I've been thinking about thin-wall enclosures that the BBC pioneered and I wonder - does it need to be ported to work right?
Most of the BBC designs were ported, and at first thought it seems it would be necessary to prevent the thin wall enclosure from being internally pressurized and flexed excessively. However, a port actually only operates as an open 'hole' to sound below its tuning frequency. At or above its tuned frequency it has a high acoustic impedance and is seen by the woofer pretty much as a closed box. So in such a speaker with a 50Hz tuned port, at 100Hz the enclosure will still be pressurized regardless. Does that seem right?
If this is indeed the case then it makes one wonder if it really matters that the cabinet is ported since it would only reduce flexing of the walls at very low frequencies. Maybe that matters though?
Comments and thoughts?
Most of the BBC designs were ported, and at first thought it seems it would be necessary to prevent the thin wall enclosure from being internally pressurized and flexed excessively. However, a port actually only operates as an open 'hole' to sound below its tuning frequency. At or above its tuned frequency it has a high acoustic impedance and is seen by the woofer pretty much as a closed box. So in such a speaker with a 50Hz tuned port, at 100Hz the enclosure will still be pressurized regardless. Does that seem right?
If this is indeed the case then it makes one wonder if it really matters that the cabinet is ported since it would only reduce flexing of the walls at very low frequencies. Maybe that matters though?
Comments and thoughts?
Actually at resonance it's much worse for the ported box - here the output of the port is in phase with the front of the cone so the box is being (de/)pressurised roughly twice as much as a sealed one of the same dimensions. Yes, it's better for the ported box below resonance but which is better overall from the viewpoint of wall flexing? I guess it depends which frequencies you're optimising for.
No. The LS3/5A was sealed whereas larger models were ported -as was the Spendor BC1.
Ahh but the LS3/5 is a pretender, its too small for the thin wall theory to work right IMO. At that size it's pretty much ridged.
Ian I think you have a good point. The idea of the BBC design was to optimise for midrange frequencies, so make the cab resonance low, and I think in this case it doesn't matter is it is closed.
Ian I think you have a good point. The idea of the BBC design was to optimise for midrange frequencies, so make the cab resonance low, and I think in this case it doesn't matter is it is closed.
I'm currently building a thin walled cabinet (9 mm) for my SEAS full range drivers. Net volume will be about 45 liters, and I'll use an aperiodic vent.
In Norwegian:
Tynne høyttalerkabinetter, BBC-style
In Norwegian:
Tynne høyttalerkabinetter, BBC-style
Tenson, what exactly is the "thin wall theory"?
I have some old celestion speakers with very thin walls. They perform surprisingly well with a bit of EQ.
This:
http://diy-audio.narod.ru/litr/1977-03.pdf
and other seminal papers are available at the BBC online archives.
We debated this quite a bit recently here:
http://www.diyaudio.com/forums/multi-way/138111-what-characteristics-better-material-enclosure.html
Follow the links to the excellent paper by Harwood at the BBC. In general he shows (proves) that you can achieve better cabinet performance with a thinner cabinet but with a greater thickness of damping material (more of the cabinet mass is from the damping material rather than the wooden structure).
This approach works just as well for sealed cabinets as vented boxes. Much confusion here. Note that the pressure in the box is related to the pressure outside the box. It is in fact the 2pi response curve integrated twice (given a 12dB per Octave bass rise EQ). So a sealed box and a vented box with the same response (same bandwidth, say) will have equal pressure inside the box. Presure at the vent frequency is no higher than for other frequencies.
David S,
http://www.diyaudio.com/forums/multi-way/138111-what-characteristics-better-material-enclosure.html
Follow the links to the excellent paper by Harwood at the BBC. In general he shows (proves) that you can achieve better cabinet performance with a thinner cabinet but with a greater thickness of damping material (more of the cabinet mass is from the damping material rather than the wooden structure).
This approach works just as well for sealed cabinets as vented boxes. Much confusion here. Note that the pressure in the box is related to the pressure outside the box. It is in fact the 2pi response curve integrated twice (given a 12dB per Octave bass rise EQ). So a sealed box and a vented box with the same response (same bandwidth, say) will have equal pressure inside the box. Presure at the vent frequency is no higher than for other frequencies.
David S,
People often get confused about this. The fact is that above port resonance, the air in the box behaves as if the box were sealed. In-box pressure changes (i.e. sound) do not "have the time" to find out that there is a port at all, if I am allowed an inappropriate anthropomorphism here. The only case that box air decompresses readily is below port resonance: in that frequency range air is just pumped in and out, no questions asked. To continue with the anthrpomorphism, box air gets its harder pounding in the area between port resonance and woofer resonance, a double whammy.
All these BTW happen way lower than panel resonance frequencies, which are usually in the three-digit range. Which is as it should be: having panels resonate in a frequency range where they will be "fed" with energy by woofer or port resonance is a sure recipe for disaster. This is just about the best plain wood can do. To go the opposite way, i.e. trying to push panel resonances way above the midrange, requires pushing for superlative rigidity. This entails complex space-frame constructions, molded polymers, metals, and what not.
Which opens another can of worms. Tweeters, which normally work with micrometer-wide displacements, don't really like sitting on a vibrating platform. This is conceptually similar as having major slack in tonearm bearings: don't expect a stylus to trace a grove faithfully if the platform it's fixed on is doing the pogo dance. Loss of resolution is inevitable, and often severe.
All these BTW happen way lower than panel resonance frequencies, which are usually in the three-digit range. Which is as it should be: having panels resonate in a frequency range where they will be "fed" with energy by woofer or port resonance is a sure recipe for disaster. This is just about the best plain wood can do. To go the opposite way, i.e. trying to push panel resonances way above the midrange, requires pushing for superlative rigidity. This entails complex space-frame constructions, molded polymers, metals, and what not.
Which opens another can of worms. Tweeters, which normally work with micrometer-wide displacements, don't really like sitting on a vibrating platform. This is conceptually similar as having major slack in tonearm bearings: don't expect a stylus to trace a grove faithfully if the platform it's fixed on is doing the pogo dance. Loss of resolution is inevitable, and often severe.
I thought I'd share a quick photo of my new thin-wall aperiodic cabinets for Exotic. No finish yet, but here goes:
9 mm birch plywood damped with bitumen and with beech fillets.
An externally hosted image should be here but it was not working when we last tested it.
9 mm birch plywood damped with bitumen and with beech fillets.
Do think that front panel should be little thicker if other panels are 13m baltic birch? Tweeter should not be vibrated so what if front panel is 25 mm and others are thin?
Read the Harwood paper referenced above. The theory (supported by measurements) is that the front panel will vibrate less if the cabinet has thinner panels but thicker damping. That is, the normal panel resonances will be most damped when we keep the panel mechanical impedances low and damping proportionately higher. The thicker you make the panel the less effective the damping becomes.
And why should the tweeter not be vibrated? This is a common audiophile precept unsupported by facts, in my humble oppinion. In the end the tweeter signal and the panel vibration will add. One does not hide the other. We need to keep panel resonance outputs below the threshold of audibility, but it has nothing to do with the tweeter.
David S.
Hi,
In some deeper than wide thinwall cabinets only the side walls and top
and bottom are damped. Typically the undamped rear is thicker than
the sidewalls/top/bottom and the undamped baffle thicker again.
rgds, sreten.
In some deeper than wide thinwall cabinets only the side walls and top
and bottom are damped. Typically the undamped rear is thicker than
the sidewalls/top/bottom and the undamped baffle thicker again.
rgds, sreten.
I suspect this is more for structural strength to allow for secure mounting of the drivers. Whatever the panel, level at resonance will drop if walls are thinner and the damping to stiffness ratio goes up.
David
David
Have you got more specific details of good thin enclosure?
If had 230x300x450 mm box, would 12-13 mm birch + 4 mm bitumen be good? No inner bracing, just damping with bitumen?
If had 230x300x450 mm box, would 12-13 mm birch + 4 mm bitumen be good? No inner bracing, just damping with bitumen?
I'd be tempted to double up the damping thickness to 8mm, if possible. The 12-13mm birch is a good shell, but the idea is to get to a more extreme damping to wall thickness ratio. Surface damping is generally inefficient, so a good thickness is desirable. (Read up on constrained layer damping.)
Can you get Finnish plywood in your area?😉
David S.
BBC recommends about the same thickness bitumen as the wall itself, but you don't need to cover the whole thing:
http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf
I've used 5 mm thick bitumen in the following way: double layers on top/bottom and large pluss smaller sheet on both sides and back.
http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf
I've used 5 mm thick bitumen in the following way: double layers on top/bottom and large pluss smaller sheet on both sides and back.
Another tip from Harbeth (which still use this method of construction) is to screw (not glue) the back panel into place. This breaks up resonances between the panels (the cracked bell effect, as they describe it).
- Status
- Not open for further replies.