2 + (Re X beta)]
    R2 = (Zin X Re X beta)[(Re X beta) - Zin]
    R2 = (15000 x 470 x 100)/[470 x 100) - 15000]
    R2 = 22,030 ohms.

We can use a 22K resistor. In general, if input impedance is not critical, for maximum stability R2 can be 10 to 20 times R_e.
The drop across R2 must be 1.20 volts so the current through R2 is 1.20/22,000, or 0.054 mA. Therefore, R1 must drop the rest of the supply voltage, which is 12 - 1.20 = 10.8 volts. The current flowing through R1 is a combination of the voltage-divider current plus the base current.
The base current is equal to the collector current divided by beta. It is found from:

    Ibeta = 1.28/100 = 0.0128 mA

So the total current through R1 is 0.054mA + 0.0128mA = 0.067mA, and R1 = 10.8/0.067mA = 160,000 ohms (160K).



Resistor R1 is the most critical resistor in the circuit. To ensure maximum voltage swing, it should bring the quiescent collector voltage to one half the supply voltage. After building the circuit, the value of R1 may have to be varied slightly to achieve that voltage swing.
We now have a circuit we can test.

Interfacing:
Connecting the circuit to the outside world will require capacitor coupling. That serves to isolate the AC signal from any DC bias voltages. Figure 3 shows our complete circuit with input and output coupling capacitors. The values of those capacitors were calculated using C = 1/(3.2 x ƒ x R), where C equals the capacitor value in farads, ƒ equals the frequency at which response will be down 1dB, and R equals the impedance on the load side of the capacitor.
To calculate the value of C1, the amplifier's input impedance (15K) is used for R. To calculate the value of C2, the input impedance of the next stage (50K) is used for R.

The value of C1 can now be calculated for a drop of 1dB at 20 Hz: C1 = 1/(3.2 x 20 x 15000) = .00000104 farad = 1.0 uF. The value of C2 = 1/(3.2 x 20 x 50000) = .00000031 farad = 0.33uF.
To increase the gain of the stage, you could bypass R_e with a capacitor, as shown in Fig. 4. Nothing comes for free, however. The price you pay for increase gain is lower input impedance, which will vary widely with beta. If that variation is not a problem, a significant gain increase can be realized by adding the bypass capacitor. Our original circuit has a gain of 10; if the emitter is bypassed the gain becomes R_c/003/I_e = 4700/(0.03/0.00129) = 4700/23 = 200 (approx).
The value of the bypass capacitor in farads is calculated from the formula C = 1/(6.2 x ƒ x R). Again ƒ is the low-frequency limit in Hz, and R is the dynamic emitter resistance (0.031/I_e). In our example, if we stick to a 20-Hz lower limit we have C = 1/[6.2 x 20 x (0.03/0.00129)] = .000344 farads = 344 uF. A 350uF unit can be used.

Thoughts:
A few thoughts on components before we finish: using 5% resistors allows closer adherence to the calculated values. Because of their temperature stability and low leakage specifications, silicon rather than germanium transistors are preferable for this type of circuit.
Finally, you've no doubt noticed that we have yet to specify a specific transistor. That's because for this type of application it really doesn't matter! Almost any small signal device will do fine.

Ricardo A monroy B
C.I. 17646658
EES

---

Circuit Applications:
Even a simple small-signal BJT has many applications related to its ability to amplify or switch. Some of the most important and practical circuit designs are described her. With few exceptions, all of the circuits are based on the 2N3904 NPN transistor. (With certain minor component value changes, other NPN transistor can be substituted.) The circuits can also be made with a PNP transistor such as the 2N3906, if the polarities are altered.

Diodes and Switches:
It was explained earlier that both the base-emitter and base-collector junctions of a silicon BJT can be considered equivalent to a zener diode. As a result, either of these junctions can perform as a fast-acting rectifier diode or zener diode, depending on the bias polarity.
Figure 6 shows two alternative ways to make an NPN BJT perform as a diode in a clamping circuit that converts an AC-coupled rectangular input waveform into a DC square wave. The input AC waveform is symmetrical above and below the zero-voltage reference. However, the output signal retains the input's form and amplitude, but it is clamped to the zero-voltage reference.


























Figure 7 shows how an NPN BJT can function as a zener diode in a circuit that converts an  unregulated supply voltage into a fixed-value regulated output voltage. Typical values range from 5 to 10 volts, depending on the characteristics of the selected transistor. The base emitter junction is the only one suitable for this application.


Figure 8 shows a BJT functioning as a simple electronics switch or digital inverter. Here the base is driven through resistor R_b by a digital input step voltage that has a positive value. The load resistor R_l can be a simple resistor, tungsten lamp filament, or a relay coil. Connect the load between the collector and the positive supply.
When the input voltage is zero, the transistor switch is cut off. Thus no current flows through the load, and the full supply voltage is available between the collector and emitter terminals. When the input voltage is high, the transistor switch is driven fully on. Maximum current flows in the load, and only a few hundred millivolts is developed between the collector and emitter terminals. Thus the output voltage signal is the inverted form of the input signal.
















Linear Amplifiers:
A BJT can function as a linear current or voltage amplifier if a suitable bias current is fed into its base, and the output signal is applied between a suitable pair of terminals. A transistor amplifier can be configured for any of three operating modes: common-emitter(Fig. 9), common-base(Fig. 10), and common-collector(Fig. 11). Each of these modes offers a unique set of characteristics.
In the common-emitter circuit of Fig. 9, load resistor R_l is connected between the collector and the positive supply, and a bias current is fed into the base through R_b. The value of R_b was selected so that the collector takes on a quiescent value of about half the supply voltage (to provide maximum undistorted signal swings).

The input signal in the form of a sine wave is applied between the base and the emitter through C1. The circuit inverts the phase of the input signal, which appears as an output between the collector and emitter. This circuit is characterized by a medium-value input impedance and a high overall voltage gain.
The input impedance of this amplifier is between 500 and 2000 ohms, and the load impedance equals R_l. Voltage gain is the change in collector voltage divided by the change in base voltage (from 100 to about 1000). Current gain is the change in collector current divided by the change in base current of H_fe.
In the common-base linear amplifier circuit of Figure 10, the base is biased through R_b and AC-decoupled (or AC-grounded) through C_b. The input signal is applied between the emitter and base through C1, and the amplified but non-inverted output signal is taken from between the collector and base. This amplifier offers very low input impedance, and output impedance equal to the resistor R_l. Voltage gain is from 100 to 1000, but current gain is near-unity.

In the common-collector linear amplifier circuit of Fig. 11, the collector is connected directly to the positive voltage supply, placing it effectively at ground impedance level The input signal is applied directly between the base and ground (collector), and the non-inverted output signal is taken between the emitter and ground (collector).
The input impedance of this amplifier is very high; it is equal to the product of h_fe and the load resistance R_l. However, output impedance is very low. The circuit's overall voltage gain is near-unity, and its output voltage is about 600 millivolts less than the input voltage. As a result, this circuit is know as a DC-voltage follower or an emitter follower. A circuit with very high input impedance can be obtained by replacing the single transistor of the amplifier of Fig. 11 with a pair of transistors connected in a Darlington configuration, as shown in Fig. 12. Here, the emitter current of the input transistor feeds directly into the base of the output transistor with an overall h_fe value equal to the product of the values for the individual BJT's. For example, if each BJT has an h_fe of 100, the pair acts like single transistor with an h_fe of 10,000. Darlington BJT's with two transistors on a single chip (considered to be discrete device) are readily available for power amplification.
The voltage-follower circuit of Fig. 11 can be modified for an alternating current input by biasing the transistor base with a value equal to half the supply voltage and feeding the input signal to the base. Figure 14 shows how this particular circuit is structured.
The emitter-follower circuits of Figs. 12 to 14 can source or feed relatively high currents into an external load through the emitter of the transistor. However, those circuits cannot sink or absorb high currents that are fed to the emitter from an external voltage source because the emitter is reverse-biased under this condition. As a result, these circuits have only a unilateral output capability.
In many applications, (such as audio amplifier output stages), a bilateral output characteristic is essential. A bilateral amplifier has equal sink and source output capabilities. This is obtained with the complementary emitter-follower circuit of Fig. 14. The series-connected NPN-PNP transistor pair is biased to give a modest quiescent current through the network consisting of resistors R1 and R2 and diodes D1 and D2. Transistor Q1 can provided large source currents, and Q2 can absorb large sink currents.

  

Phase Splitters:
Transistor linear amplifiers can be used in active filters or oscillators by connecting suitable feedback networks between their inputs and outputs. Phase splitting is another useful linear amplifier application. It provides a pair of output signals from a single input signal: one is in phase with the input phase, and the other is inverted or 180° out of phase. Fig. 16 and 17 show these alternative circuits.
In the circuit shown in Fig. 15, the BJT is connected as a common-emitter amplifier with nearly 100% negative feedback applied through emitter resistor R4. It has the same value as collector resistor R3. This configuration provides a unity-gain inverted waveform at output 1 and a unity-gain non-inverted waveform at output 2.
The phase-splitter circuit shown in Fig. 16 is known as a long-tailed pair because the two BJT's share common-emitter feedback resistor R7. An increasing waveform applied at the base of transistor Q1 causes the voltage to increase across resistor R7, reducing the bias voltage on transistor Q2. This results in the generation of an inverted waveform at the collector of Q1 (at output 1), and an in-phase waveform at the collector of Q2, (at output 2).

   

Phase Splitters:
Transistor linear amplifiers can be used in active filters or oscillators by connecting suitable feedback networks between their inputs and outputs. Phase splitting is another useful linear amplifier application. It provides a pair of output signals from a single input signal: one is in phase with the input phase, and the other is inverted or 180° out of phase. Fig. 16 and 17 show these alternative circuits.

In the circuit shown in Fig. 15, the BJT is connected as a common-emitter amplifier with nearly 100% negative feedback applied through emitter resistor R4. It has the same value as collector resistor R3. This configuration provides a unity-gain inverted waveform at output 1 and a unity-gain non-inverted waveform at output 2.
The phase-splitter circuit shown in Fig. 16 is known as a long-tailed pair because the two BJT's share common-emitter feedback resistor R7. An increasing waveform applied at the base of transistor Q1 causes the voltage to increase across resistor R7, reducing the bias voltage on transistor Q2. This results in the generation of an inverted waveform at the collector of Q1 (at output 1), and an in-phase waveform at the collector of Q2, (at output 2).

        

Multivibrators:
Figures 17 to 20 show BJT's in the four different kinds of multivibrator circuit: bistable, astable, monostable, and Schmitt trigger.
The bistable multivibrator is a simple electronic circuit that has two stable states. It is more often known as the flip-flop, but is also called a binary multivibrator, or an Eccles-Jordan circuit. The circuit is switched from one state to the other by a pulse or other external signal. It maintains its state to the other by a pulse or other external signal. It maintains its state indefinitely unless another input signal is received.
Figure 17 is a simple, manually-triggered, cross-coupled bistable multivibrator. The base bias of each transistor is obtained from the collector of the other transistor. Thus one transistor automatically turns OFF when the other turns ON, and this cycle can be continued in definitely as long as it is powered.
The output of the multivibrator in Fig. 17 can be driven low by turning off transistor Q2 with switch S2. The circuit remains "locked" or stable in this state until transistor Q1 is turned off with switch S1. At that time, the output is locked into its high state, and the process is repeated. It can be seen that this action makes it a simple digital memory circuit that holds its state until manually or electronically switched.
Figure 18 is the schematic for a monostable multivibrator or one-shot pulse generator. It has only one state. The output of this circuit, a manually triggered version, is normally low, but it switches high for a period determined by the values of capacitor C1 and resistor R2 if transistor Q1 is turned off with switch S1. It then returns to tits original state.
The pulse duration time of the monostable multivibrator can be determined from the equation: T = 0.69 RC
Where: T is in microseconds, R is in ohms, and C is in microfarads.
Monostable multivibrators are used as pulse generators and weep generators for cathode-ray tubes.

        

Figure 19 is the schematic for an astable multivibrator or free-running, square-wave oscillator. The transistors are in a common-emitter configuration so that the output of one is fed directly to the input of the other. Two resistance-capacitor networks, R3 and C1, and R2 and C2, determine the oscillation frequency.
The output of each transistor is 180° out of phase with the input. An oscillating pulse might begin at the base of Q1. It is inverted at the collector of Q1 and is sent to the base of Q2. It is again inverted at the collector of Q2 and therefore returns to the base of Q1 in its original phase. This produces positive feedback, resulting in sustained oscillation.
The astable multivibrator is frequently used as an audio oscillator, but is not usually used in radio-frequency circuits because its output is rich in harmonics.
Figure 20 is a schematic for a Schmitt Trigger, a form of bistable multivibrator circuit. It produces rectangular waves, regardless of the input waveform. The circuit is widely used to convert sine waves to square waves where these is a requirement for a train of pulses with constant amplitude.
The Schmitt trigger circuit remains off until the rising input waveform crosses the preset threshold trigger-voltage level set by the value of resistors R1 and R2. When transistor Q1 is switched 'on', transistor Q2 is 'off' and, the Schmitt trigger's output voltage rises abruptly.
When the input signal falls back below its drop-out level, Q1 switches 'off' and Q2 switches 'on'. The output voltage of the Schmitt trigger drops to zero almost instantly. This cycle of events will then be repeated in definitely, as long as the input signal is applied.

Ricardo A monroy B
C.I 17646658

EES

---

History of the Transistor

Dr. John Bardeen(left), Dr. Walter Brattain(right), and Dr. William Shockley(center) discovered the transistor effect and developed the first device in December, 1947, while the three were members of the technical staff at Bell Laboratories in Murray Hill, NJ. They were awarded the Nobel Prize in physics in 1956.



The PN Junction:
What the Bell Labs scientists discovered was that silicon was comprised of two distinct regions differentiated by the way in which they favored current flow. The area that favored positive current flow they named "P" and the area that favored negative current flow they named "N." More importantly, they determined the impurities that caused these tendencies in the "P" and "N" regions and could reproduce them at will. With the discovery of the P-N junction and the ability to control its properties, the fundamental ground work was laid for the invention of the transistor. This Bell Labs discovery was instrumental in the development of all semiconductor devices to come.



The Invention of the First Transistor
November 17 - December 23, 1947

The First Silicon Transistor:
It was late afternoon in 1954 at a conference for the Institute of Radio Engineers. Many people giving talks had complained about the current germanium transistors--they had a bad habit of not working at high temperatures. Silicon, since it's right above germanium on the periodic table and has similar properties, might make a better gadget. But, they said, no one should expect a silicon transistor for years.

Then Gordon Teal of Texas Instruments stood up to give his talk. He pulled three small objects out of his pocket and announced: "Contrary to what my colleagues have told you about the bleak prospects for silicon transistors, I happen to have a few of them here in my pocket."

That moment catapulted TI from a small start-up electronics company into a major player. They were the first company to produce silicon transistors -- and consequently the first company to produce a truly consistent mass-produced transistor.

Scientists knew about the problems with germanium transistors. Germanium worked, but it had its mood swings. When the germanium heated up--a natural outcome of being part of an electrical circuit--the transistor would have too many free electrons. Since a transistor only works because it has a specific, limited amount of electrons running around, high heat could stop a transistor from working altogether.

While still working at Bell Labs in 1950, Teal began growing silicon crystals to see if they might work better. But just as it had taken years to produce pure enough germanium, it took several years to produce pure enough silicon. By the time he succeeded, Teal was working at Texas Instruments. Luring someone as knowledgeable about crystals as Teal away from Bell proved to be one of the most important things TI ever did.

On April 14, 1954, Gordon Teal showed TI's Vice President, Pat Haggerty, a working silicon transistor. Haggerty knew if they could be the first to sell these new transistors, they'd have it made. The company jumped into action -- four weeks later when Teal told his colleagues about the silicon transistors in his pocket, TI had already started production.

Getting 'Wet':

On November 17, 1947, Walter Brattain dumped his whole experiment into a thermos of water. The silicon contraption he'd built was supposed to help him study how electrons acted on the surface of a semiconductor -- and why whatever they were doing made it impossible to build an amplifier. But condensation kept forming on the silicon and messing up the experiment. To get rid of that condensation, Brattain probably should have put the silicon in a vacuum, but he decided that would take too long. Instead he just dumped the whole experiment under water -- it certainly got rid of the condensation!

Out of the blue, the wet device created the largest amplification he'd seen so far. He and another scientist, Robert Gibney, stared at the experiment, stunned. They began fiddling with different knobs and buttons: by turning on a positive voltage they increased the effect even more; turning it to negative could get rid of it completely. It seemed that whatever those electrons had been doing on the surface to block amplification had somehow been canceled out by the water--the greatest obstacle to building an amplifier had been overcome.

Putting the Idea to Use.

When John Bardeen was told what had happened he thought of a new way to make an amplifier. On November 21, Bardeen suggested pushing a metal point into the silicon surrounded by distilled water. The water would eliminate that exasperating electron problem just under the point as it had in the thermos. The tough part was that the contact point couldn't touch the water, it must only touch the silicon. But as always, Brattain was a genius in the lab. He could build anything. And when this amplifier was built, it worked. Of course, there was only a tiny bit of amplification--but it worked.

Big Amplification.

Once they'd gotten slight amplification with that tiny drop of water, Bardeen and Brattain figured they were on the road to something worthwhile. Using different materials and different setups and different electrolytes in place of the water, the two men tried to get an even bigger increase in current. Then on December 8, Bardeen suggested they replace the silicon with germanium. They got a current jump, all right--an amplification of some 330 times--but in the exact opposite direction they'd expected. Instead of moving the electrons along, the electrolyte was getting the holes moving. But amplification is amplification -- it was a start.

Brattain Makes a Mistake

Unfortunately this giant jump in amplification only worked for certain types of current -- ones with very low frequencies. That wouldn't work for a phone line, which has to handle all the complex frequencies of a person's voice. So the next step was to get it to work at all kinds of frequencies.

Bardeen and Brattain thought it might be the liquid which was the problem. So they replaced it with germanium dioxide -- which is essentially a little bit of germanium rust. Gibney prepared a special slab of germanium with a shimmering green oxide layer on one side. On December 12, Brattain began to insert the point contacts.

Nothing happened:

In fact the device worked as if there was no oxide layer at all. And as Brattain poked the gold contact in again and again, he realized that's because there wasn't an oxide layer. He had washed it off by accident. Brattain was furious with himself, but decided to fiddle with the point contact anyway. To his surprise, he actually got some voltage amplification -- and more importantly he could get it at all frequencies! The gold contact was putting holes into the germanium and these holes canceled out the effect of the electrons at the surface, the same way the water had. But this was much better than the version that used water, because now, the device was increasing the current at all frequencies.

Bringing it All Together:
In the past month, Bardeen and Brattain had managed to get a large amplification at some frequencies and they'd gotten a small amplification for all frequencies -- now they just had to combine the two. They knew that the key components were a slab of germanium and two gold point contacts just fractions of a millimeter apart. Walter Brattain put a ribbon of gold foil around a plastic triangle, and sliced it through at one of the points. By putting the point of the triangle gently down on the germanium, they saw a fantastic effect -- signal came in through one gold contact and increased as as it raced out the other. The first point-contact transistor had been made.

Telling the Brass:
For a week, the scientists kept their success a secret. Shockley asked Bardeen and Brattain to show off their little plastic triangle at a group meeting to the lab and the higher-ups on December 23. After the rest of the lab had a chance to look it over and conduct a few tests, it was official -- this tiny bit of germanium, plastic and gold was the first working solid state amplifier.

---------------
Shockley Invents the Junction Transistor
January and February, 1948

A Solitary New Year's Eve

William Shockley spent New Year's Eve alone in a hotel in Chicago. He was there for a Physical Society meeting, but he was most excited about having some time to himself to concentrate on his work. There may have been a party going on downstairs, but Shockley wanted nothing to do with it. He had more important things to think about. He spent that night and the next two days working on some of his ideas for a new transistor-one that would improve on Bardeen and Brattain's ideas.

Scratching page after page into his notebook, one of Shockley's ideas was to build a semiconductor "sandwich." Three layers of semiconductors all piled together, he thought, just might work like a vacuum tube-with the middle layer turning current on and off at will. After some 30 pages of notes, the concept hadn't quite come together so Shockley set it aside to do other work.

The Idea Comes Together

Shockley's January was pretty dismal. He thought he should get sole credit for inventing the transistor--the initial research ideas, after all, had been his own. The Bell Labs attorneys didn't agree. They refused to even put him on the patent. The only thing to do, Shockley decided, was to build a better mouse trap.

As the rest of the group worked merrily away on improving Brattain and Bardeen's point-junction transistor. Shockley concentrated on his own ideas -- never letting anyone else in the lab know what he was up to.

On January 23, unable to sleep, Shockley was sitting at the kitchen table bright and early in the morning. He suddenly had a revelation. Building on the "sandwich" device he'd come up with on New Year's Eve, he thought he had an idea for an improved transistor. This would be three-layered sandwich. The outermost pieces would be semiconductors with too many electrons, while the bit in the middle would have too few electrons. The middle layer would act like a faucet--as the voltage on that part was adjusted up and down, it could turn current in the sandwich on and off at will.

Shockley told no one about his idea. The physics behind this amplifier was very different from Bardeen's and Brattain's, since it involved current flowing directly through the chunks of semiconductors, not along the surface. No one was sure if current even could flow right through a semiconductor and possibly Shockley wanted to test it before discussing it. Or possibly he felt that Bardeen and Brattain had "taken" ideas of his for the point-contact transistor and he didn't want to risk that happening again.

The Eureka Moment

Then, on February 18, Shockley learned it could work. Two members of the group, Joseph Becker and John Shive, were working on a separate experiment. Their results could only be explained if the electrons did in fact travel right through the bulk of a semiconductor. When they presented their findings to the group, Shockley knew he had the proof he needed. He jumped up and for the first time shared his concept of a sandwich transistor to the rest of his team.

Bardeen and Brattain were stunned that they hadn't been filled in before now. It was clear that Shockley had been keeping this secret for weeks. It added still more space to the ever-widening gap that was growing between them.
---------------

Telling the Military
June 23, 1949

They had no way of knowing all that the transistor could do, but the administrators at Bell Labs still knew they were on to something big. They were about to hold a huge press conference to announce what they'd invented -- but before telling the public they had to check with the military. At the very least, the transistor could revolutionize communications and radio signals, something that would give the US Army an advantage if the invention was kept a secret from other countries. Bell's president, Mervin Kelly, hoped the army wouldn't want to classify this research, but he knew it just might happen.

On June 23, Ralph Bown gave a presentation to a group of military officers. He showed the way the tiny bit of crystal and wire could amplify an electrical signal much more efficiently than a bulky vacuum tube could. He also told them this was the same demonstration he was preparing to give to the press the next week. What he didn't do was ask permission. Bown and Kelly didn't want to make it easy for the military to classify the transistor. If they wanted to keep it a secret, the army would have to bring up the subject itself.

The armed services went home to their various offices and discussed whether to classify Bell's work. There were certainly those who thought that, at the very least, it should be kept secret until it was better understood just what the transistor could do. But in the end, nobody said a word. Bell Labs went on to its big press conference without a hitch.
---------------

A Working Junction Transistor
1948-1951

There was no doubt about it, point-contact transistors were fidgety. The transistors being made by Bell just didn't work the same way twice, and on top of that, they were noisy. While one lab at Bell was trying to improve those first type-A transistors, William Shockley was working on a whole different design that would eventually get rid of these problems.

Early in 1948, Shockley conceived of a transistor that looked like a sandwich, with two layers of one type of semiconductor surrounding a second kind. This was a completely different setup which didn't have the shaky wires that made the point-contact transistors so hard to control.

Not Just on the Surface

A working sandwich transistor would require that electricity travel straight across a crystal instead of around the surface. But Bardeen's theory about how the point-contact transistor worked said that electricity could only travel around the outside of a semiconductor crystal. In February of 1948, some tentative results in the Shockley lab suggested this might not be true. So the first thing Shockley had to do was determine just what was going on.

Careful experiments led by a physicist in the group, Richard Haynes, helped. Haynes put electrodes on both sides of a thin germanium crystal and took very sensitive measurements of the size and speed of the current. Electricity definitely flowed straight through the crystal. That meant Shockley's vision of a new kind of transistor was theoretically possible.

Growing Crystals

But Haynes also discovered that the layer in the middle of the sandwich had to be very thin and very pure.

The man who paved the way for growing the best crystals was Gordon Teal. He didn't work in Shockley's group, but he kept tabs on what was going on. He'd even been asked to provide crystals for the Solid State team upon occasion. Teal thought transistors should be built from a single crystal-as opposed to cutting a sliver from a larger ingot of many crystals. The boundaries between all the little crystals caused ruts that scattered the current, and Teal had heard of a way to build a large single crystal which wouldn't have all those crags. The method was to take a tiny seed crystal and dip it into the melted germanium. This was then pulled out ever so slowly, as a crystal formed like an icicle below the seed.

Teal knew how to do it, but no one was interested. A number of institutions at the time, Bell included, had a bad habit of not trusting techniques that hadn't been devised at home. Shockley didn't think these single crystals were necessary at all. Jack Morton, head of the transistor-production group, said Teal should go ahead with the research, but didn't throw much support his way.

Luckily, Teal did continue the research, working with engineer John Little. Three months later, in March of 1949, Shockley had to admit he'd been wrong. Current flowing across Teal's semiconductors could last up to one hundred times longer than it had in the old cut crystals.

Growing Even Better Crystals

Nice crystals are all well and good, but a sandwich transistor needed a sandwich crystal. The outer layers had to be a semiconductor with either too many electrons (known as N-type) or too few (known as P-type), while the inner layer was the opposite. Under Shockley's prodding, Teal and Morgan Sparks began adding impurities to the melt while they pulled the crystal out of the melt. Adding impurities is known as "doping," and it's how one turns a semiconductor into N- or P-type.

As they pulled the seed crystal out of an N-type germanium melt, they quickly added some gallium to turn the melt into P-type. As a layer of P-type formed on the ever-lengthening crystal, they added antimony, which compensated for the gallium and turned the melt back into N-type. Once the process was done, there was a single, thin crystal formed into a perfect sandwich.

By etching away the surface of the outside layers, Sparks and Teal left a tiny bit of P-type crystal protruding. To this they attached a fine electrode-creating a circuit the way Shockley had envisioned. On April 12, 1950, they tested what they had built. Without a doubt, more current came out of the sandwich than went in. It was a working amplifier.

The First Junction Transistor

The first junction transistor had been born.

But It Wasn't a Very Good One . . . Yet

This transistor could amplify electrical signals, but not particularly complicated ones. If the signal changed rapidly, as a voice coming over a phone line does, the transistor couldn't keep up and would garble the output. The problem lay in the middle of the sandwich: it was too easy for electric current to spread out and become unfocused as it crossed the P-type layer. To solve the problem, the layer had to be even thinner.

In January of 1951, Morgan Sparks figured out a way to accomplish that. By pulling the crystal out more slowly than ever, while constantly stirring the melt, he managed to get the middle layer of the sandwich thinner than a sheet of paper.

This new, improved sandwich did all that the researchers hoped. They still weren't up to the point-contact transistor's ability to handle signals that fluctuated extremely rapidly, but in every other way they were superior. They were much more efficient, used very little power to work, and they were so much quieter that they could handle weaker signals than the type-A transistors ever could.

In July of 1951, Bell held another press conference -- this time announcing the invention of a working and efficient junction transistor.
---------------

Sharing the Technology: Bell Hosts Transistor Symposia
1951-1952

Bell Labs had an important realization: development of the transistor was going to move a lot more quickly if they opened up the field to other companies. So in September 1951, Bell Labs hosted a symposium to spread the gospel about what the transistor could do.

Attending the conference were some 300 scientists and engineers. The attendees all went home to their respective companies with a great sense of what the transistor could do -- but little idea of how to build one. For that knowledge, Bell announced, a company would have to pay a licensing fee of $25,000. Twenty-six companies, from both the US and abroad, signed up for the privilege. The companies were both big, such as IBM and General Electric, and small, such as then-unknowns like Texas Instruments.

Over one hundred registrants from the select companies returned for the Transistor Technology Symposium in April of 1952. For eight days Bell Labs worked the attendees day and night -- but at the end, they were equipped to go off and build transistors for themselves.

Bell took all the information from the meeting and bound it into a two volume book set called "Transistor Technology." The book became fondly known as "Mother Bell's Cookbook."
----------------

William Shockley Moves to California
1956

William Shockley had gone as far as he was going to go at Bell Labs. He had watched the people underneath him get promoted above him -- and with good reason. Too many top quality scientists hadn't been able to work with him . A genius he may have been, but a good manager he was not.

Shockley decided he needed a big change. The first thing to go was the car -- he traded in the fancy MG for a Jaguar convertible. Next: the job. He spent a semester at Caltech and then a year working for the Weapons Systems Evaluation Group in Washington DC., but nothing completely satisfied him. Eager to be able to run things his own way, he finally decided to strike out on his own -- get some funding and start his own company.

In August of 1955, Shockley flew to LA to spend a week with his new friend Arnold Beckman, a California chemist and businessman. Shockley shared his dream of starting a company to build cutting edge semiconductor devices. Beckman was sold on the idea and agreed to underwrite the venture.

Shockley was lured to the Palo Alto area by Stanford's provost, Fred Terman who thought that a solid research institution in the area would benefit Stanford. With a location picked out, Shockley just had to find the people. He wanted to staff his company with only the best and the brightest. He first sought to employ his colleagues from Bell Labs, but they wouldn't make the jump to the west coast -- or perhaps they couldn't make the jump to working with Shockley again. So Shockley began traveling all over the country recruiting young scientists.

At a lavish luncheon in February of 1956, Shockley and Beckman announced the formation of their brand new lab. They only had four employees at the time, but Shockley Semiconductor Laboratory had officially opened for business. Shockley's was the first company of its kind to settle in the Palo Alto area, but over the years more and more semiconductor labs -- and the computer industries they initiated -- flocked to the area. It wasn't long before the region had earned a new name: Silicon Valley.
---------------

The Future of Transistors

The first announcement of the invention of the transistor met with almost no fanfare. The integrated circuit was originally thought to be useful only in military applications. The microprocessor's investors pulled out before it was built, thinking it was a waste of money. The transistor and its offspring have consistently been under valued -- yet turned out to do more than anyone predicted.
Today's predictions also say that there is a limit to just how much the transistor can do. This time around, the predictions are that transistors can't get substantially smaller than they currently are. Then again, in 1961, scientists predicted that no transistor on a chip could ever be smaller than 10 millionth of a meter -- and on a modern Intel Pentium chip they are 100 times smaller than that.

With hindsight, such predictions seem ridiculous, and it's easy to think that current predictions will sound just as silly thirty years from now. But modern predictions of the size limit are based on some very fundamental physics -- the size of the atom and the electron. Since transistors run on electric current, they must always, no matter what, be at least big enough to allow electrons through.

On the other hand, all that's really needed is a single electron at a time. A transistor small enough to operate with only one electron would be phenomenally small, yet it is theoretically possible. The transistors of the future could make modern chips seem as big and bulky as vacuum tubes seem to us today. The problem is that once devices become that tiny, everything moves according to the laws of quantum mechanics -- and quantum mechanics allows electrons to do some weird things. In a transistor that small, the electron would act more like a wave than a single particle. As a wave it would smear out in space, and could even tunnel its way through the transistor without truly acting on it.

Researchers are nevertheless currently working on innovative ways to build such tiny devices -- abandoning silicon, abandoning all of today's manufacturing methods. Such transistors are known, not surprisingly, as single electron transistors, and they'd be considered "on" or "off" depending on whether they were holding an electron. (Transistors at this level would be solely used as switches for binary coding, not as amplifiers.) In fact, such a tiny device might make use of the quantum weirdness of the ultra-small. The electron could be coded to have three positions -- instead of simply "on" or "off" it could also have "somewhere between on and off. This would open up doors for entirely new kinds of computers. At the moment, however, there are no effective single electron transistors.

Even without new technologies, there's room for miniaturization. By improving on current building techniques, it's likely that current transistors will be at least twice as small by 2010. With nearly a billion transistors on Intel's latest processor that would mean four times as many transistors on a chip are theoretically possible. Chips like this would allow computers to be much "smarter" than they currently are.

Ricardo A Monroy B
C.I 17646658

EES   

---