I just found this when Googling for some other stuff. Not exactly an in-depth explanation of each design, but the basics are covered. Sean > Simple explanations of various loudspeaker cabinet designs. PS... I don't know what their affiliation was, but i think that Xtant was somehow affiliated with McIntosh i.e. possibly their car stereo division at one point in time prior to Clarion purchasing Mac. I could be wrong about this though.... |
Think of an Infinite Baffle ( IB ) design as a dipole radiator ( Maggie's, E-stat's, etc... ) but the rear radiation never comes in contact with the front wave. The result is no cancellation or reinforcement. The driver is basically "unloaded" due to having no "pressure" on it from a small cabinet.
This is completely different from an "acoustic suspension" design, which maintains constant pressure on the woofer. In an "acoustic suspension" design, the electro-acoustic characteristics of the box volume contribute to the suspension ( compliance ) of the driver. In effect, a "sealed box" is typically a generic term for an "acoustic suspension" design, but not always. As such, an IB does NOT have to be in a sealed box and a sealed box is not always an acoustic suspension design. It can be quite confusing even though the end results are somewhat similar in performance when all is said and done.
What constitutes an IB design is that the front wave and back wave are completely isolated OR isolated well enough that the length of the path where the two waves do meet is below the usable bandwidth of the driver itself. This means a HUGE baffle area ( a wall ) if mounting it in an "open box" design. Some people with basements will mount the woofers in the floor, utilizing the entire basement as the "cabinet" for the back wave of the driver.
In effect, "infinite baffle" simply means that the front wave is "separated" from the back wave due to the use of a "baffle" of "infinite" size. That baffle can be either phenomenally tall and wide OR it can be very deep and "acoustically absorbing". Both end up making for a long signal path that encourages isolation between the front and back waves of a driver.
Some "sealed" designs are consided to be IB's when the internal box volume is much larger than the driver's electrical operating characteristics as mounted in free space. That's because the volume of the sealed box is so great that there isn't enough "pressure" generated within it when the woofer moves to alter the tuning of the driver itself. In effect, the back wave is simply "lost" within the volume of the absolutely huge sealed and stuffed box. Sean
>
PS... Many drivers that are designed for use in a "free air" system work well in an Infinite Baffle design. As a general rule, IB's do not have quite as tight of bass or as good of transient response as a properly designed low Q "acoustic suspension" design, but they can be very, very good none the less. |
Golix: differences basically as above, in my 1st answer to your post.
It dawns on me (belatedly) that what you read in the handbook probably refers to the fact that the closed box is an approximation of an IB -- since it's quite impractical to install drivers on the front wall of our house, the back wave radiating into the street beyond:)
Some differences, as above, the IB completely separates front & back waves; the box does not (because the enclosure radiates sound). The IB (take a wall of yr house, for example) is larger (i.e. infinite) than the longest wave-length radiated by the driver...
Of course the resulting ectromechanical system parametres of the two systems are completely different too. Cheers! |
golix@macunlimited.net...The term "sealed box" means that the box is relatively small and air tight so that air compressed within the box contributes much of the spring effect on the driver cone. Because of the air spring (which is very linear) the mechanical cone suspension of the driver must be very soft and floppy.
Infinite baffle is also a sealed box, but it is so large, and imperfectly sealed, that there is essentially no spring provided by air pressure. Accordingly the mechanical cone suspension must be designed to do the job without any help. A typical IB installation would be a large LF driver mounted in a door leading to the attic.
Parts Express sells some drivers that are designed for IB use. |
Ok everywhere I searched I can only find IB=sealed box,even in some white papers and the relevant formulae are identical. Also at least two manufacturers(Linn,Quested) refer to their closed box designs as IB's. Driver manufacturers too use the terms interchangebly.
So could you elucidate a bit more on the differences?
Thanks |
Completely different system parametres... one is a relatively small box trying to absorb the back wave and operating partly in 2pi and 4pi, the other has complete separation of the front & back waves (as in the driver mounted in a wall) the separator being larger ("infinite")than the longest wavelength reproduced by the transducer (i.e. radiates 2pi always).
Cheers |
I give you the one about the circular baffle, was mostly to show the size problem with open baffles.
With regards to the infinite baffle let me quote the "Audio & HiFi Engineers handbook":
'Infinite Baffle
The rear wave is'smothered' by mounting the speaker at the front of a sealed box.'
Where,then, is the difference between IB and closed box?
Now I am baffled. |
Ouch, golx! infinite baffle otherwise known as the closed box design IB, closed boxes, ported, OB etc are all different... the lowest frequency transmitted depends on the smallest(!) distance between driver edge and baffle edge, thats why they are usually circular ??? Were they circular, there would NO smallest distance. A circular baffle has the worst response pattern and is rarely used. Maybe your eye caught a sphere somewhere? The response pattern is much better there. BTW the smallest diametre is a valid point -- if that's what you're referring to. they stopped being used sometime prior to WW2 You have me baffled! Cheers! |
In an open baffle scenario the lowest frequency transmitted depends on the smallest(!) distance between driver edge and baffle edge, thats why they are usually circular!
The baffle needs to extent a quarter wavelength in every direction,ie. to achieve 32Hz cut off the baffle has to have a radius of 8feet8inch or a diameter of 17ft4inch.
Thats why they stopped being used sometime prior to WW2 and have been replaced by the infinite baffle otherwise known as the closed box design.
From the top of my head Fiddlers use of 30inch board should result in a bass roll off starting between 300 and 400 Hz. |
You might test it out using clamps... we actually drilled the appropriate holes on both baffles |
That should work. Try to make it so that it won't have buzzing vibrations where it contacts the ring behind the driver. That is a point which may vibrate, if not affixed in some way.
It sounds like a very cool idea, which offers alot of options, and fits the WAF, and looks cool, and should sound good too.
I'm interested to hear how you like it overall, after break in. |
Thanks for the tip, Gregm. That is terrific food for thought! Shouldn't be very hard to experiment with. |
I can take the big Lexan baffles out of our coat closet, lean them against the smaller, wooden Koa baffle/driver, slipping the hole in the big baffles over the Lexan donut & driver for a really nice fit Good waf idea! Tried something similar with a friend (using plexi for the "normal" baffle). The problem was securing together the two baffles, if you will, without transmitting too many vibrations from the driver baffle to the outer baffle. The compromise was to use 3 screws, two bottom one on top with a very thin layer of damping matl around the connection points. Of course, the "outer" baffle was calculated to be flush with the driver... Cheers |
|
Thanks for the info Tom.
4' x 29 1/2" is what I am experimenting with using Birch ply, but the drivers are brand new and they need to be broken in. When I have about 250 hrs. on them I will do some in-room measurements to see exactly what type of bass response I am getting.
When I actually decide on the final dimensions of the baffle, I will be using .5 inch Lexan (stiff). I don't think it will flex!
Also I am going to try a unique baffle plan. I have an idea to make a wooden baffle like the Maple one pictured in the first link I posted. My baffle will be made from Koa, a beautiful wood native to Hawaii. But here's the unique part. Mounted between the driver and the wooden baffle, I am going to place a donut of 1/2" Lexan about 1/8" larger than the outside edge of the driver. So sandwiched between the driver and the Koa baffle, will be a 1/2" thick, Lexan donut. The big 4' x 29 1/2" Lexan baffle will have a hole cut about 1/16 larger than the Lexan "donut". (BTW, the Koa baffle will be 1/2" wider than the Lexan baffle around the sides and top so the 1/16" gap between the large Lexan baffle and the Lexan donut will have a backing.
When I am ready to listen, I can take the big Lexan baffles out of our coat closet, lean them against the smaller, wooden Koa baffle/driver, slipping the hole in the big baffles over the Lexan donut & driver for a really nice fit. This will be a removable baffle that is completely flush from the driver to the edge of the baffle. I suspect this may also reduce some resonance of the large Lexan baffle since there will be a 1/16 gap between the Lexan and the Lexan donut (which is sandwiched between the driver & Koa baffle.)
I plan on using an easily adjustable spike system underneath the "base" of the Koa baffle/driver to slightly raise or lower the Koa baffle/driver to perfectly fit the hole in the large Lexan baffles. The up & down adjustment will be from the top side of the "base" for convenience.
If everything was sitting directly on the hardwood floor, raising or lowering the Koa baffle/driver wouldn't be necessary, but since we have a 13 x 20 Persian rug (plus pad) which the speakers will be sitting on, there will be some settling of the Lexan baffle, as well as with the Koa baffle/driver.
When I am not listening, it will take less than a minute to remove the large Lexan baffles and place them in the closet for WAF. All that will be left in the room will be a rather small and neat 36" high x 15" wide Koa baffle/driver. (Plus subs)
To my knowledge this idea hasn't been tried, but I am pretty confident it will work pretty well (I mean the removable baffle). I simply need to break the drivers in and experiment with baffle size.
Thanks for your interest.
PS - and I apologize if my description is not clear, but it really is rather simple, just a little hard to describe on paper. If you look at the pic of the OB in the first link I posted with maple baffles it may help to visualize the removable baffle.
|
Fiddler, that's cool.
If you keep the basic quarter-wave formula in mind, then you'll have something to lean on when deciding on baffle-related issues.
Slightly off-center driver will be a bit smoother, as would a slightly curved panel.
I wouldn't recommend going "wildly" off center for driver mounting, because as you get nearer any edge of the baffle, there could be exacerbated cancellation issues.
Assuming a four foot height rectangle, which may be (what?) around 3 feet wide? You may expect to have a decent protection against wraparound cancellation(and have bass frequency boundary reinforcement) down to around 80-90Hz.
Remember that this baffle must be rigid enough to properly allow the driver to do its thing, without flexing and causing doppler and info loss, and also must be rigid enough to not flex from the SPL which it is being directly subjected to. If you are using Lexan or plexi, make sure the thickness is good enough, and that it is somewhat braced so that it doesn't act as a "secondary radiator". It can do this, because it has a large surface area, and if it is flexing with the music, it is re-radiating the soundwaves out of phase, and will cause smear and possibly other unwanted issues. Strong is good.
Basic numbers:
For quarter-wave reinforcement above the stated freq.
Frequency / Wavelength / 1/4wavelength / Baffle size
32Hz 32' 8' 8'
48Hz 24' 6' 6'
64Hz 16' 4' 4'
96Hz 12' 3' 3'
128Hz 8' 2' 2'
If you want to use a certain frequency between these numbers, you can interpolate. |
"this info is mostly for jordans only or other OB drivers, OB drivers need much help to produce any kind of range.Other drivers wouldnt perform well in theses type of designs."
Johnk, if you are referring to the PHY driver, then sorry to disagree. Qt primarily determines whether a driver is suitable in OBs.
The manufacturer recommends the PHY KM 30 SAG for open baffle designs. For OBs, one just has to accept a compromise between bass response vs. baffle size. With subs, baffle size options can more real-world.
Thanks Twl, I am starting with a baffle about twice that size (but rectangular, not square) with the driver mounted off-center. In my earlier post, when I said smaller baffles, I didn't mean "reeeeal" small, I just meant in the four foot range with a plan to use subs crossed-over where the PHY rolls off. |
Fiddler, basically the baffle size requirement is directly related to the bass "wraparound" cancellation that is inherent in open baffle speakers. If you use subwoofers with these systems, then the baffles can be much smaller, depending upon how high you want to cross-in the subs.
For example, if you want the OB mains to only go down to 128Hz, then the quarter-wave formula would show a baffle dimension of 2' square. |
Nzera, do you have a link for Nelson Pass' comparision of the Jordan driver? I can't find it at 6moons archive |
Johnk: "OB drivers" = open baffle drivers??? If so, disabuse yourself. There no such thing as an "open baffle driver" (although certain drive units' T/S parametres may indicate good performance on an open baffle...). Cheers
Fiddler: what Twl notes above about diametres is a (unfortunate maybe) necessity. The links you give refer to ("quasi quasars") actually show baffles that should be ~0,5m wide (~20,5"). Good luck, the Phy is a beautiful unit! |
this info is mostly for jordans only or other OB drivers, OB drivers need much help to produce any kind of range.Other drivers wouldnt perform well in theses type of designs |
From the experience of numerous other people, the baffle does not have to be very large to be effective if you aren't concerned about low bass frequencies. 50-60HZ is about what one could expect with a small baffle (see link below) and the subs come in where the bass begins to roll-off determined by the in-room response of the baffle.
Here is a link with several pictures:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=19883&perpage=10&highlight=lexan&pagenumber=7
And another:
http://www.dietiker-humbel.ch/micromag.htm
There are many more pics and threads elsewhere online about OB, but this is an example of the type of OB I am experimenting with.
I know jack**** about all this, but I trust both my friends and their ears. They both currently have really high-end systems and have been through several other high-end systems. My friend said his OB PHYs wiped the floor with his previous $$K speakers and all who have heard his OB's have been very impressed. |
Fiddler, what's an open baffle?
Maybe up against the wall does not ruin left - right imaging but depth sure does suffer IMHO. Interesting what some folks consider a myth. Maybe Jordan is one of those "flat earth guys".
There aren't many speakers that measure as flat as Jordan (at least on axis) and those measurements don't take into consideration the effect of crossvers.
Sean, my computer is so slow I don't always post a link. =:) |
TWL, et'al
keep in mind the downside of having the baffle stop reinforcing at one frequency will be a very specific baffle step point. I would use a rectangle or maybe a slightly squished octagon with folding wings to extend the cancelation points toward the bass. What do you think?
Clever dimensioning could help give an extended smoother response. personally i'd at these sizes I'd build a new wall in the room with the drivers mounted in it and be done,
Best bass best everything. |
Placing the driver in the center of a circular baffle will produce increased non-linearities in response. This is due to the consistent spacing in all directions of the baffle and the baffle related diffraction / reflections that take place. If using a circular baffle, the driver should be slightly offset. This causes a staggering of center frequencies that the diffraction related lobing takes place at. Radiusing the baffle and / or treating it with acoustic damping material can help, but there are tricks to this too.
When radiusing a baffle, the radius has to taper at a very gradual rate in order to achieve maximum effectiveness. Rounding the corners of a flat baffle does little to nothing. It might look prettier and produce "flowing lines", but it does next to nothing acoustically.
According to the studies i've seen, the most linear response from a driver is achieved when placed in an "egg shaped" cabinet with a slightly flattened baffle. This staggers both the quantity and frequency of baffle related diffraction taking place due to the driver having varying distances to the edge on each part of the baffle. The contoured shape minimizes the diffraction that does take place because the reflected sound waves are seamlessly directed away from their source due to the radius of the cabinet.
This is the reason that Waveform used this design on their monitors. The fact that they "textured" the cabinet also helped to break up standing waves on the baffle even further. While i did not personally like specific design aspects on some of their products, they truly were doing some innovative things that nobody else had attempted to conquer. Sean
> |
Fiddler...Please tell us all about your open baffle system. |
Fiddler, you can calculate the size of the needed baffle by using the quarter wave formula.
It depends on how deep in the bass you want these things to produce.
Pick a frequency above the F3(resonant freq) of the driver, and deep enough into the bass region to satisfy you.
For example, if you want 32Hz to be able to be reproduced by the system(assuming the driver goes that low), you will need a baffle that is at least 1/4 wavelength of the (32')32Hz wavelength, which would be 8 feet across. If you want to get only to 64Hz(16' wavelength), then you'd only need a baffle that is 4 feet across. If you want to "split the difference, and get 48Hz, then you could use a baffle that is 6 feet across.
All frequencies above the frequency determined by the baffle board will be reinforced by the boundary effects of the baffle board, and will not be subject(much) to the "wrap-around" cancellation effects that will adversely affect the frequencies below the capability of the baffle board to handle.
While it is a matter of discussion whether to mount the driver in the center of the baffle board, due to the relative "smoothness" of the response curve, mounting it in the center would give the best result in getting to the deepest bass for a given size baffle board.
The determining factor(in size) for the baffle board's effectiveness is based upon the smallest outside dimension that goes across the driver. So the smallest outside dimension will dictate your reinforcement frequency, and any larger dimensions will have little effect, but they may do a little. Thus, a circle shape will be the minimum sized shape that you could use, but it is more difficult to stand up and use. A square baffle is typical, but only the largest diameter circle that you can draw around the driver will be doing the work, and the corners doing very little.
Remember, if you make it tall and narrow, you are losing your effectiveness, so make it just as wide as it is tall.
Having the piano hinges, and angling the sides backwards to a small degree will do just as well, and helps to make the baffle more visually acceptable. It also keeps the edges out of the diffraction plane.
The best thing would be to make the whole baffle shape like one of those "snow saucers". It would allow the driver to be placed at the foremost part(center), and have uniform shallow radius sloping backward all the way around, for most smooth response and minimum diffraction. But it would need to be big, and I've never seen a 6 foot snow saucer!
For the easiest construction, flat and square works. |
No Gregm, no piano hinges.
I am going to experiment with plywood baffles until I find the size of baffle that seems to be optimum and then I will have the baffles made out of Lexan if I like the results of the plywood. It will be a flat baffle with no hinges. Others have had great success with this design.
I have a friend that has just done this and he and another friend said the results have been nothing short of phenomenal. But he used piano hinges, but many others haven't. The open-baffle PHY's may suck in my room, but time will tell.
No additional tweet. The PHY KM 30 SAG has a time-aligned piezo-electric tweet that goes to 30kHZ.
They will require subs for sure. |
Wow, Fiddler. I expect you're trying out Phy's recommended open baffle with the piano hinges on the side flaps... Planning to use a tweet??? |
I have owned Carolina Audio JTM speakers for about a year now. These speakers use the Jordan driver in a wide baffle transmission line enclosure. I have owned many other speakers, including five other brands at present. So far, compared with these other speakers I have found this implementation of the Jordan driver to be the best all-around speaker for typical rooms I have yet owned. As Sean pointed out, the wide baffle helps with the bass. With the JTMs bass is very accurate and satisfying without a sub. They are easier to place than any speaker I have owned- much less sensitive to side and rear wall distances (perhaps the limited off-axis dispersion of single driver speakers helps with this as well- this is a liability as well). The design does a good job of highlighting the strengths of this 4'' Jordan driver, while achieving in-room bass comparable to many floorstanders out there.
The Jordan drivers themselves measure flatter than most other single-drivers (see the Nelson Pass measurements on 6moons comparing different single drivers), and when used in this kind of enclosuse, can deliver fairly flat in-room response without a lot of tweaking. Considering how good these drivers are, surprising there aren't more speakers out there using them. Seems like most interest in single drivers is for use with low power SETs. A shame since this kind of speaker works well with more common solid state and tube audiophile gear. |
|
Fiddler, what drivers are you using /filter configuration (if any)? |
While the Jordan driver is not high-sensitivity like many single-driver systems, it has many of the pros and cons of these kinds of systems.
The Jordan driver was awarded best sound in at least one national "DIY speaker" competition, and has gotten some follwers for its sound quality. |
Cdc, I am currently in the process of putting together an "open baffle" design. If you are interested, I will let you know my impressions when the drivers are broken in. |
Thanks for providing a link ; )
1) Larger baffles increase bass response, which is what he's after with this specific design. There are trade-off's associated with increased baffle dimensions though, but he's obviously willing to take them in order to get what he's after for use with that driver.
2) Placing the speaker cabinet closer to the wall increases bass response, which is what he's after. Like the above response, there are trade-off's involved, but he's willing to deal with them in order to get what he's after for use with that driver.
3) Placing the speaker cabinet nearer a side wall will increase bass response, which is what he's after. Like the above two responses, there are trade-off's involved, but he's willing to deal with them in order to get what he's after for use with that driver.
Have you noticed any specific similarities? Figure it out. It is a small driver that he's trying to use full range. As such, it is going to be deficient in bass response. On top of that, due to the size of the driver, it will become beamier as frequency rises. As such, the problem with sidewall reflections will be somewhat reduced as compared to designs using a tweeter or smaller midrange drivers.
With all of that in mind, I read that Nelson Pass really liked the Jordan drivers. This tells me that it obviously has some very alluring qualities to it. Whether or not it is truly suitable as a full range driver with no auxillary support above or below the midband is up to the end user. Sean
>
Now Listening to: Jazz At the Pawnshop 2 / Redbook CD
|
I also know very little, but:
1. I think he's making the argument that sound can diffract off the edges of baffles. Some manufacturers address this issue by rounding the edges of the baffles or the shapes of the cabinets in various ways. Jordan seems to recommend using a larger baffle instead to decrease such diffraction, but it really does come across as though he's advocating in-wall speakers. The flip side, of course, is that many folks seem to report better results (less "boxy" sound) with narrower baffles.
2. It's a bit ambiguous, but I believe that "rear wall" refers to the wall behind the speakers, as he refers the advantages of larger baffles. In essence, I think he's saying that backing the speakers up against the wall makes the wall behind the speakers act almost as a large extension of the speakers' baffle, especially because at least one of his designs is very shallow.
3. As this question follows the previous one, I believe that he's saying that having the speakers backed up against the (what would be front to us) wall isn't detrimental to stereo imaging, not having the side of the speaker against the side wall.
Young-Ho |
I'm not an expert on speaker design, but I can assure you of one thing - several of these questions and answers would certainly disinterest me in this man's designs.
For the life of me I can't even imagine an infinitely large baffle - for a home system thats comparable to saying a wall mount speaker is ideal. The next two questions are as interesting - which wall is the rear wall - in my experience the rear wall is the wall behind the listener. A couple of feet from the front wall is going to make a materiel difference in minimizing the time delay etc - where in a phone booth?
And, lastly, he can't see any reason why his speaker placed next to a wall would ruin a stereo image - his site has dynamic drivers, do the have no off axis dispersion - perhaps only a narrow beam of sound and there is no first reflections to degrade the signal?
But as I said I know nothing about speaker design so YMMV. |
