Any techniques for obtaining a 'best fit' digital filter (1st choice FIR, 2nd choice IIR) for an analog filter. The analog filter specifications be from an equation (e.g. butterworth etc) or a set of measured frequency response values. This will allow me to simulate digitally an analog process.
Simulating analog filter with FIR
Started by ●September 18, 2006
Reply by ●September 18, 20062006-09-18
Peterpan wrote:> Any techniques for obtaining a 'best fit' digital filter (1st choice FIR, > 2nd choice IIR) for an analog filter. The analog filter specifications be > from an equation (e.g. butterworth etc) or a set of measured frequency > response values. > This will allow me to simulate digitally an analog process.The most straightforward ways are the BLT and the impulse invariant design. Just make sure the sample rate is at least 4 times higher then the highest frequency of interest. If it is not, close matching of the analog prototype becomes a really complicated task. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
Reply by ●September 18, 20062006-09-18
Peterpan wrote:> Any techniques for obtaining a 'best fit' digital filter (1st choice FIR, > 2nd choice IIR) for an analog filter. The analog filter specifications be > from an equation (e.g. butterworth etc) or a set of measured frequency > response values. > This will allow me to simulate digitally an analog process.P.S. Building the digital filter by the analog prototype and the simulation of the analog process are the two very different things. If your goal is a filter, Google "BLT". If your goal is the simulation, Google "Spice". Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
Reply by ●September 18, 20062006-09-18
On Mon, 18 Sep 2006 14:21:41 GMT, Peterpan <xxxx@uun.com> wrote:>Any techniques for obtaining a 'best fit' digital filter ... for an >analog filter.http://groups.google.com/group/comp.dsp/browse_thread/thread/501b31898ae74f47/79dfa6d61e95c6b7 Greg Berchin
Reply by ●September 19, 20062006-09-19
"Peterpan" <xxxx@uun.com> wrote in message news:Xns9842551001B93dnyarkoinamecom@140.99.99.130...> Any techniques for obtaining a 'best fit' digital filter (1st choice FIR, > 2nd choice IIR) for an analog filter. The analog filter specifications be > from an equation (e.g. butterworth etc) or a set of measured frequency > response values. > This will allow me to simulate digitally an analog process.Google Parks McClellan HTH, Syms.
Reply by ●September 20, 20062006-09-20
Greg Berchin skrev:> On Mon, 18 Sep 2006 14:21:41 GMT, Peterpan <xxxx@uun.com> wrote: > > >Any techniques for obtaining a 'best fit' digital filter ... for an > >analog filter. > > http://groups.google.com/group/comp.dsp/browse_thread/thread/501b31898ae74f47/79dfa6d61e95c6b7 > > Greg BerchinGreg, Thanks for pointing out a very interesting thread that slipped past me at the time. Too bad your first attempts at publishing this method did not come through; some times one wonders what's on the minds of journal editors... oh well. How is that SPM article coming? It would be a very interesting one, for sure. Rune
Reply by ●September 20, 20062006-09-20
On 20 Sep 2006 01:02:36 -0700, "Rune Allnor" <allnor@tele.ntnu.no> wrote:>How is that SPM article coming? It would be a very interesting one, >for sure.The article is complete, and will be the subject of an upcoming episode of "Tips & Tricks". I have Matlab code that implements the algorithm. Anybody may email me privately for a copy. (Remove the "w" from my email address.) Greg
Reply by ●September 22, 20062006-09-22
On 20 Sep 2006 01:02:36 -0700, "Rune Allnor" <allnor@tele.ntnu.no> wrote:> >Greg Berchin skrev: >> On Mon, 18 Sep 2006 14:21:41 GMT, Peterpan <xxxx@uun.com> wrote: >> >> >Any techniques for obtaining a 'best fit' digital filter ... for an >> >analog filter. >> >> http://groups.google.com/group/comp.dsp/browse_thread/thread/501b31898ae74f47/79dfa6d61e95c6b7 >> >> Greg Berchin > >Greg, > >Thanks for pointing out a very interesting thread that slipped past me >at the time. Too bad your first attempts at publishing this method >did not come through; some times one wonders what's on the minds >of journal editors... oh well. > >How is that SPM article coming? It would be a very interesting one, >for sure. > >RuneHi Rune (& Greg), to elaborate a little on Greg's phrase "an upcoming episode of "Tips & Tricks", Greg's article will appear in the 'DSP Tips & Tricks' column of January 2007 issue of the IEEE Signal Processing magazine. See Ya', [-Rick-]