Of course, there are many designs for headphone amplifiers, more or less successful, more simple or more. The design described in this article is simple, it sounds quite good and can be constructed using universally used components.
Currently, if you want to find a headphone amplifier in the store, it is not easy. They do exist, especially in the high-fidelity world, but they match the price tag. The design described here is the following high-end circuits, but can be constructed using easy-to-giving components and still have a relatively good sound quality.
Circuit
The circuit can be described as a power amplifier type, which is built with a separation component (see Figure 1). In the input, we found volume control (P1, through the needle connection) and the coupled capacitor (C1), thereafter thereon has a differential amplifier (T1, T2), a constant current source (T3) in the emitter branch (T3). The preset between T1 and T2 (P2) is used to set symmetry, or in other words, the output voltage is set to 0 V DC (compared to ground). In order to achieve the best sound quality, we should make the two transistors have the same collector flow current. This can be seen at the circuit map voltage test point F and G, both of which are basically the same. The input offset of R1 is caused by the base current flow to T1. This can cause a slight negative voltage at point A (V (a)). The rapid measurement of prototype development shows that the base current of T1 is approximately 3 μA. If the offset compensation is not provided with TrimPot P2, the output offset voltage Vo will exceed 0.2 V:
VO = (1 + R6 / R5) × V (a)
VO = (1 + 10 / 1.5) × 0.028 = 0.215 V
The above image is used for the circuit of the simple headphone amplifier uses components that are easy to order (displayed single channels).
Therefore, it is slightly running by setting the differential amplifier removal. Although this is not the best method, it is really simple to make the circuit in terms of sound quality.
Constant current settings
The current source in the emitter branch (T3) is set to a large about 3 mA, with diodes D1, D2, and resistor R4, which can cause T4 to drive as linear as possible. Then, the audio signal to the driver stage, T4, can drive more powerful output transistors (T6 and T7). The addition of C4 provides greater internal gains. The T5 and R9 output static static current is set to 5 mA. Assume that the output transistor gain (HFE) is 50, theoretically, this 5 mA can provide linear 0.005 a × 50 × 32 Ω = 8 V peak to 32 Ω. However, the constant current source T5 proposes some restrictions and is pressed in the bubble pole wiring T7 (about 1.5 V). We should also take into account the compressors around R11 and R12 (R10 and R12). Then, the maximum voltage Vmax in the load (RL) becomes
Vmax = rl / (RL + R11 + R12) × (9 - 1.5) vmax = 4.6 V peak
This corresponds to approximately 3.26 V Valid values, as we measure, you can view in specifications. This means that the circuit provides 330 mW, 32 Ω performance (3.262 / 32), which is enough to ensure that most popular and rock music fans are excited. Resistor R12, accompanied by an output stage, limiting the output current and holds the circuit stabilization, such as a long shield cable to the headset when the capacitive load is connected. This prevents the output transistor overheating when shorting. R10 and R11 can be kept symmetrical. Although the value of C2 is in the feedback circuit, the bandwidth is still greater than the audio bandwidth (see the specification). To get the input corner frequency, we will use 4.7 μF for C1. 2.2 μF (easier) capacitors still produce 7 Hz acceptable corner frequencies (-0.6 dB, 20 Hz). One measurement of prototype design is shown in the circuit diagram. These can be deemed to be guided value standards rather than accurate needs. Of course, the PN connector and the gain of the transistor are different from the manufacturer (this is also applied to the current consumption provided in the specification).
experiment
For those who don't mind larger noise (although it is silent with most headphones), you can increase the impedance of the feedback loop to approximately 10 kΩ. This can be achieved by adding R5 and R6 in the parallel circuit to 10 kΩ. In this case, the base current of T1 and T2 compensates each other. If you want to perform an experiment, use 12 kΩ resistor to replace R5, and use 68 kΩ resistor to replace R6 (pursue perferable 11.5 kΩ and 76.8 kΩ models). This is impossible to provide audio improvement, but this approach may produce a small offset.
structure
The small printed circuit board used in this circuit (see Figure 2) can be ordered by [1]. From here, you can download the board layout in the PDF format. The component layout is shown in Figure 3. Welcome to the lowest component (resistor, diode), then continue to install higher components (capacitors, transistors, connection fitts). For stereo models, you need two panels, in which case P1 is replaced with stereo potentiometers, so it can control the volume on the two channels simultaneously. If your sound source includes volume control, P1 can be ignored (put the jumper on the needle or welding pin holder 1 and 2 boards, not the actual needle). We recommend the input impedance of the circuit (including P1) minimum 5 kΩ (P1 setting to maximum). This should not be the issue of most modern sound sources. Note the pin spacing of the decoupling capacitor C1; the board provides 5 mm and 7.5 mm models.
You can use two 9V batteries. In addition, there are 2x6 V, 5 VA transformers, with 1.5 A bridge rectifiers and options for 8200 μF / 16 V per supply rail. This can have a set of a pair of voltage regulators. Perhaps, in fact, the output crystal tube (T6 EN T7) does not require a radiator, although the small heat exchanger ensures short circuit protection. We decided to fix this circuit in Elektor ProjectCase. This is easy to do, it provides a unique electronic product appearance and a good view (see Figure 4).
Our other product:
