passive rc-bandpass with 0dB attenuation in the passband
in [Design]
|
Prev: Peak value
Next: NimH AAA to 5V, suggestions?
From: Tim Wescott on 3 Aug 2010 13:35 On 08/03/2010 10:23 AM, cassiope wrote: > On Aug 2, 1:35 pm, Tim Wescott<t...(a)seemywebsite.com> wrote: >> On 08/02/2010 01:06 PM, nukeymusic wrote: >> >>> On Aug 2, 7:04 pm, cassiope<f...(a)u.washington.edu> wrote: >>>> On Aug 2, 7:50 am, nukeymusic<nukeymu...(a)gmail.com> wrote: >> >>>>> Is it possible to make a passive rc-bandpass filter which has 0dB >>>>> attenuation in the passband with only 4 components? >> >>>>> nukey >> >>>> How close to 0dB? What sort of Q (bw/f0)? What's the load? >> >>> exactly 0dB, unloaded, Q to be determined from the other specifications >> >> Q is almost meaningless in this case -- any passive RC bandpass filter >> is going to have a damping ratio greater than 1, and the various >> definitions of Q only converge for damping ratios much less than one. >> >> -- >> >> Tim Wescott >> Wescott Design Serviceshttp://www.wescottdesign.com >> >> Do you need to implement control loops in software? >> "Applied Control Theory for Embedded Systems" was written for you. >> See details athttp://www.wescottdesign.com/actfes/actfes.html > > Ok, so I have used a more generalized definition of Q (bw/f0). > Fortunately, > definitions rarely have convergence problems. Applying them in > specific instances > is another matter ;) > > The simpleminded 4-component RC filter (the obvious serial-parallel > arrangement) > won't get to exactly 0dB... would only approach it for truly wide > bandwidths, even with > no load. Yes, I cited that circuit to contradict Jim's statement that "you can't make a passive RC bandpass circuit", not to answer the OP's question. As I told the OP, I do dimly remember a passive RC circuit that had voltage (but certainly not power!) gain at some frequency. But I don't think it was a bandpass, and I don't know if one could build a bandpass version of it. If you could design a bandpass, passive, RC circuit with gain, then it would be a snap to design one with a 0dB peak. Realizing exactly a 0dB peak with component tolerances and all would be more of a trick, I'm sure. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" was written for you. See details at http://www.wescottdesign.com/actfes/actfes.html
From: Jim Thompson on 3 Aug 2010 14:19 On Tue, 03 Aug 2010 10:35:41 -0700, Tim Wescott <tim(a)seemywebsite.com> wrote: >On 08/03/2010 10:23 AM, cassiope wrote: >> On Aug 2, 1:35 pm, Tim Wescott<t...(a)seemywebsite.com> wrote: >>> On 08/02/2010 01:06 PM, nukeymusic wrote: >>> >>>> On Aug 2, 7:04 pm, cassiope<f...(a)u.washington.edu> wrote: >>>>> On Aug 2, 7:50 am, nukeymusic<nukeymu...(a)gmail.com> wrote: >>> >>>>>> Is it possible to make a passive rc-bandpass filter which has 0dB >>>>>> attenuation in the passband with only 4 components? >>> >>>>>> nukey >>> >>>>> How close to 0dB? What sort of Q (bw/f0)? What's the load? >>> >>>> exactly 0dB, unloaded, Q to be determined from the other specifications >>> >>> Q is almost meaningless in this case -- any passive RC bandpass filter >>> is going to have a damping ratio greater than 1, and the various >>> definitions of Q only converge for damping ratios much less than one. >>> >>> -- >>> >>> Tim Wescott >>> Wescott Design Serviceshttp://www.wescottdesign.com >>> >>> Do you need to implement control loops in software? >>> "Applied Control Theory for Embedded Systems" was written for you. >>> See details athttp://www.wescottdesign.com/actfes/actfes.html >> >> Ok, so I have used a more generalized definition of Q (bw/f0). >> Fortunately, >> definitions rarely have convergence problems. Applying them in >> specific instances >> is another matter ;) >> >> The simpleminded 4-component RC filter (the obvious serial-parallel >> arrangement) >> won't get to exactly 0dB... would only approach it for truly wide >> bandwidths, even with >> no load. > >Yes, I cited that circuit to contradict Jim's statement that "you can't >make a passive RC bandpass circuit", not to answer the OP's question. With horrible skirts. "Band-pass" usually implies skirt-rate relative to bandwidth. > >As I told the OP, I do dimly remember a passive RC circuit that had >voltage (but certainly not power!) gain at some frequency. But I don't >think it was a bandpass, and I don't know if one could build a bandpass >version of it. If you could design a bandpass, passive, RC circuit with >gain, then it would be a snap to design one with a 0dB peak. Realizing >exactly a 0dB peak with component tolerances and all would be more of a >trick, I'm sure. ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | Spice is like a sports car... Performance only as good as the person behind the wheel.
From: Tim Wescott on 3 Aug 2010 14:54 On 08/03/2010 11:19 AM, Jim Thompson wrote: > On Tue, 03 Aug 2010 10:35:41 -0700, Tim Wescott<tim(a)seemywebsite.com> > wrote: > >> On 08/03/2010 10:23 AM, cassiope wrote: >>> On Aug 2, 1:35 pm, Tim Wescott<t...(a)seemywebsite.com> wrote: >>>> On 08/02/2010 01:06 PM, nukeymusic wrote: >>>> >>>>> On Aug 2, 7:04 pm, cassiope<f...(a)u.washington.edu> wrote: >>>>>> On Aug 2, 7:50 am, nukeymusic<nukeymu...(a)gmail.com> wrote: >>>> >>>>>>> Is it possible to make a passive rc-bandpass filter which has 0dB >>>>>>> attenuation in the passband with only 4 components? >>>> >>>>>>> nukey >>>> >>>>>> How close to 0dB? What sort of Q (bw/f0)? What's the load? >>>> >>>>> exactly 0dB, unloaded, Q to be determined from the other specifications >>>> >>>> Q is almost meaningless in this case -- any passive RC bandpass filter >>>> is going to have a damping ratio greater than 1, and the various >>>> definitions of Q only converge for damping ratios much less than one. >>>> >>>> -- >>>> >>>> Tim Wescott >>>> Wescott Design Serviceshttp://www.wescottdesign.com >>>> >>>> Do you need to implement control loops in software? >>>> "Applied Control Theory for Embedded Systems" was written for you. >>>> See details athttp://www.wescottdesign.com/actfes/actfes.html >>> >>> Ok, so I have used a more generalized definition of Q (bw/f0). >>> Fortunately, >>> definitions rarely have convergence problems. Applying them in >>> specific instances >>> is another matter ;) >>> >>> The simpleminded 4-component RC filter (the obvious serial-parallel >>> arrangement) >>> won't get to exactly 0dB... would only approach it for truly wide >>> bandwidths, even with >>> no load. >> >> Yes, I cited that circuit to contradict Jim's statement that "you can't >> make a passive RC bandpass circuit", not to answer the OP's question. > > With horrible skirts. "Band-pass" usually implies skirt-rate relative > to bandwidth. Yes. It's more of a "scholar's bandpass" than anything you might want to use in real life. It's not a bad mental tool to use when cooking up an active filter, because an active bandpass can be made by "sharpening up" a passive one, ditto with an active lowpass and (with due respect for stability) an active highpass. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" was written for you. See details at http://www.wescottdesign.com/actfes/actfes.html
From: nukeymusic on 4 Aug 2010 05:20 On Aug 3, 8:54 pm, Tim Wescott <t...(a)seemywebsite.com> wrote: > On 08/03/2010 11:19 AM, Jim Thompson wrote: > > > > > On Tue, 03 Aug 2010 10:35:41 -0700, Tim Wescott<t...(a)seemywebsite.com> > > wrote: > > >> On 08/03/2010 10:23 AM, cassiope wrote: > >>> On Aug 2, 1:35 pm, Tim Wescott<t...(a)seemywebsite.com> wrote: > >>>> On 08/02/2010 01:06 PM, nukeymusic wrote: > > >>>>> On Aug 2, 7:04 pm, cassiope<f...(a)u.washington.edu> wrote: > >>>>>> On Aug 2, 7:50 am, nukeymusic<nukeymu...(a)gmail.com> wrote: > > >>>>>>> Is it possible to make a passive rc-bandpass filter which has 0dB > >>>>>>> attenuation in the passband with only 4 components? > > >>>>>>> nukey > > >>>>>> How close to 0dB? What sort of Q (bw/f0)? What's the load? > > >>>>> exactly 0dB, unloaded, Q to be determined from the other specifications > > >>>> Q is almost meaningless in this case -- any passive RC bandpass filter > >>>> is going to have a damping ratio greater than 1, and the various > >>>> definitions of Q only converge for damping ratios much less than one.. > > >>>> -- > > >>>> Tim Wescott > >>>> Wescott Design Serviceshttp://www.wescottdesign.com > > >>>> Do you need to implement control loops in software? > >>>> "Applied Control Theory for Embedded Systems" was written for you. > >>>> See details athttp://www.wescottdesign.com/actfes/actfes.html > > >>> Ok, so I have used a more generalized definition of Q (bw/f0). > >>> Fortunately, > >>> definitions rarely have convergence problems. Applying them in > >>> specific instances > >>> is another matter ;) > > >>> The simpleminded 4-component RC filter (the obvious serial-parallel > >>> arrangement) > >>> won't get to exactly 0dB... would only approach it for truly wide > >>> bandwidths, even with > >>> no load. > > >> Yes, I cited that circuit to contradict Jim's statement that "you can't > >> make a passive RC bandpass circuit", not to answer the OP's question. > > > With horrible skirts. "Band-pass" usually implies skirt-rate relative > > to bandwidth. > > Yes. It's more of a "scholar's bandpass" than anything you might want > to use in real life. It's not a bad mental tool to use when cooking up > an active filter, because an active bandpass can be made by "sharpening > up" a passive one, ditto with an active lowpass and (with due respect > for stability) an active highpass. > > -- > > Tim Wescott > Wescott Design Serviceshttp://www.wescottdesign.com > > Do you need to implement control loops in software? > "Applied Control Theory for Embedded Systems" was written for you. > See details athttp://www.wescottdesign.com/actfes/actfes.html Dear Tim, You probably did not see my question in the other message, therefore I'd like to ask it here once more: Can you prove the following statement or point to a proof for it: any passive RC bandpass filter is going to have a damping ratio greater than 1, regards, nukey
From: Tim Wescott on 4 Aug 2010 11:01
On 08/04/2010 02:20 AM, nukeymusic wrote: > On Aug 3, 8:54 pm, Tim Wescott<t...(a)seemywebsite.com> wrote: >> On 08/03/2010 11:19 AM, Jim Thompson wrote: >> >> >> >>> On Tue, 03 Aug 2010 10:35:41 -0700, Tim Wescott<t...(a)seemywebsite.com> >>> wrote: >> >>>> On 08/03/2010 10:23 AM, cassiope wrote: >>>>> On Aug 2, 1:35 pm, Tim Wescott<t...(a)seemywebsite.com> wrote: >>>>>> On 08/02/2010 01:06 PM, nukeymusic wrote: >> >>>>>>> On Aug 2, 7:04 pm, cassiope<f...(a)u.washington.edu> wrote: >>>>>>>> On Aug 2, 7:50 am, nukeymusic<nukeymu...(a)gmail.com> wrote: >> >>>>>>>>> Is it possible to make a passive rc-bandpass filter which has 0dB >>>>>>>>> attenuation in the passband with only 4 components? >> >>>>>>>>> nukey >> >>>>>>>> How close to 0dB? What sort of Q (bw/f0)? What's the load? >> >>>>>>> exactly 0dB, unloaded, Q to be determined from the other specifications >> >>>>>> Q is almost meaningless in this case -- any passive RC bandpass filter >>>>>> is going to have a damping ratio greater than 1, and the various >>>>>> definitions of Q only converge for damping ratios much less than one. >> >>>>>> -- >> >>>>>> Tim Wescott >>>>>> Wescott Design Serviceshttp://www.wescottdesign.com >> >>>>>> Do you need to implement control loops in software? >>>>>> "Applied Control Theory for Embedded Systems" was written for you. >>>>>> See details athttp://www.wescottdesign.com/actfes/actfes.html >> >>>>> Ok, so I have used a more generalized definition of Q (bw/f0). >>>>> Fortunately, >>>>> definitions rarely have convergence problems. Applying them in >>>>> specific instances >>>>> is another matter ;) >> >>>>> The simpleminded 4-component RC filter (the obvious serial-parallel >>>>> arrangement) >>>>> won't get to exactly 0dB... would only approach it for truly wide >>>>> bandwidths, even with >>>>> no load. >> >>>> Yes, I cited that circuit to contradict Jim's statement that "you can't >>>> make a passive RC bandpass circuit", not to answer the OP's question. >> >>> With horrible skirts. "Band-pass" usually implies skirt-rate relative >>> to bandwidth. >> >> Yes. It's more of a "scholar's bandpass" than anything you might want >> to use in real life. It's not a bad mental tool to use when cooking up >> an active filter, because an active bandpass can be made by "sharpening >> up" a passive one, ditto with an active lowpass and (with due respect >> for stability) an active highpass. >> >> -- >> >> Tim Wescott >> Wescott Design Serviceshttp://www.wescottdesign.com >> >> Do you need to implement control loops in software? >> "Applied Control Theory for Embedded Systems" was written for you. >> See details athttp://www.wescottdesign.com/actfes/actfes.html > > Dear Tim, > You probably did not see my question in the other message, therefore > I'd like to ask it here once more: > Can you prove the following statement or point to a proof for it: > > any passive RC bandpass filter is going to have a damping ratio > greater than 1, No, I can't. I'm not sure if that was ever even proved to me -- it's just one of those "obvious truths" that get presented early in one's educational career, and is never questioned thereafter. I'm certain that it _is_ true -- if it weren't I'd have seen a circuit that took advantage of a resonant RC network. Further, I'm certain that some clever network theorist has proved it. I just don't know where to look for such a proof. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" was written for you. See details at http://www.wescottdesign.com/actfes/actfes.html |