Introduction
Download this software at http://GuitarFx.Net >>
The sound of rock music, in particular of hard rock and heavy metal, is largely based on a specially distorted guitar sound, for which electronic “distortion” devices, tube amplifiers in “overloaded” mode, computers with appropriate software and digital processors are used. increasingly using neural network algorithms.
The distorted sound of electric guitars began to gain popularity around the 1960s. Since that time, the sound of overloaded tube amplifiers, connected to powerful dedicated guitar speakers with large dedicated speakers, has been considered the benchmark in rock music. But tube amplifiers were relatively expensive and inconvenient to operate. Therefore, semiconductor distortion devices were developed.
At that time, the electrical circuitry of distortion devices was relatively simple and the signal output from their output only vaguely resembled the sound of an overloaded tube amplifier. Nevertheless, it was still somewhat similar to the “sound of a lamp” and this provided a powerful incentive for designers of analog semiconductor distortion circuits to continue their research, complicate circuits and propose new circuit solutions. The heyday of analog solid-state distortion was around 1995-2010. The most popular were electrical circuit diagrams like those shown in the figure below.
Subsequently, the distortion circuits became somewhat more complicated. To soften the clipping of the signal and more like a smooth clipping in the amplifier tubes, the number of diodes increased and their number was often different for the positive and negative half-waves of the signal.
Not entirely successful (but not failed) attempts were made to emulate tube distortion on field-effect transistors as in the diagram below.
In the above phase, progress in electrical circuitry for electric guitar distortion has slowed down significantly. Unfortunately, electronics engineers have not been able to develop high-precision transistor, diode, and op-amp simulators of tubes and tube amplifiers. Probably the most notable pinnacle in this direction is the SansAmp GT2 Tech 21 NYC. The circuitry of this device contains an input high-pass filter, several op-amps, an equalizer-type filter and output stages.
The main idea of SansAmp is overload without diodes in the OOS of the op-amp and without shunt diodes, but right inside the op-amps, possibly on field-effect transistors inside the op-amp. It is difficult to find a 100% reliable diagram of this device on the Internet. And most importantly, for this circuit, the op-amp overload and the output signal critically strongly depend on the type of op-amp used. In fact, radio amateurs have not been able to fully achieve the sound quality of the original SansAmp and solder the full analog sound. This was probably due to the lack of the very old op-amps on field-effect transistors used in the original SansAmp on the free market.
The incompleteness of the analog age and promising ideas for digital modeling of distortion circuits
Digital software modeling of existing guitar amplifiers and effects has received a very perfect solution in the form of processors on neural networks (DSP). All significant and legendary guitar amplifiers, as well as guitar speakers and microphones received good adequate models, worthy of confronting the originals in a blind listening test. The DSP-generated sound of distorted electric guitars is now difficult to distinguish from the sound of real tube amplifiers. However, it is likely that digital processors with software signal processing have retired many developers of semiconductor distortion devices a little early, and many interesting projects from a schematic point of view remain unfinished.
But creative engineering is willing to move on. The fact is that it seems problematic to develop something fundamentally new in sound or significantly improve the sound of existing guitar amplifiers using neural networks, since a neural network needs a training sequence of signals, which means it is necessary to create these signals by some kind of prototype from an undistorted signal the sound of an electric guitar.
Ham radio enthusiasts are trying to invent similar distortion prototypes with more interesting sound than standard branded tube guitar amplifiers. There is a demand, because, as a rule, rock band guitarists want an original powerful sound. There are good conditions for new developments in the field of electric guitar signal processing. Many significant and powerful analog semiconductor digital simulations are free. There are large archives with distortion circuits, tube guitar amplifiers, reviews of radio amateurs who soldered these circuits, sound samples of various circuits. All this, probably, allows you to analyze and rethink the experience of the analog era and, on this basis, come up with, quickly simulate in digital and hear how a new guitar effect will sound.
An interesting, but little-known, not fully explored direction is multi-band and multi-stage distortion. There are very few analog implementations due to the complexity of manufacturing and customization. There are “software” (ie software) implementations. They, as a rule, are not analog prototype models, but are simplified software implementations of digital signal processing algorithms (DSP or DSP). The sound of such implementations is not very convincing. They do not produce enough “fat” distortion of the electric guitar signal and it looks more like a soft “overdrive”. But the idea itself is beautiful.
A diode distortion circuit with a floating bias, simulating the grid current of overloaded tubes of the output stage of a guitar amplifier, is being actively studied. Its output sound is reminiscent of overdriven tubes, and in some respects it can even compete with the overdriven tube output stages of guitar amps. A worthy idea.
The original ideas of the principles of distortion functioning may be found on the forums of radio amateurs and “home-made” electronic devices. For 2020..2021, more than 700 posts were found. Often the discussion was conducted at a high professional level with samples, circuit models on LTSpice, Microcap, Native Instrumets, TINA TI, VST and real oscillograms from real overloaded stages of tube guitar amplifiers.
Digital modeling of distortion circuits as a creative process
The free TINA TI (Texas Instrumets) is considered a good modeling system. The main window of this program is a virtual artboard. The developer assembles a circuit from virtual resistors (resistances), capacitors, transformers, chokes, diodes, transistors, lamps, operational amplifiers, power supplies, signal generators and other elements. It is possible to send a signal recorded in a regular wav file to the input of a virtual circuit, for example, an electric guitar signal recorded through a computer sound card.
From the output of the virtual circuit, the processed signal is output to a wav file, to a virtual oscilloscope or directly to a sound card and can be listened to through headphones or speakers. Several virtual oscilloscopes and voltmeters can be connected to all nodes of the circuit and the developer can easily customize the circuit according to his ideas, setting the required voltages by changing the values of virtual resistors (resistances) and other circuit elements.
A new idea of signal processing is required from the developer itself, i.e. creative idea. The implementation of this idea in the form of a virtual circuit and its testing on real guitar signals is carried out by assembling the circuit in TINA TI from ready-made virtual elements without low-level programming of digital signal processing algorithms.
A successful virtual circuit is likely to have commercial value if it is transformed into a DSP, VST module loaded into a hardware digital guitar processor, or fabricated as a real analog device for promotion through conventional electronic music equipment sales channels.
