Multichannel Systems MC_Rack Manual - Manual (Page 144)

MC_Rack Manual
138
Bandstop Resonator
In signal processing, a bandstop filter is a filter that passes most frequencies unaltered,
but attenuates those in a specific range to very low levels. A bandstop filter with a high
Q factor has a narrow stopband. Please see the following pictures.
Savitzky Golay
The parameters of the Savitzky-Golay filter are the order (2 or 4) and the number of points which
are included from the right and the left side of a data point for averaging the calculated data
point (2, 4, 8, 16, 32, 48). This filter can be used to smooth a noisy signal. However, it is possible
to accidental filter out fast signals, like spikes.
The y-axis is the normalized amplitude, that is, the output amplitude divided by the input
amplitude. The x-axis represents the frequency of a sinus-wave function. The red line indicates
the input signal amplitude, the purple line indicates the theoretical value at the cutoff frequency,
and the green vertical line indicates the cutoff frequency. The data was obtained using a sine
wave generator and MC_Rack for the measurement. The cutoff frequency was set to 100 Hz.
All filters except of Savitzky Golay are calculated with the help of Fidlib 0.9.10 Copyright
2002-2004 Jim Peters http://uazu.net. This library is released under the GNU Lesser General
Public License (LGPL) version 2.1 as published by the Free Software Foundation.
4.12.3 Filtering and Sampling Rate
A sampling rate of at least five times the highest frequency of the signal of interest is generally
enough for a reasonable representation of the signals. If you are using a broadband MEA1060
amplifier and a digital low pass filter for removing high frequency noise, however, a higher
sampling rate might be necessary.
According to the Nyquist-Shannon sampling theorem, the sampling rate should equal twice the
bandwidth of the analog (hardware) low pass filter. The 1/2 bandwidth frequency is also called
Nyquist frequency. You may ignore this if saving hard disk space is more important for your
application than the noise level.
This is the case because the full amplifier bandwidth is recorded and then high frequency noise is
removed with a digital low pass filter after recording. Frequencies (noise) that are above half the
sampling rate (for example above 2.5 kHz at a 5 kHz sampling rate) will be transformed into lower
frequencies. This is called aliasing. This low frequency noise passes the digital low pass filter and
increases your noise level.
The other possibility to prevent aliasing is to use a MEA1060 amplifier with an appropriate analog
low pass (so-called anti-aliasing) filter instead of the digital filter instrument.