gavinbirss
Head-Fier
- Joined
- Oct 19, 2001
- Posts
- 83
- Likes
- 11
DISCRETE PREAMPLIFIER
Quality-conscious audio buffs still prefer discrete designs. And quite rightly so, because although there are very good operational amplifiers available, discrete designs offer just that little bit extra.
The present preamplifier is a symmetrical Class A design. The input is a double differential amplifier consisting of dual transistors Type MAT02 or MAT03. A stable d.c. operating point is ensured by current sources T3 and T4 which use LEDs as reference – D1 and D2 respectively. The current through the LEDs is held stable by current source T5. It is essential for good thermal stability that the transistors and associated diodes (T3 and D1 and T4 and D2 are mounted In close contact.) The input signals are applied to push-pull drivers T6 and T7 which feed the output stages, consisting of emitter followers T10 and T11. Transistors T8 and T9 ensure a constant quiescent current through the emitter followers. It is necessary for good thermal stability that T8 and T10, and T9 and T10 are mounted in close contact. To this end, their flat sides, with heat conducting paste in between, are juxtaposed. The pairs are held together with a loop of bare copper wire.
Before the mains is switched on, set P1 to maximum resistance. Switch on the mains, wait for about a minute and then adjust P for a quiescent current through T10 and T11 of 15 mA, corresponding to a voltage drop of 150 mV across R23 and R24.
Since the amplifier is d.c. coupled throughout, the likelihood of a fairly high direct voltage at the output would be great, the more so because the input transistors are not truly complementary. This is, however, obviated by an active d.c. correction that holds the direct voltage at the output at zero in all circumstances. For this purpose, the output signal is passed via low-pass filter R26-C13 to integrator IC1. This arrangement does not affect fast variations of the signal. If, however, the output signal has a d.c. component. T12 will conduct to some degree, so that the bases of T1 and T2 are pulled into a negative direction. In a negative direction, because T1 (n-p-n) has an inherently greater voltage amplification (x3) than T2 (p-n-p).
Adjust P Immediately on switch-on for as low a direct voltage at the output as possible. From then on, any variations caused by temperature changes will be corrected by IC1. The speed at which the correction takes place can be increased by giving R26 and R27 lower values.
It is important for optimum symmetry that the currents through T1 and T2 (and thus the I voltage drops across R9 and R10 are equal. This can only be if the potentials across and D1 and D2 are equal, and it is, therefore, advisable to match these diodes for equal voltage with a test current through them of 3 mA. When the diodes are matched, the drops across R13 and R14 should not differ by more than a few millivolts.
The same applies to T6 and T7 for good symmetry they should be matched for equal base/emitter voltage, with a current through them of 5 mA. This matching can not be done in the circuit, because the voltage drops across R17 and R18 will be equal whatever, other wise the output would not be zero.
Low-pass filter R2-C2 is designed for maximum slew rate at a cut-off point of 9-1O MHz. If this large bandwidth results in high sensitivity to interference, it may be advisable to lower the cut-off point. If the value of C2 is increased to 680 pF, the cut-off point drops to about 400 kHz. At the same time, the slew rate deteriorates to about 20 V us.
The preamplifier is best built on the PCB in Fig. 2, which is available ready-made.
The supply lines should be stabilized by a suitable voltage regulator.
Fig. 1. Circuit diagram of the discrete preamplifier
here is a link to the html document :
discrete preamplifier
Sorry about the banner ads.
Gavin
Quality-conscious audio buffs still prefer discrete designs. And quite rightly so, because although there are very good operational amplifiers available, discrete designs offer just that little bit extra.
The present preamplifier is a symmetrical Class A design. The input is a double differential amplifier consisting of dual transistors Type MAT02 or MAT03. A stable d.c. operating point is ensured by current sources T3 and T4 which use LEDs as reference – D1 and D2 respectively. The current through the LEDs is held stable by current source T5. It is essential for good thermal stability that the transistors and associated diodes (T3 and D1 and T4 and D2 are mounted In close contact.) The input signals are applied to push-pull drivers T6 and T7 which feed the output stages, consisting of emitter followers T10 and T11. Transistors T8 and T9 ensure a constant quiescent current through the emitter followers. It is necessary for good thermal stability that T8 and T10, and T9 and T10 are mounted in close contact. To this end, their flat sides, with heat conducting paste in between, are juxtaposed. The pairs are held together with a loop of bare copper wire.
Before the mains is switched on, set P1 to maximum resistance. Switch on the mains, wait for about a minute and then adjust P for a quiescent current through T10 and T11 of 15 mA, corresponding to a voltage drop of 150 mV across R23 and R24.
Since the amplifier is d.c. coupled throughout, the likelihood of a fairly high direct voltage at the output would be great, the more so because the input transistors are not truly complementary. This is, however, obviated by an active d.c. correction that holds the direct voltage at the output at zero in all circumstances. For this purpose, the output signal is passed via low-pass filter R26-C13 to integrator IC1. This arrangement does not affect fast variations of the signal. If, however, the output signal has a d.c. component. T12 will conduct to some degree, so that the bases of T1 and T2 are pulled into a negative direction. In a negative direction, because T1 (n-p-n) has an inherently greater voltage amplification (x3) than T2 (p-n-p).
Adjust P Immediately on switch-on for as low a direct voltage at the output as possible. From then on, any variations caused by temperature changes will be corrected by IC1. The speed at which the correction takes place can be increased by giving R26 and R27 lower values.
It is important for optimum symmetry that the currents through T1 and T2 (and thus the I voltage drops across R9 and R10 are equal. This can only be if the potentials across and D1 and D2 are equal, and it is, therefore, advisable to match these diodes for equal voltage with a test current through them of 3 mA. When the diodes are matched, the drops across R13 and R14 should not differ by more than a few millivolts.
The same applies to T6 and T7 for good symmetry they should be matched for equal base/emitter voltage, with a current through them of 5 mA. This matching can not be done in the circuit, because the voltage drops across R17 and R18 will be equal whatever, other wise the output would not be zero.
Low-pass filter R2-C2 is designed for maximum slew rate at a cut-off point of 9-1O MHz. If this large bandwidth results in high sensitivity to interference, it may be advisable to lower the cut-off point. If the value of C2 is increased to 680 pF, the cut-off point drops to about 400 kHz. At the same time, the slew rate deteriorates to about 20 V us.
The preamplifier is best built on the PCB in Fig. 2, which is available ready-made.
The supply lines should be stabilized by a suitable voltage regulator.
Fig. 1. Circuit diagram of the discrete preamplifier
here is a link to the html document :
discrete preamplifier
Sorry about the banner ads.
Gavin