音箱音频放大器电路 (Speaker Box Audio Amp)Design PhilosophyI did this design after getting a new computer. After buying the processor, monitor, and printer, I wasn"t willing to spring for a set of speakers too. After going "soundless", I decided to add speakers. Of course, this was the perfect excuse warm up the soldering iron to try out a new design. This original design is a variation on a well-known design, examples of which can be found in a great many texts. My variation was to add a second differential stage to replace the usual common emitter-plus-constant current source. Doing so opens up a second inverse feedback path. The signal path being a series connection of the differential amp inverting inputs, and the feedback being the non-inverting inputs. To this end, the constant current source does not include bypass capacitors to render it a DC-only stage. The other feedback path is the usual one: output to the non-inverting input of the first differential stage. The idea is to add lots of inverse feedback in order to linearize the output transistors, as the non-linearity of these devices makes them rather poor analog amplifiers.click on image for higher resolutionConstruction: Power SupplyThe power supply is simply a straight forward symmetrical plus/minus supply constructed from a transformer with a balanced secondary and a bridge rectifier module. Even though the transformer was a self-wound unit, any transformer with a 25.2VCT secondary at 2.0A will be a good substitute. Such transformers oughht to be available as VT "heater transformers" for high power "ham" RF amplifiers. It is essential that the primary side of the transformer be connected exactly as shown. This PS uses a MOV for over voltage protection. These things have a nasty habit of not turning off once they turn on. The fuse needs to be connected between this MOV and the mains. Don"t attempt to "go cheap (or lazy)" and fail to include it. A 2.0A, "slow blow" type works well here. It will stand up to the initial surge as the filter capacitors charge up, and will prevent damage or worse should the MOV be triggered. This supply is somewhat over rated for the application, and is heavily bled for good voltage regulation. There is still some 120Hz hum that could be eliminated with a more involved PS circuit. However, as this wasn"t intended as a "premium" amp, that wasn"t necessary. The hum is most noticeable when running with no input signal. While actually playing sound, it"s no longer noticeable. If you want to substitute a better supply, then that"s OK, too.ImprovementsThese units worked quite well, and do sound good, despite the inherent limitations of BJTs as analog amplifiers. There are a couple of improvements possible. The first would be the use of a full-complimentary final. If this is done, then any complimentary BJTs with a PC of 5.0W or more will work. Again, the most important part of this will remain the input Darlingtons. Even better would be the substitution of complimentary (N-Channel/P-Channel) VMOSFET power transistors. These, of course, don"t require Darlington pairs in order to operate. If VMOS power transistors are used, it"s important to include a 100 ohm resistor in series with the gate terminal, mounted close to the transistor body with the shortest lead length. This will help prevent the possibility of RF oscillation. (Unlike the BJT, the VFET is a high frequency device, and the inadvertant construction of VHF oscillators with these is quite easy.)Parts List: AmplifierCR1 -- 6: 1N914Q1: 2N3819NPN: 2N3904 (or other small signal NPN)PNP: 2N3906 (or other small signal PNP)Q8: 2N3904 (Critical)Q10: 2N3906 (Critical)Q9, 11: 2N3053R: 1.0K/¼W miniature pot.Parts List: Power SupplyC1 -- 2: 20000uF/35WVDCCR1 -- 4: 50VPRV/5.0A Bridge ModuleCR5: Green LED Panel MountF1: 2.0A, Slow Blow FuseT1: PRI: 120VRMS SEC: 25.2VCT/2.0AZ1: 120VRMS MOV