With no BSC? two drivers in parallel gives +6db, 3 gives +9.5 or so. I know BSC is rarely mentioned in commercial design information, but I think almost all have it, unless they are designed to sit right against the back wall.
I am sure as you look all the different drivers available out there, you are running smack into the unbreakable Laws of Physics. Low Fcb in a small box always gives low sensitivity......higher sensitivity requires a bigger box.....
I spent a long, long time looking at many, many, many drivers for use in sealed boxes trying to figure how to "have it all" (and still not go below 4 ohms -- one restriction you don't have to worry about). I finally decided I would have to build huge boxes if I wanted to achieve that goal.
I am still following this thread with great interest. I still plan on running the simulation mentioned previously, just had a busy week dismantling base camp and driving 1300 miles home.
JJ
More food for thought
I don't recall if this aspect of BSC has been mentioned earlier in the thread, but keep in mind that a woofer mounted close to the floor is close to being in a 2-pi environment with the floor being the planar bi-section. If you cross at 250Hz you have a wavelength of 54.2". For the range where the driver is nearly omni-directional, it is 2-pi or less because of the small reinforcement of the baffle for any part above the baffle step region. Since the floor bounce is at a point between the listener and driver, any cancellation effects will be far different than for the case with for the direct floor distance seen vertically above the driver.
In essence, the closer to the floor, the higher the frequency where a null would appear in the woofer's response at the listener's position. This helps to put the first null into the stop-band if the Fc and driver positions are chosen well. The converse holds for the midrange. The higher off the floor and the lower the Fc, the more likely that the floor reflection null is within the passband. It's a balancing act. Neglecting room modes, floor bounce is one of the most important issues for the woofer response at the listener position.
Multiple woofers will complicate this, but they will at least spread the floor bounce null effect. It will be higher in frequency for each driver mounted higher off of the floor. Good examples of 3-way design where this was well handled would be some of those by Hales Design Group and Thiel.
The woofers are not placed at the floor, they are part way between the mid and the floor. This trades off full woofer floor bounce mitigation against better integration with the midrange, since the midrange must remain as close to the tweeter as reasonably possible. The small height of the woofer above the floor retains most of the floor support due to the small increase in path difference between the direct arrival from the woofer and the floor bounce path.
In any case, you will not see the full 6db baffle step loss for most of the woofer's response for a single driver. It will be a bit more for a stacked multi-driver arrangement. For the range in which the woofer is truly omni-directional, there will be essentially no step loss.
Dave
G.Kleinschmidt said:
I don't recall if this aspect of BSC has been mentioned earlier in the thread, but keep in mind that a woofer mounted close to the floor is close to being in a 2-pi environment with the floor being the planar bi-section. If you cross at 250Hz you have a wavelength of 54.2". For the range where the driver is nearly omni-directional, it is 2-pi or less because of the small reinforcement of the baffle for any part above the baffle step region. Since the floor bounce is at a point between the listener and driver, any cancellation effects will be far different than for the case with for the direct floor distance seen vertically above the driver.
In essence, the closer to the floor, the higher the frequency where a null would appear in the woofer's response at the listener's position. This helps to put the first null into the stop-band if the Fc and driver positions are chosen well. The converse holds for the midrange. The higher off the floor and the lower the Fc, the more likely that the floor reflection null is within the passband. It's a balancing act. Neglecting room modes, floor bounce is one of the most important issues for the woofer response at the listener position.
Multiple woofers will complicate this, but they will at least spread the floor bounce null effect. It will be higher in frequency for each driver mounted higher off of the floor. Good examples of 3-way design where this was well handled would be some of those by Hales Design Group and Thiel.
The woofers are not placed at the floor, they are part way between the mid and the floor. This trades off full woofer floor bounce mitigation against better integration with the midrange, since the midrange must remain as close to the tweeter as reasonably possible. The small height of the woofer above the floor retains most of the floor support due to the small increase in path difference between the direct arrival from the woofer and the floor bounce path.
In any case, you will not see the full 6db baffle step loss for most of the woofer's response for a single driver. It will be a bit more for a stacked multi-driver arrangement. For the range in which the woofer is truly omni-directional, there will be essentially no step loss.
Dave
sreten said:
Hmmm.......
is certainly true.
Not much wrong with the Seas midrange driver for what is.
Never seen any tests that have a good thing to say about Morel tweeters.
Mating the above with > $2K worth of bass drivers ? Yawn .....
It makes no difference to me other than it ceases to be interesting.
I have not run out of "reasons", I've given up trying to help (almost).
You do not seem to want to understand my posts, so what is the point .....
Well I’m not particularly interested in pandering to your interests.
And now Morel tweeters, like ribbon tweeters are bad too. I'm getting the feeling it wouldn't matter what I suggested, it would still be bad.
I’m pretty sure I understand your posts just fine. What you don’t seem to grasp is that I’m interested in building a critically damped, sealed bass enclosure with reasonable sensitivity.
If you can suggest a cheaper bass driver that will perform as well in this application than the Scan-Speak unit I found then I am all ears, but I’m not holding my breath.
jupiterjune said:
The Scan-Speak woofer has a 2.83V/1m sensitivity of 83dB, so ~92dB with three in parallel.
A agree that the only way you can have your cake and eat it too WRT to sealed bass enclosures is to go well below a nominal impedance of 4 ohms.
I'm in a position to do this, so it is why I'm pursuing this path ATM. And that 9” Scan-Speak woofer does a critically damped 31Hz f3 in a ~35 litre sealed box, which is pretty good for a 9” driver – and it has a linear cone excursion of +/-13mm with an Xmax of +/-20mm.
Cheers,
Glen
Re: More food for thought
Hi.
Thanks for this post. This hasn't been mentioned before and there is a lot to absorb here.
With three 9" bass drivers in a WWWMT arrangement there will always be an issue with one bass driver being rather close to the floor. I don’t know how much of an issue these floor bounce effects will be in practice, but I guess a safe bet would be to delete the bottom bass driver, which would only come with a marginal sensitivity loss of ~3dB.
WRT to baffle step correction, I’m not overly fussed if I come up with a design that is theoretically less than perfect. If the average midrange and treble response deviates by + or – 3 dB or so from the average bass response, I feel that this is something easily rectified with a little active equalization if desired. When speaker placement / bass reinforcement / etc issues are drawn into the equation the effects of baffle step seem small in comparison. Well, as far as I can tell at this stage anyway.
Cheers,
Glen
dlr said:
Hi.
Thanks for this post. This hasn't been mentioned before and there is a lot to absorb here.
With three 9" bass drivers in a WWWMT arrangement there will always be an issue with one bass driver being rather close to the floor. I don’t know how much of an issue these floor bounce effects will be in practice, but I guess a safe bet would be to delete the bottom bass driver, which would only come with a marginal sensitivity loss of ~3dB.
WRT to baffle step correction, I’m not overly fussed if I come up with a design that is theoretically less than perfect. If the average midrange and treble response deviates by + or – 3 dB or so from the average bass response, I feel that this is something easily rectified with a little active equalization if desired. When speaker placement / bass reinforcement / etc issues are drawn into the equation the effects of baffle step seem small in comparison. Well, as far as I can tell at this stage anyway.
Cheers,
Glen
This is an excellent tweeter Glen, more efficient than the Millennium; I used it in my best design:
http://www.madisound.com/catalog/product_info.php?manufacturers_id=173&products_id=778
http://www.madisound.com/catalog/product_info.php?manufacturers_id=173&products_id=779
I believe that it even tests better than the Millennium based on Zaph's tests:
http://www.zaphaudio.com/tweetermishmash/
Note the much lower F2 distortion at LF for the T25CF001, also the super clean CSD and flat FR. The low F2 indicates that they got the motor right with good symmetry.
Note that Zaph does not do his tests at constant SPL, and therefore tweeters with a more extended low end and/or higher efficiency are at a disadvantage due to increased excursion.
Pete B.
http://www.madisound.com/catalog/product_info.php?manufacturers_id=173&products_id=778
http://www.madisound.com/catalog/product_info.php?manufacturers_id=173&products_id=779
I believe that it even tests better than the Millennium based on Zaph's tests:
http://www.zaphaudio.com/tweetermishmash/
Note the much lower F2 distortion at LF for the T25CF001, also the super clean CSD and flat FR. The low F2 indicates that they got the motor right with good symmetry.
Note that Zaph does not do his tests at constant SPL, and therefore tweeters with a more extended low end and/or higher efficiency are at a disadvantage due to increased excursion.
Pete B.
Re: Re: More food for thought
Glen, the woofer closest to the boundary should have the flatest response due to it having the least delay for the reflection. Think of the floor as a reflector and then you have 3 drivers also below the floor. You can actually simulate the 3 real drivers and the image drivers in CALSOD, I believe that this is covered in the manual. Still, often room modes have more influence than the one boundary so it is questionable if the in room environment is ever really 2pi. I find that large systems that can handle the full boost of BSC often sound best with a full 6 dB but it does depend on many factors.
Note that Revel uses 3 woofers in the design I linked above, you might want to have a look at the measurements in that article.
Pete B.
G.Kleinschmidt said:
Glen, the woofer closest to the boundary should have the flatest response due to it having the least delay for the reflection. Think of the floor as a reflector and then you have 3 drivers also below the floor. You can actually simulate the 3 real drivers and the image drivers in CALSOD, I believe that this is covered in the manual. Still, often room modes have more influence than the one boundary so it is questionable if the in room environment is ever really 2pi. I find that large systems that can handle the full boost of BSC often sound best with a full 6 dB but it does depend on many factors.
Note that Revel uses 3 woofers in the design I linked above, you might want to have a look at the measurements in that article.
Pete B.
Rooms are Cavity Resonators
Simulated:
http://www.silcom.com/~aludwig/images/hbl_opt.gif
Theory:
http://www.silcom.com/~aludwig/Room_acoustics.html
Simulated:
http://www.silcom.com/~aludwig/images/hbl_opt.gif
Theory:
http://www.silcom.com/~aludwig/Room_acoustics.html
I like you passionate, no holds-barred approach Glen!
Of course you will make some mistakes, but thats no reason not to set a high standard from the beginning.
As for the bass driver arrangement, have you read B&Ws white-paper on the Signature 800? Here are some excerpts from the development of the bass section which will give you more to consider :
"While the measured results looked
very promising, subjective assessment failed to
excite; the speed and timing of multiples were
always inferior to a single driver of equivalent area."
"The new 12-inch bass
drivers would play louder without audible protestation
but not loud enough. Only more cone area and
a larger cabinet would really suffice. Simply using
two of the improved 12-inch units proved to be
disappointing. Even though sensitivity and headroom
were approaching the required levels, perceived
scale, timing and speed were below expectations.
The psychoacoustics of bass is sometimes difficult to
quantify but, particularly at high levels, the body’s
contribution to space and time location means that
the brain is even less easily fooled. Two large
drivers operating in free field conditions will
produce a largely coherent wavefront comparable
to a point source at a position equidistant from the
two sources. Divergence from the central position
will produce an anomaly in arrival times. Now
place the same two drivers in an environment with
at least one near field boundary at a normal to
the array. The two drivers now experience different
driving point impedances. Their outputs will vary
in magnitude and phase and the coherency of
the wavefront is now further degraded. Take into
account the effects of the inconsistent directivity of a
large two element line source on a multi-modal room
environment, as well as a host of other reservations,
and the arguments for using multiple drivers become
less clear cut when large amounts of air are to be
displaced in a domestic environment, and there is
floor space for a single unit. A single circular cone
is simply the most spatially efficient dynamic
radiator available. This was part of the rationale
behind the development of the 15-inch unit for the
Nautilus™801."
Cheers
David
Of course you will make some mistakes, but thats no reason not to set a high standard from the beginning.
As for the bass driver arrangement, have you read B&Ws white-paper on the Signature 800? Here are some excerpts from the development of the bass section which will give you more to consider :
"While the measured results looked
very promising, subjective assessment failed to
excite; the speed and timing of multiples were
always inferior to a single driver of equivalent area."
"The new 12-inch bass
drivers would play louder without audible protestation
but not loud enough. Only more cone area and
a larger cabinet would really suffice. Simply using
two of the improved 12-inch units proved to be
disappointing. Even though sensitivity and headroom
were approaching the required levels, perceived
scale, timing and speed were below expectations.
The psychoacoustics of bass is sometimes difficult to
quantify but, particularly at high levels, the body’s
contribution to space and time location means that
the brain is even less easily fooled. Two large
drivers operating in free field conditions will
produce a largely coherent wavefront comparable
to a point source at a position equidistant from the
two sources. Divergence from the central position
will produce an anomaly in arrival times. Now
place the same two drivers in an environment with
at least one near field boundary at a normal to
the array. The two drivers now experience different
driving point impedances. Their outputs will vary
in magnitude and phase and the coherency of
the wavefront is now further degraded. Take into
account the effects of the inconsistent directivity of a
large two element line source on a multi-modal room
environment, as well as a host of other reservations,
and the arguments for using multiple drivers become
less clear cut when large amounts of air are to be
displaced in a domestic environment, and there is
floor space for a single unit. A single circular cone
is simply the most spatially efficient dynamic
radiator available. This was part of the rationale
behind the development of the 15-inch unit for the
Nautilus™801."
Cheers
David
Glen,
If you use multiple woofers, it is best to not have the lowest driver placed right at the floor. Doing this looks appealing from the standpoint of getting boundary gain (free output). However, putting a driver right against the floor means that it will excite all of the vertical standing waves in the room as much as is possible. In most rooms, the shortest dimension is usually the height, which means that those standing wave frequencies are the highest and have the worst effect on sound quality.
NHT has built a number of speakers with the woofers located directly against the floor, so I'm very famillar with the compromises involved.
If you do this, I would keep the woofers as high as possible, without compromising your tweeter height (usually around 42"-48" high). With multiple woofers at different heights, standing wave excitation and the floor bounce interference notch are all averaged out.
Another option, if you can use an even number of woofers is to put the woofers on the left and right side of the cabinet (a la AR9, NHT Xd). Doing this means that the reaction force from the driver motions cancel. The woofers are much less likely to excite cabinet panel vibrations or vibrations of things in the room. If you use four of the Scan Speak woofers this would still average out the standing wave and floor bounce behavior. You could wire them in parallel/series and still get 6dB of gain from this.
If you don't mind building a small active eq circuit to get the low end bass response you want, that would radically expand your possible woofer choices. You may have already commented on this, if so, ignore me.
If you use multiple woofers, it is best to not have the lowest driver placed right at the floor. Doing this looks appealing from the standpoint of getting boundary gain (free output). However, putting a driver right against the floor means that it will excite all of the vertical standing waves in the room as much as is possible. In most rooms, the shortest dimension is usually the height, which means that those standing wave frequencies are the highest and have the worst effect on sound quality.
NHT has built a number of speakers with the woofers located directly against the floor, so I'm very famillar with the compromises involved.
If you do this, I would keep the woofers as high as possible, without compromising your tweeter height (usually around 42"-48" high). With multiple woofers at different heights, standing wave excitation and the floor bounce interference notch are all averaged out.
Another option, if you can use an even number of woofers is to put the woofers on the left and right side of the cabinet (a la AR9, NHT Xd). Doing this means that the reaction force from the driver motions cancel. The woofers are much less likely to excite cabinet panel vibrations or vibrations of things in the room. If you use four of the Scan Speak woofers this would still average out the standing wave and floor bounce behavior. You could wire them in parallel/series and still get 6dB of gain from this.
If you don't mind building a small active eq circuit to get the low end bass response you want, that would radically expand your possible woofer choices. You may have already commented on this, if so, ignore me.
Jack Hidley said:
Thanks Jack
I like the idea of 2 woofers on each side, but that is starting to really stretch the budget
I guess the most sensible option would be to delete one of the woofers.
Maybe even a WWTM layout instead of a WWMT?
With the tweeter, at a height of, say, 45", the lower of the two 9" woofers can be kept well off the floor and with a bass crossover at 250Hz the midrange-driver centre can still be kept at about 1/3 wavelength from the centre of the lower woofer.
Cheers,
Glen
Hi Glen,
My suggestion here would be two series connected mids in D'Apollito with 4ft tweeter height; same sensitivity but reducess reproduction changes with vertical listening angle.
Also means reduced need for depth in the bass enclosure.
M
T
M
W
W
Mids take over from tweeter + W does the baffle step = happy days !
Cheers ......... Graham.
My suggestion here would be two series connected mids in D'Apollito with 4ft tweeter height; same sensitivity but reducess reproduction changes with vertical listening angle.
Also means reduced need for depth in the bass enclosure.
M
T
M
W
W
Mids take over from tweeter + W does the baffle step = happy days !
Cheers ......... Graham.
PB2 said:
Hi Pete.
That looks like a nice unit, but I'm I have a bit of a preference for stuff I can buy over here.
What would you make of the specs for this Visaton unit?
http://www.soundlabsgroup.com.au/p/V-1191-KE25SC/KE+25+SC+-+8+Ohm
Say, with a 6db pad (to account for the BSC with a pair of the 83dB 9" woofers) and a XO of 2500Hz?
Graham Maynard said:
Hi Graham.
I really like the idea of an MTM with the mid-range drivers connected in series. That gives you the same mid-range SPL, but only 1/4 the power dissipated in mid-range each voice coil and half the cone excursion. That can only be good for linearity and fidelity at high SPL’s.
However I don't like the lobing issues of the MTM. Also, adding the extra M between the WW and the T shifts the WW closer to the floor, and the top M is then even further away from the bottom woofer.
If I were to implement two series connected mid-range drivers while avoiding the MTM lobing problems, I suppose that both mid-range drivers would have to be mounted as close together as possible.
However, a mid-range driver generally needs to be as close to the tweeter as possible.....
.....so how about the outre........
MM
T
W
W
........arrangement as shown in my drawing attached?
I’d really appreciate some thoughtful comments on this. Feasible? Or more issues created than alleviated?
Cheers,
Glen
Attachments
inertial said:David Gatti,
Interesting argument/conclusion by B&W engeneers, but a few years later they contradicts themselves with the launch of the new top of the line 800 nautilus ( 2 x 10" woofers) !
Cheers,
Paolo
I always pays to be wary of subjective evaluations and white papers from those with a marketing agenda.
Cheers,
Glen
inertial said:David Gatti,
Interesting argument/conclusion by B&W engeneers, but a few years later they contradicts themselves with the launch of the new top of the line 800 nautilus ( 2 x 10" woofers) !
Cheers,
Paolo
Hmm, but I don't think they said this new one was better, just slimmer for wives
Don't they still sell the 15" one anyway...?
Simon
- Status
- Not open for further replies.