Digital Signal Processors (DSPs) consist of customized hardware to efficiently manipulate sequences of numbers or symbols that comprise a signal used in a wide variety of applications including digital communications and scientific instrumentation. DSPs are used in applications where discrete, raw signals can be captured and turned into digital data used for real-time analysis. DSPs are capable of processing either fixed point numeric data, or floating point data. Applications where DSPs are deployed include industrial automation, scientific instrumentation, medical instrumentation, scientific instrument, and test and measurement. To facilitate development of applications using DSP, Critical Link offers the MityDSP, a family of highly-configurable, small form factor System on Modules (SoMs). The MityDSP combines a TI DSP and a Xilinx FPGA, and comes in a range of configurations to meet varying processing requirements. (There is also a MityDSP engine that includes a TI OMAP containing both a DSP and ARM processors alongside FPGA.)
More on DSP
DSPs are used to measure, filter and/or compress analog signals. DSPs are commonly used because of their small form factor, relatively low power consumption, and computational speed. The computational power of DSPs often makes digital signal processing preferable to analog processing, even given the overhead required for analog to digital conversion (ADC). Using real-world analog signals, a DSP clarifies or standardizes the levels or states of these signals to eliminate chaotic noise. For example, an incoming signal may be analog transmitted via a standard television broadcast station. An ADC converts the television broadcast analog signal into a number stream or digital signal. The incoming digital signal, representing either voltage or current measurements (or both), may contain noise preventing the standard value signal transmission. The DSP can be used to process the digital stream of data for a variety of purposes, including noise reduction, power amplification, or signal recovery. A digital-to-analog (DAC) converter will, as needed, convert the digital signal back to analog.
Digital signals generally fall into one of the following domains: time, spatial (multidimensional signals), frequency, autocorrelation, or wavelet.