A POWERFUL AMPLIFIER NAMED DIRTY HARRY
The big module presented in the Photo1 below, come from a general circuit suitable for building amplifiers in any power level. This concrete benefit of the circuit presented in Fig.1 owed to the extensive use of constant current sources (CCS) in particular thermal compensated. The CCS ensures that, with slight modifications in some passive elements such resistors we can obtain a greater power output by simply increasing the power supply level and substituting the type of transistors with those of higher Vce and manipulating the number of output devices to sustain this supply level. The type of architecture used in this amplifier, after the spreading of Douglas Self book "Audio Power Amplifier Design Handbook", most people like to call it "Blameless" a term established by the author of this book. I must to point out that the only novelty come from the D. Self propositions which is used in this circuit, it is the emitter follower before the voltage amplifier stage (VAS) which abolish the previous used complex cascode topology in this place. The rest topologies of circuit come from different sources such as I.C. constructors, power amplifier manufacturers etc. in reality. Before 14 years i constructed P.A. type amplifiers without all of the novelties included in the present circuit. Due to circumstances (driving of big loudspeakers equipped with hard suspension system and heavy voice coil and loaded with large horns) always my priority was the good transient response of the amplifier, because this is the main audible feature of such systems. This was my philosophy until now. With simple words it is the killing bass reproduction. And this is based in the transient response of an amplifier exactly in low frequencies. Afterwards i decided to build a more stable model without the loss of this beautiful characteristic. I remembered that, my previous amplifier had a rise time in 10 KHz about 1,7μsec. This was the main target which will be drive me the next years to obtain more stable circuits with little as possible deviation from this rise time. Finally, all of my present models have a rise time from 1,8 to 2 μsec as much. And proved that, the evolution of my previous used architecture in a most modern for today it has little aggravation in the transient response of amplifiers.