Speaker Q, box Q and xo Q. Help me think through

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm designing my midbass section in a 4-way active system. I want it to be sealed, and I'm using 80Hz as xo point from subs to midbass.

The active nature will be implemented through digital crossovers, using high-order, linear-phase crossovers, driver linearization and room correction.

I was working with a design using Qtc=0.707 for the speaker for optimum transient response. For my chosen drivers, and using UniBox, using 6.5 liter sealed boxes achieves that, with an F3 of 121Hz and a relatively steep drop such that at 80Hz I'll be at -8.5dB. Through DSP I can easily correct to a flat response from 80Hz upwards.

But I read Linkwitz-Riley approach: use a sealed box with a Q=0.5 and a 2nd order LR xo with Q=0.5 to end up with a speaker with a Qtc=0.707. So the xo Q is affecting the speaker system Q. And my digital high order xo will have a Q...do I need to calculate it to figure out what my box Q should be so to end up with a speaker Q=0.707?

A 19 liter sealed box would provide Qtc=0.5, with a F3 of 118Hz and a flatter drop showing -6dB at 80Hz. So increasing the box size would help reach lower flatter. But would it adversely impact transient response, or other aspects of sound?

I'm confused as to what should be the box Q I should be designing towards in light of the overall system context. Would very much appreciate your guidance.
 
Qtc=0.707 is not optimum transient response

Why not make the cross at 120Hz to simplify matters? Make sure you model the excursion implications of extending your 120Hz driver to 80Hz, the result may shock you.

4th order linkwitz riley is cascaded Q=0.707 to get -6dB at crossover point and a flat sum when HP and LP Fc's are coincident. It is not dual Q=0.5 sections to get 0.707.
2nd order linkwitz-riley is cascaded 1st order with coincident Fc (Q is not defined for 1st order) to result in Q=0.5, again -6dB at Fc.

You don't understand what you think you understand and I am not sure how to untangle it in a reasonable amount of time other than point out your mistakes.
 
Qtc=0.707 is not optimum transient response

Why not make the cross at 120Hz to simplify matters? Make sure you model the excursion implications of extending your 120Hz driver to 80Hz, the result may shock you.

4th order linkwitz riley is cascaded Q=0.707 to get -6dB at crossover point and a flat sum when HP and LP Fc's are coincident. It is not dual Q=0.5 sections to get 0.707.
2nd order linkwitz-riley is cascaded 1st order with coincident Fc (Q is not defined for 1st order) to result in Q=0.5, again -6dB at Fc.

You don't understand what you think you understand and I am not sure how to untangle it in a reasonable amount of time other than point out your mistakes.

Fair enough. Indeed, I don't understand this piece. But I'm a pretty fast learner if you care to guide me.

Moving the xo point to 120Hz isn't a good idea as the subwoofers aren't good to playing that high and the midbass is good to play low. Excursion at 80hz will be very small. I will have twin drivers per side, thus reducing excursion. Playing at 100dB SPL at 80Hz, each driver will be displacing 20% of Xmax.

With what Q do you think I should design the box?
 
diyAudio Moderator
Joined 2008
Paid Member
Two second order 0.7 can make 4th order 0.5 . What concerns me is that your room is going to have its effect on top. You might end up using multiple subs to tame this and then your response should be flat, no problem. A filter that has a critical transient response is just one part of the puzzle, and thereby questionable in the bigger picture.
 
Two second order 0.7 can make 4th order 0.5 . What concerns me is that your room is going to have its effect on top. You might end up using multiple subs to tame this and then your response should be flat, no problem. A filter that has a critical transient response is just one part of the puzzle, and thereby questionable in the bigger picture.

OK, so what's your recommendation for the bigger picture and then the box Q?

BTW, is your concern about my planned path or about speaker design in general? Any commercial speaker will have a box Q, and crossover Q, and the manufacturer wouldn't know the room it will go into...so my case, where I can control box and crossover parameters, and have a known room doesn't seem that bad a situation.
 
Two second order 0.7 can make 4th order 0.5 . What concerns me is that your room is going to have its effect on top. You might end up using multiple subs to tame this and then your response should be flat, no problem. A filter that has a critical transient response is just one part of the puzzle, and thereby questionable in the bigger picture.

Show me one credible reference that talks about a 4th order "Q".
No 4th order filter has a "critical transient response"
The room statement is kind of silly.

Without knowing what your system is, I would make the midbass Q=0.707 and cross over second order with Q=0.707 to attain a 4th order L-R crossover, at 120Hz, then filter your "sub" at 120Hz 4th order L-R.

I know of very few "subs" that will not work admirably at 120Hz, most will indeed reach 300.

This whole 80Hz midbass thing is something I often see from enthusiastic youngsters who don't know better, often into home theater or car stereo.
 
diyAudio Moderator
Joined 2008
Paid Member
The plan sounds fine, just trying to take the pressure off where possible.

I'm of the belief that the combined response, when flat and when created using minimum phase sections, ends up good especially at lower frequencies. I have used filters above Q=1 to good result below the schroeder frequency. I certainly wouldn't set out to do this but you know, it happens sometimes. Then, considering what the room does, and how a good result can be had once fixed..

I know of very few "subs" that will not work admirably at 120Hz, most will indeed reach 300.
Yes, I've used spatial fillers to 1k. I see discrete modes (produced by only some of the walls) as a separate microcosm.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.