As I browsed many "diy" speaker set vendor, most of them advertised their enclosures are strengthened with a square frame.
This is a typical example:
5" 2-way unit:
H(300mm) X W(180mm) X D(269MM)
Material: 18mm MDF.
For 6.5" 2-way unit:
H420mm*W230mm*D296mm
The rest are the same with 5" enclosure.
One I talked to a person of Silver Flute, he said its a market gimmick. Given the material is strength enough(I guess he refers to 18mm MDF), no strengthening needed.
Thanks for any suggestion.
This is a typical example:
5" 2-way unit:
H(300mm) X W(180mm) X D(269MM)
Material: 18mm MDF.
For 6.5" 2-way unit:
H420mm*W230mm*D296mm
The rest are the same with 5" enclosure.
One I talked to a person of Silver Flute, he said its a market gimmick. Given the material is strength enough(I guess he refers to 18mm MDF), no strengthening needed.
Thanks for any suggestion.
Depends on the speaker design. Conventional box speakers tend to have trouble with the Fourier series resonances supported by rectangular enclosures, requiring bracing to manage mechanical resonances and damping for acoustic ones. The most complex alternatives support something approximating Legendre polynomials---look at the B&W Nautilus---but Bessel type solutions like the Linkwitz Pluto are pretty easy to build.
Another mechanically simple alternative is dipoles as panel resonance is not generally too hard to control. Or just go with a nude dipole.
Another mechanically simple alternative is dipoles as panel resonance is not generally too hard to control. Or just go with a nude dipole.
Dipole may consider in future project.
Do you think a square frame inside the box can damp the resonance?
For 12inch or bigger speakers I always use 18mm plywood and 2inch square bracing.
I have a 2 by 12 inch 450WRMS speaker cabinet and a 650WRMS 18 inch eminence bass speaker cabinet.
They need to be strong though because they sometimes get taken out on the road to do discos or guitar work.
I guess the bracing needs to stop cabinet vibrations mostly and after that not fall apart if moved.
I have a 2 by 12 inch 450WRMS speaker cabinet and a 650WRMS 18 inch eminence bass speaker cabinet.
They need to be strong though because they sometimes get taken out on the road to do discos or guitar work.
I guess the bracing needs to stop cabinet vibrations mostly and after that not fall apart if moved.
Depends on the frame and box you have in mind; as posed this question is too vague to answer meaningfully. If you're not familiar with Fourier basis functions your best option for predicting what'll happen is probably FEM simulation. If you can share 3D renderings or plan drawings of the design that's likely sufficient for informed speculation but, ultimately, that's no substitute for taking the time to develop your own design and analysis skills.
Very broadly speaking, what I was saying in the previous post is it's often easier to change the geometry to avoid problems than it is to solve them head on. If you're asking about typical rectilinear bracing in typical rectilinear enclosures usually it's more about shifting resonances out of band, though how feasible that is depends on the enclosure geometry, crossover frequencies and slopes, and material and bracing choices.
Hi,
no, it can not. It will shift the panel resonance to higher frequencies, and will probably even increase the Q, so the resonance may become more harmful.
If you're lucky the resonances will be shifted out of band, but that is very difficult, and nearly impossible with wooden enclosures. "Out of band" means above 20 kHz so you won't here them any more. It does not mean "above operating range of the woofer", because the midrange and even the tweeter may excite the panel resonances. I tested this last friday with an accelerometer, and this was very surprising to me (2 6.5" very light midranges were exciting the panel resonance as strong as 2 10" heavy woofers, and even the tweeter did excite them).
Read also the posts of Speaker Dave on this topic, he put me in the right direction recently.
Bracing is not completely useless, though. If you build a big floorstand, it will most probably collapse (either instantly or after some time) without bracing. Do not use thicker material instead, which is even worse than bracing.
In such a case, I would use thin wood (9 to 12 mm plywood), careful bracing so that the enclosure does not collapse, and apply a big amount of damping material on each inner surface, at least same thickness as the wood. It may be useful to strenghten the edges of the enclosure, especially with the 9 mm plywood. Decouple all drivers, at least with rubber grommets.
Baseballbat
no, it can not. It will shift the panel resonance to higher frequencies, and will probably even increase the Q, so the resonance may become more harmful.
If you're lucky the resonances will be shifted out of band, but that is very difficult, and nearly impossible with wooden enclosures. "Out of band" means above 20 kHz so you won't here them any more. It does not mean "above operating range of the woofer", because the midrange and even the tweeter may excite the panel resonances. I tested this last friday with an accelerometer, and this was very surprising to me (2 6.5" very light midranges were exciting the panel resonance as strong as 2 10" heavy woofers, and even the tweeter did excite them).
Read also the posts of Speaker Dave on this topic, he put me in the right direction recently.
Bracing is not completely useless, though. If you build a big floorstand, it will most probably collapse (either instantly or after some time) without bracing. Do not use thicker material instead, which is even worse than bracing.
In such a case, I would use thin wood (9 to 12 mm plywood), careful bracing so that the enclosure does not collapse, and apply a big amount of damping material on each inner surface, at least same thickness as the wood. It may be useful to strenghten the edges of the enclosure, especially with the 9 mm plywood. Decouple all drivers, at least with rubber grommets.
Baseballbat
Any chance you could post the data somewhere? Good accelerometer measurements are rather hard to find, particularly on three ways.
The opposite is more typically true. Raising the frequency of a (potential) panel resonance higher will (usually) make it less likely to get excited (as will increasing its Q).
dave
You'll get lots of differeing opinion, but i don't think MDF is a good material for making loudspeakers.
dave
Hi,
no, I'm sorry, I accidentally hit the wrong button and all data was lost. But I will do some more tests in this case, and then it will be easy to repeat the measurements.
It will shift the frequency higher, increase the Q, and as the input force is constant, it will be excited like the lower resonances, but with at least equal amplitude. That's what all publicly available and my own measurements show.
Best
Baseballbat
no, I'm sorry, I accidentally hit the wrong button and all data was lost. But I will do some more tests in this case, and then it will be easy to repeat the measurements.
planet10 said:
It will shift the frequency higher, increase the Q, and as the input force is constant, it will be excited like the lower resonances, but with at least equal amplitude. That's what all publicly available and my own measurements show.
Best
Baseballbat
What I have not seen discussed much here is the advantages of forcing materials away from a slab. I have an idea for a low diffraction box of only simple curves. Think about a wide oval cylinder on end with the top a simple curve leading away from the baffle. ( my use case requires it to sit on a shelf, so a round bottom is not viable.) The two edges basically perpendicular to the wavefront. Yes, MDF looks like it will be the material, but laminating several layers of thin hardboard may be better. This is a couple of projects away. Can't say 3/4" MDF is a great material, but I have not had better luck in the 10 to 20L size with other materials. I have no idea where to get that super-duper CF reinforced 9-ply 1/2 inch spruce grown from only north side trees in Norway. I have to deal with what Home Despot sells. I can BUY quarter sheets of MDF which automatically makes it a suitable material.
hi dave,
you are arguing against Newton's third law.
The acceleration of the driver's cone/membran/anything is constant above its resonance frequency*. Otherwise, you wouldn't have a linear frequency response. Thus, the opposing force on the cabinet is constant. There is no mysterious 1/f², only for the energy, but energy is not equal to amplitude (a 1h signal of level 1 has the same energy as a 1 Minute signal of level 60).
* In real applications there will be usually a 6 dB down step in acceleration due to the baffle step, so this might be beneficial.
Baseballbat
you are arguing against Newton's third law.
The acceleration of the driver's cone/membran/anything is constant above its resonance frequency*. Otherwise, you wouldn't have a linear frequency response. Thus, the opposing force on the cabinet is constant. There is no mysterious 1/f², only for the energy, but energy is not equal to amplitude (a 1h signal of level 1 has the same energy as a 1 Minute signal of level 60).
* In real applications there will be usually a 6 dB down step in acceleration due to the baffle step, so this might be beneficial.
Baseballbat
Still waiting to see your data on this since all the plots I've seen show fairly even distribution of resonant peak level across the frequency band. Harwood and Shorter showed that greater wall stiffness actually raised peak level (made damping less effective).
There is no simple relationship between Q and likelyhood of excitation. With a mechanical system, if Q is increased (resistance reduced) then peak height goes up. Area under the curve increases and likelyhood of excitation goes up. The only way to reduce resonant bandwidth while keeping the same resonant frequency would be to increase mass reactance and stiffness reactance in step, hard to do.
The fact that peak heights are fairly constant is proof that the resonances are equally likely to be stimulated.
David S
Internal SPL does tend to roll downwards with frequency, but accelerative force from the woofer chassis (reaction force) is constant. Baseballbat is correct that all published data shows a fairly broad trend to resonant output, rather than any 1/f^2 trend.
David S
Here is some more.
http://downloads.bbc.co.uk/rd/pubs/reports/1958-31.pdf
Page 20 (25 of the PDF) shows the output of a cabinet with and without a damping layer. Strong sequence of resonances from 200 to 3000 Hz.
This was a fully sealed cabinet with a buried internal driver.
David S
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.