IF there are signals with frequency content above the Nyquist frequency ( half the sample rate ), AND analog anti-aliasing filtering is not used prior to Analog to Digital Conversion, the signals will appear as signals between DC and the Nyquist frequency ... AND once digitized, the aliased signals cannot be separated nor recognized as being invalid - can be a Major issue.
From my experience anti-alias filtering is usually a compromise and typically you set the sample rate higher than the desired max needed frequency to allow some roll-off without significantly compromising the data of interest.
Modern ADC systems typically sample at their max rate and use anti-alias filters set appropriately high... the data is then Digitally low pass filtered and Decimated to create lower effective sample rates with no analog anti-aliasing filter artifacts...
Hi Jon, As always, great technical article. I'll definitely be waiting for the additional articles on Anti-Alias Filters so I can share this information with my Integrated Circuits class at ITT Tech Institute. Keep the tech articles coming!
Hi, Rob. You can get some unusual results because signals at higher frequencies alias, or fold, into the bandwidth over which you want to make measurements. The alising arises from the sampling of signals at discrete time intervals. Upcoming columns get into more details and explain how to use aliasing to "shift" signals to lower "frequencies" on purpose. Stay tuned. --Jon
Are they robots or androids? We're not exactly sure. Each talking, gesturing Geminoid looks exactly like a real individual, starting with their creator, professor Hiroshi Ishiguro of Osaka University in Japan.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.