For sealed designs the woofer has to be designed for that purpose and for one such woofer there can be only one optimal box design.
Vladimir said:
In the days when acoustic suspension designs were popular a woofer would have a very loose suspension in order for the air trapped in a small sealed box to act as the main spring. Such a driver could not be reasonably used for other box loadings but there are almost no drivers around these days designed to work in such a manner. Woofers and midwoofers for sealed boxes today tend to have stiffer suspensions and are intended for use in larger volumes where the driver suspension and not the air volume is the main spring. This allows the same driver to work with more than one type of cabinet loading.
There is no such thing as a single optimal design for a sealed speaker because the various factors in the design are not given the same weight by different people. For example, what should the Q of the box resonance be? Critical damping would suggest 0.5, the flatest response 0.7, 0.85 can sound a bit less dry and allow a smaller volume,... How much stuffing should be used? There is no single correct answer but there are plenty of wrong ones.
hg said:
When I saw AR 3a for the first time I was surprised from the look of the woofer (the first Alnico version). Very loose, very wide textile suspension. The damping had to be exactly right in grams, everything about the box and the driver was made to work together and you couldn't replace the woofer or the box and get same result. I learned this from Ken Kantor's explanations. From personal experience I refoamed my AR 11B speakers with surrounds that were adequate for 11" woofers, but the width of the roll was not large enough. The speakers simply didn't work as well. There was no deep deep bass of 27Hz.
Vladimir said:
With the exception of some subwoofers which are not designed to work upto a midrange, woofers in modern sealed designs, of which there are few, are not designed to work in tiny "acoustic suspension" cabinets but larger "infinite baffle" cabinets like the SP Acoustics speakers. Here is a competent DIY example which may help with the physics.
Jota180
Well-known member
Interesting thread guys and I appreciate hg's insights explaining things in a rational way and making it easy for the laymem among us.
Electro
Well-known member
SteveR750 said:
These might be of interest to you , they are not concrete but they are made of 7mm cast Aluminium tensioned with push rods to 2500N .
http://www.electrocompaniet.no/products/classic/speakers/
http://www.hifiplus.com/articles/electrocompaniet-nordic-tone-model-1/
SteveR750
Well-known member
Electro, those are indeed interesting! Especially this part:
Frequency response: 28 - 35000 Hz ±2dB
Sensitivity: 90 dB, re 2.83V @ 1m
Frequency response: 28 - 35000 Hz ±2dB
Sensitivity: 90 dB, re 2.83V @ 1m
Best thing to do if possible is to get hold of some sealed Spendors and do a home demo. If you are happy with the bass buy them if not look for some front ported speakers.
Ignore all the technical stuff quoted unless you are interested in researching and understanding more about speaker design.
matt49
Well-known member
Vladimir said:
"Sir, can we play 'guess the speaker by its FR as measured by John Atkinson and published in Stereophile' please, sir?"
"Oh, if you insist ..."
hg, would you say that the most obvious way to compare diesel and petrol cars would be to put the same sized engine in each?
SteveR750 said:
Except they don't give the conditions which is rather important when it comes to bass performance.
They look to be using a pair of 8" SEAS Excel drivers in about a 100 litre enclosure. The standard driver is roughly 8 ohm and 88 dB sensitivity into half space giving 94 dB for a pair in parallel. 100 litres is about the right size for a Q of 0.7 in an "infinite baffle" enclosure for these drivers. Into half space the -3 dB point would be around 42 Hz. The midrange looks like a 5" Scan-Speak Revelator and the standard driver has a sensitivity of 90 dB which is in line with the speaker's quoted senstivity.
If the speaker were to radiate into the full space of an anechoic chamber below about 300 Hz for this baffle width sound will progressively start to radiate in all directions rather than just forward like it does at higher frequencies. Below about 50 Hz sound will be radiating equally in all directions which will require the woofers to generate 6 dB more than they would into half space in order to maintain a flat frequency response at the listening position in front of the speaker. If you look at the measured frequency response close to the woofers of this similar speaker you can see what is required.
The woofers have got 4 dB to use for this effect which would mean in an anechoic chamber the -3 dB point at the listening position in front of the speakers would be higher than 42 Hz at say 55 Hz or so. So we can state with confidence the frequency response has not be taken in anechoic conditions.
In a room at low frequencies we have room resonances which are not helpful but we also have boundary reinforcement which can be helpful. The low frequency sound that travels backward will hit the rear wall and be reflected forwards. At low frequencies the wavelength of sound is long, e.g. 10 metres at 33 Hz, and so what is reflected back is largely in phase with the direct sound and will add making it louder. This can also occur from the other boundaries like the floor, walls and ceiling. Unfortunately when the sound has passed our head it will come back from the boundaries behind us which may add or subtract depending on the wavelength and the distance travelled but it will also be quieter due to the sound continuing to spread plus some absorption by the walls. So the additive effects of the early reflections will tend to dominate and raise the level at the listening position.
In truth there is a bit more to it such as room resonances and pressurisation but a gain of 10 dB or so to push the in room -3 dB point down to 25 Hz or so is reasonable. So my guess would be it is an in room measurement of some kind but not stating the conditions is a bit naughty although effective as we can see from the posts above.
Electro said:
I hope that they are quite cheap, a Revelator midwoofer?
I would not settle for anything less than Illuminators!
SteveR750
Well-known member
hg said:
Thanks for the explanation. The FR does appear to be remarkable for a box that isn't the size of a garden shed! It also does highlight the problem of just looking at paper specifications.
matt49
Well-known member
steve_1979
Well-known member
steve_1979
Well-known member
matt49
Well-known member
Electro
Well-known member
SteveR750 said:
Apparently the measurements were made in an anechoic chamber at Seas .
http://translate.google.co.uk/translate?hl=en&sl=da&u=http://www.lydogbillede.dk/test/electrocompaniet-the-nordic-tone-model-1/&prev=search
SteveR750
Well-known member
Electro said:
Some measurements may have been taken in the SEAS anechoic chamber but I see no reason to assume that is where the extension to 28 Hz has come from in a raw form (i.e. without some processing to simulate a room). A small anechoic chamber like that will start to suffer from significant reflections below about 100 Hz and so it would be a very poor sign if the very low frequency measurements came from a microphone placed a few metre in front of the speakers.
It is not difficult to measure the low frequencies by placing a microphone close to the woofer to minimise the influence of reflections and doing it outside to avoid issues like leaky room pressurisation. However, what is measured is not what would be heard at the listening position and what is heard at the listening position is strongly influenced by the room. Some assumptions need to be made to get a reasonable figure. The first is what is often called "baffle step correction" where the low bass stops radiating mainly forward like the high frequencies and progressively shifts to radiating equally in all directions. The second is the boundaries of the room taking that omni directional radiation and directing it towards the listener. The first effect could be fairly successfully measured in the SEAS anechoic chamber because most of it occurs at high enough frequencies. The second varies from room to room and placement within the room and so it is unreasonable to include it without stating what has been assumed. Since a speaker with a perfectly flat frequency response in an anechoic room would normally have too much low bass at the listening position in a real room it is not unreasonable to want to include the second effect for a speaker that has too little low bass in an anechoic room but is likely to sound about right in some rooms.
Stereophile take their low frequency measurements by measuring outside close to each woofer and port and then adding them up. They then join this measurement onto the gated (removes reflections) in room high frequency measurement at a few hundred Hz (I haven't looked up the exact figure and how/if they smooth the join). However, they make no attempt to correct for either the room or "baffle step correction" and so speakers that a have a reasonably flat frequency response at the listening position are shown with a rising bass response like this one from Sony which has a similar pair of 8" woofers in a tower. I suspect a fair few readers will fail to correctly interpret such plots particularly as boundary gain in most room locations tends to kick in and reduce the required amount of baffle step correction to less than the full 6 dB that Sony seem to have implemented.
Facebook
Twitter
Instagram