I’m wanting to use the DC-block block, and I have questions about its
use.
My scenario is to use it as a cheap high-pass, with a corner frequency
of a few milli-hertz, with samples arriving at 50Hz.
It’s not clear how the “length” parameter relates to the desired corner
frequency and sample rate.
Another approach I was considering was to use a long-time-constant
single-pole IIR, and subtract the output of that from the unmolested
sample stream.
This is for a correlation interferometer, where we want to eliminate
stability issues caused by low-level interference from satellite signals
that basically show up as slowly-varying DC, at a frequency that’s about
1/10th the fringe frequency.
On Fri, Mar 20, 2015 at 9:45 AM, Marcus D. Leech [email protected]
wrote:
I’m wanting to use the DC-block block, and I have questions about its use.
My scenario is to use it as a cheap high-pass, with a corner frequency of a
few milli-hertz, with samples arriving at 50Hz.
It’s not clear how the “length” parameter relates to the desired corner
frequency and sample rate.
I believe this is implemented as described as a linear phase DC
blocker in this paper:
http://www.digitalsignallabs.com/dcblock.pdf
I believe the “length” parameter is the “D” parameter in the moving
average calculations.
Hopefully this helps.
Brian
On 03/20/2015 10:25 AM, Brian P. wrote:
http://www.digitalsignallabs.com/dcblock.pdf
I believe the “length” parameter is the “D” parameter in the moving
average calculations.
Hopefully this helps.
Brian
Thanks muchly. This helps a lot.
The graph at 4(a) is quite enlightening.