Mottlefeeder Posted October 21 Share Posted October 21 I'm confused! Rod Elliott's 36dB rumble filter uses two cascaded 18dB sections and he says it has a combined Q of 0.707, making it Butterworth. FDeck's 3rd generation 24dB HPF uses two 12dB sections and global feedback to sharpen the corner and make it Butterworth when both the variable and fixed frequency sections are tuned the same. However, WinISD gives you different responses for two cascaded 12dB Butterworth HPF filters compared with one 24dB Butterworth HPF filter. My question - Does cascading two HPF filters, each with a Q of 0.707, give you an HPF filter with a Q of 0.707, or is the Q related to the sharpness of the knee, so a steeper filter will have a higher Q? David Quote Link to comment Share on other sites More sharing options...
nekomatic Posted October 24 Share Posted October 24 I always understood that qualitatively a steeper filter will have a higher Q, but I’m not up to speed with the actual theory of Butterworth and so on, so would be interested to know the answer. Quote Link to comment Share on other sites More sharing options...
tauzero Posted October 24 Share Posted October 24 You might find some useful information in chapter 16 of https://web.mit.edu/6.101/www/reference/op_amps_everyone.pdf - it's somewhat above my head. Quote Link to comment Share on other sites More sharing options...
Mottlefeeder Posted October 24 Author Share Posted October 24 Thanks @tauzero - towards the end of chapter 16 there is a comment that to achieve an overall Butterworth response in a multi-section filter, each section should have a different Q. That implies that the copy-and-paste sections in the FDeck and Elliott designs are approximations, and the WinISD calculated responses are correct. I'll have a play in Win ISD and see if I can get a two section filter to produce the same response as their composite one. @nekomatic - as far as I can tell, the Q defines the sharpness of the corner, and is not related to the slope of the curve further downhill, so for example with a high Q, low order filter, you could have a high peak at the corner followed by a relatively gentle slope. David Quote Link to comment Share on other sites More sharing options...
Suburban Man Posted October 27 Share Posted October 27 I’d be interested to know how much ripple is created in the passband by a filter with 36dB per octave slope? Does that original design give you a graph of the frequency response? Quote Link to comment Share on other sites More sharing options...
Mottlefeeder Posted October 27 Author Share Posted October 27 12 hours ago, Suburban Man said: I’d be interested to know how much ripple is created in the passband by a filter with 36dB per octave slope? Does that original design give you a graph of the frequency response? My understanding is that the Q defines the sharpness of the knee, and that also affects the likelihood of ripple. The Butterworth 'design' maximises the range of the flat response, and the WinISD Butterworth graphs show no ripple. Chapter 16 of the op-amp cookbook @tauzero linked to does show graphs for various filter configurations. David 1 Quote Link to comment Share on other sites More sharing options...
Mottlefeeder Posted October 27 Author Share Posted October 27 (edited) Feeding WinISD'a filter calculator with the cookbook recommended Q per section for a two-section filter, I get two 12dB filters to give the same response as a 24dB filter. Unfortunately I cannot model the Rod Elliott twin 18dB filter design as WinISD only allows a triple 12dB design. But that gives me a design that I can work with. David Edited November 1 by Mottlefeeder Spelling Quote Link to comment Share on other sites More sharing options...
Passinwind Posted November 1 Share Posted November 1 (edited) On 27/10/2024 at 14:30, Mottlefeeder said: Feeding WinISD'a filter calculator with the cookbook recommended Q per section for a two-section filter, I get two 12dB filters to give the same response as a 24dB filter. Unfortunately I cannot model the Rod Elliott twin 18dB filter design as WinISD only allows a triple 12dB design. But that gives me a design that I can work with. David IMO LTspice is a much better tool for the job, FWIW. You can even run .wav or .mp3 files through your model and have a listen, if desired. Edited November 1 by Passinwind Quote Link to comment Share on other sites More sharing options...
Mottlefeeder Posted November 1 Author Share Posted November 1 (edited) It's funny how when you only have a hammer, every problem starts to look like a nail! Thanks for the info - I'll check it out. (And thanks for correcting my typo which I only noticed last night) David Edited November 1 by Mottlefeeder 1 Quote Link to comment Share on other sites More sharing options...
Passinwind Posted November 2 Share Posted November 2 23 hours ago, Mottlefeeder said: It's funny how when you only have a hammer, every problem starts to look like a nail! Thanks for the info - I'll check it out. (And thanks for correcting my typo which I only noticed last night) David Always happy to share starter files for fellow travelers breaking into the LTspice world, just PM me if you might find that helpful. I typically use a bit different 2nd order/ 2nd order format from fdeck et al, mine doesn't need a multisection pot and is meant to interact with the bass control in some pretty specific ways. There are lots of workable ways to skin felis catus though! For steeper alignments like 36dB/oct I've just used digital solutions, FWIW. Quote Link to comment Share on other sites More sharing options...
Mottlefeeder Posted November 2 Author Share Posted November 2 I'm trying to squeeze as much bass out of a small cab as I can, and the constraints I have accepted are that I need the steeper filter to reduce the port velocities. So my filter does not need to be variable, which makes things easier. With regard to digital options, the last digital circuit I designed was the logic behind a discharge-then-charge battery charger. I think it had eight gates in it. Also, learning new languages has never been easy for me, so I'm pretty much 100% anogue. I'll check out LTspice and see how steep the learning curve will be. David Quote Link to comment Share on other sites More sharing options...
Passinwind Posted November 4 Share Posted November 4 On 02/11/2024 at 11:43, Mottlefeeder said: I'm trying to squeeze as much bass out of a small cab as I can, and the constraints I have accepted are that I need the steeper filter to reduce the port velocities. So my filter does not need to be variable, which makes things easier. With regard to digital options, the last digital circuit I designed was the logic behind a discharge-then-charge battery charger. I think it had eight gates in it. Also, learning new languages has never been easy for me, so I'm pretty much 100% anogue. I'll check out LTspice and see how steep the learning curve will be. David For a digital possibility, you might want to take a look at the MiniDSP stuff, for starters. No programming needed, just a simple WYSIWYG interface and nothing to build yourself in many cases. I've never looked at chaining two of my open source HPF boards to make a 6th order filter, but it might be feasible. If I needed that transfer function I'd probably just do a new layout though. Fixed frequency certainly makes things much easier in any case! 1 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.