ARC - Artificial Neural Network

Overview

A collection of sub-modules for creating and processing CVs, for various boolean logic functions, switching operations and wave shaping. The core are two neuronal networks for combined and sequential logic use which change their function depending on how you patched them (keyword: patch programmable).

The module's idea originates from research on artificial neurons in the fields of cybernetics, artificial intelligence and machine learning.

Views

Details

The ANN consists of a number of unusual and circuits from the field of analog computing and science. For that reason it is advisable to read the user manual for more background information and patch examples.

Boolean Logic Neuron (available twice):
Basically this is a logic AND gate with a number of special fucntions: besides the two inputs A and B (the blue sockets to the left and in the center) there's also the Inhibit input which suppresses the logic result at the output. The left output (red socket) is an AND gate while the other output is an inversion of it and hence is an NAND gate.

Threshold Logic Neuron (available twice)
The most complex sub-module of the ANN works as an artificial neuron based on the "Threshold Logic Input" nervous system model. Three inputs with adjustable (even negative!) weight are summed before passing through an adjustable threshold stage (which also can be negative) which produces a gate when all logic states exceed the threshold.
Plain-talking: depending on the weight, thhreshold and number of used or active inputs the Threshold Logic Neuron can be programmed in a number of ways: as a simple AND, OR or NOT gates with two inputs, as complex logic gates with three inputs and various permutations as well as complex output possibilities. Using feedback they turn into a one-bit memory for sequential use. Combined with the other sub-modules or another Threshold Logic Neuron the possibilities are broadend even more. Complex rhythmicity, gate delays, automated pulse streams and so much more.

Dual Schmitt Trigger:
Two identical Schmitt Trigger circuits in one function block. It is a comparator with a hysteresis, i.e. it works with two different thresholds for turning the gate on or off. The gate (0V or 5V) goes high when the input is higher than 2.6V but it goes low at 2.1V. Any fluctuations in between the two values are ignored.

Comparator:
It compares if the amplitude of the signal at + IN is higher than the amplitude at - IN. If this is the case the gate goes high. The input - IN is pre-patched to +5V.

Dual NOT Gate:
This submodule contains two identical logic inverters. If the input is high, the output will be low and vice versa. For instance an OR gate becomes an NOR and so on. This is not a voltage inverter!

connectors

per Boolean Logic Neuron: AND gate inputs A and B, inhibit input, non-inverted and inverted logic outputs
per Threshold Logic Neuron: three voltage inputs, one logic output
per Schmitt Trigger: voltage input, logic output
comparator: + input, - input, logic output
per NOT gate: logic input, logic output

dimensions

3U Eurorack module, 18HP wide, 25mm deep
current draw: 100mA at +12V and 10mA at -12V

Weight
0.147 kg
Article Number
ARCann
Status: The article can be picked up in the shop right now / article is ready for dispatch
availability ? immediately available

Regular Price: €225.00

Special Price: € 19% VAT excl. €167.23 19% VAT incl. €199.00

Excl. 19% VAT, Excl. shippingLieferzeit 1 Tag