Building Resonance Free Subwoofers

@daledeee1 Good question. I know they are very low, but I do not have the equipment to measure it. I can tell you that all other things being equal, among the finest drivers, the primary source of distortion is the size of the driver. As the cone moves farther the suspension starts to get stiffer until the driver can go no further. That nonlinearity creates distortion. Larger drivers do not have to travel as far to create a given volume. My subs, with two 12" drivers, are equal to one 15" driver in distortion levels. The drivers I am using are Dayton's Reference Series "low distortion" drivers for smaller enclosure volumes. They have very strong motors, have very light and stiff aluminum cones and are extremely well vented. They are without question the finest drivers I could find for my design and money was not a factor at all. Having said all that the greatest source of "distortion" in subwoofers is the enclosure. If you have subs, turn up the volume with a bass heavy number and put your hand on the enclosure. Any vibration you feel is audible distortion. These subwoofers do not vibrate at all. Another interesting characteristic is you can hear each bass note clearly on records where it was previously difficult. You know there is bass there but you could not discern the note clearly. The large surfaces you have in a box enclosure are resonance prone. In order to get the resonance frequencies up above the subs operating range you have to use internal bracing. Even with bracing the forces generated by a single driver are going to cause even extremely heavy enclosures to vibrate a lesson I learned by building 200 lb Corian enclosures. 

How does the sub you made measure for distortion vs the standard sub box shaped enclosure.

@forrestc  Thank you Forrest. I should add that this is the fourth try at making a high performance subwoofer that can match up to an ESL. The other three were not so successful, but they were a learning experience. What I learned was that any amount of reasonable mass is not enough to control a subwoofer driver and it is extremely difficult to control the resonances of a box structure. The balance force concept is a major advance and I did not invent this myself. I first heard of it from KEF, a brilliant idea and boy does it work. The enclosure design otherwise is my own. The problem for commercial manufacturers is the complexity of the design increases the labor involved by at least a factor of 10 pushing the retail price up into the ozone. I am sure they could find some efficiencies, but still. I published the pictures so that DIYers with woodworking skills could make their own versions. The key design features are balanced force and cylindrical enclosure. My favorite thought experiment involves cutting 15" diameter aluminum pipe to length, welding on ends and feet then getting the whole affair anodized. I have no idea how to weld aluminum but the hired labor cost would be minimal. The problem with a perfectly round cylinder is it is rather bland from an aesthetic standpoint. The decagon with facetted ends is far more interesting. 

Great job, Mike!  Your fine engineering, design, and workmanship inspires me to get out to my workshop start a new project.  But after seeing your workshop, I think that I need to do more than just a few updates first to it to say the the least.   Everything looks great, and I would sure love to hear it one day.

@deep_333 technically you are correct, however if you can not hear or feel any resonance, they are from a human sensory perspective, resonance free. Any resonance frequency is well beyond the subs operational limit and extremely well damped. I am speaking about the enclosure only. The drivers do have their own resonance frequency which I think is in the 50 hz region. I do not have the equipment to measure it. Run with a flat sine sweep at 6 inches they are down 3 dB at 20 Hz with a very smooth curve. Using room control and digital EQ they are plus 15 dB at 20 Hz, remain flat to 60 Hx then drop smoothly to zero dB at 100 Hz. This is as measured at the listening position and by design. The crossover is at 100 Hz 10th order.  As long as the drivers are capable and there is ample power you can make a sub do anything you want with digital signal processing. 

The title is just a tad misleading. There is no such structure that’s "Resonance Free". If a structure has mass and stiffness associated with it, there are resonant frequencies. You may have pushed it out of the sub’s operational range or not (test or sim tool could let you know), but, there are some resonances around..

You can either dump your CAD and material properties for the wood (modulus, poisson’s ratio, etc) in some FEA tool and all resonant frequencies can be computed. You will need to know how to work with FEA tools in this case.

Or get a DAQ and a couple of accelerometers... Start sweeping the driver and determine if any resonances can be caught (coarse). You will need to know some basics of signal processing in this case..

@phusis No argument from me. Dirty bass messes up everything, no bass kills any elusion of a live performance. I might point out that you are doing exactly the same as me. You use a slower crossover but counter by lowering the crossover point to keep the sub out of the midrange. Exactly what slope sounds best I think would depend. on the main speakers you are using. Such steep filters are not good at higher frequencies. I cross from my bass transformer to the treble transformer at 500 Hz 2nd order. I stepped into the twilight zone by bi amping the Sound Lab's transformers using a digital electronic crossover. My initial experience included totally trashing a pair of JC 1s, right into the dumpster. 

I do not think I am adverse to large loudspeakers, but I only have so much room on a 16 foot wall which is mostly taken up by a 113" diagonal screen and two 36" wide loudspeakers. Two of the subwoofers had to go outside the panels to create the line source and I only had 15 " left for each sub. 12" drivers just made it!

@mijostyn --

Looks to be excellent craftsmanship. 

Size is always an issue for aesthetic reasons.

It is, with all that typically entails. However, large size subs (or speakers in general) is not an insurmountable challenge to the minority who wills it, just like you willed it through with great effort via another path from a smaller size factor (in multiples) to keep even minute enclosure resonances at bay, not to mention the magnitude of the added weight of such sturdy built cabs with two woofers per unit. 

The best solution is to cross out at 100 Hz, but you have to use a very steep curve or you wind up with subwoofer in your midrange which is poison. I use a 10th order filter, 48 dB/oct. The test is listening for voice in the subwoofer system, there should not be any or you smear the image. With shallow filters, even 18 dB/oct, you have to cross out below 60 Hz to keep the sub out of the midrange.

I find using 8th order slopes (i.e.: 48dB/octave) to produce a less desirable sonic outcome than 6th order L-R dittos (36dB/octave), but I also prefer using the same slopes throughout the entire frequency range actively, and I cross a bit lower to my subs - just below 85Hz. At the end of the day the ears are the final judges here, and context (incl. personal taste) is a factor as well. 

Another advantage of having your subwoofer system active from below 100 Hz is increasing the gain on this segment between 10 and 100 Hz allows you to get the feel of a live performance without having to resort to ear damaging volume levels. 

Also: the cleaner the bass the more gain can typically be applied without tipping the balance, and this way one can more readily appreciate the physicality of reproduction that's afforded here. Having the reminder of the frequency range configured actively gives you even further options to work with. 

@curiousjim Thank you Jim. I wish to reiterate. There is nothing I can not do with a band saw and hand tools, both powered and not. The big machines just make me faster. For me the fun part is what I call process engineering. How the heck am I going to do that?

@terry9 Ahmen 

@mijostyn,

You have no idea how jealous I am of your workshop! 🤗 It’s beautiful, well equipped and large! Your workmanship is impressive as well. It was a pleasure to look at all the pictures you posted.

Jim

And DIY forever!

@terry9 Go for it! IMHO there is no such thing as too much power for subwoofers. With the amount of gain required at 20 Hz and the corrections room control makes some frequencies are going to be plus 15 to 20 dB. That is 6 to 8 times the amount of power required!

Mini DSP makes a very inexpensive digital subwoofer crossover. You might want to try playing around with one. You can get slopes up to 96 dB/oct! Even if you do not use it in the end it will help you determine what the best crossover points and slopes are.

ESLs are the best! 

Subs are also a dipole isobaric, premium 10" Scanspeak woofers, crossed at 50Hz, 18dB. Both DWM’s and Scanspeaks are unusual in having good transients and linearity extending more than 4 octaves above the crossover point, so that the transitions are pretty seamless (to me, another Quad user, and a Magnepan owner). Sub drivers are also Bryston powered - bass doesn’t need more, but it also sounds a lot better than less. IMO.

You are so right about active control over the bass. Would never go back. Both of my bass systems are actively controlled in terms of volume, but only the sub has an electronic crossover.

I use four modern Quads, modified. They are crossed over at 18dB, 200 Hz (factory), but they need stabilizing, which is easily done with a platform box of baltic birch with a non-resonant top. Each box is stabilized with a Bryston mono block on the lower shelf, so I use the mass instead of it using me. Soon the DWM's will be further stabilized by Al bracing, a la Mye.

I am building an active crossover, must get back to it.

@terry9 Thanx!  I have never heard the DWM. I assume you are using Magnepan loudspeakers. Where does the Isobaric sub come in?

It is important to note for the gallery that properly made cylindrical subwoofer enclosures do not require any bracing as they are inherently extremely stiff. Another interesting trait of the decadron cylinder is if you travel around the long axis in radians the enclosure wall thickness is constantly changing from 1 15/32" to just over 2 " at the joints. The enclosure wall dampens itself.  

Most impressive in all respects. I just finished my bass system consisting of 2 Magnepan DWM's for the lower 3 octaves and an isobaric sub, which I thought was heroically braced until I saw your design. Nicely done!

@clio09 The swarm is built to a price, but I agree crossover flexibility is an important factor when is comes to integration. The most flexible by far are digital.

@bdp24 You won't get an argument from me. Roger Modjeski was absolutely right. Big flat panels have low frequency resonance issues. Sound Labs deals with it by using, I think it is 10 different sizes of panel. Acoustat used felt pads on the back side of the panels to dampen them. The best solution is to cross out at 100 Hz, but you have to use a very steep curve or you wind up with subwoofer in your midrange which is poison. I use a 10th order filter, 48 dB/oct. The test is listening for voice in the subwoofer system, there should not be any or you smear the image. With shallow filters, even 18 dB/oct, you have to cross out below 60 Hz to keep the sub out of the midrange. Another advantage of having your subwoofer system active from below 100 Hz is increasing the gain on this segment between 10 and 100 Hz allows you to get the feel of a live performance without having to resort to ear damaging volume levels. 

With multiple subwoofers placement is not as big an issue with point source systems. Having a line source system I have to arrange them to form a linear array or they will fall behind the main speakers. Another benefit is the bass response throughout the room is very even with a slight increase at the boundaries. 

When I spoke with Roger Modjeski (of Music Reference) about mating subs with the old QUAD ESL loudspeaker---of which he was a huge fan and owner, he recommended employing a crossover frequency of 100Hz, with 4th-order filters in both directions (high pass and low). He maintained that the bass panels of the ESL had a rather pronounced resonance in the 50Hz-100Hz frequency band, and benefited from not allowing the panels to reproduce those frequencies.

The subject of whether or not to use a high pass filter with the main speakers is a matter of some disagreement. Making a seamless transition from speakers to subs is not easy, but having as good a set of subs as possible is of course the place to start, whatever crossover filter characteristics one prefers.

It is my opinion that relieving the loudspeakers the duty of reproducing very low frequencies can greatly benefit the loudspeaker’s reproduction of the higher frequencies the woofer must also reproduce (commonly into the midrange). Employing a crossover frequency of 100Hz with 4th-order filters is a good general recommendation. Finding the best room locations is the next order of business.

Place them where they best address the room’s high and low pressure zones, where bass frequencies either disappear into black holes or "ring" far past the point where the signal has ended, the result of the room’s dimensions creating those zones, referred to as "eigenmodes". Tho location of the "modes" of any given room can be found by entering the room’s dimensions into one of the mode calculators findable via a Google search. If at all possible, do NOT place your subs in those locations.

@mijostyn IIRC the Dayton Audio amp that Duke LeJeune recommends for his swarm has a high pass filter but it's fixed at 12 dB per octave. So not very flexible. It's why I decided to use my own active crossover for the swarm I put together.

@bdp24 Exactly! If you want to get anywhere near a live performance a powerful subwoofer system is mandatory. I might add that approaching realistic levels requires more gain than one would expect. Standard speakers, even large ones can not handle this. Room control, digital EQ and crossovers are important additions. I would even go so far as to say they are mandatory for the best performance. They also make integration a breeze. 

My last set of commercial subwoofers were early Velodynes and like yours the foam surround disintegrated in 5 years. Garbage. That is when I started building my own. As you note there are many roads to Rome. Kits are a wonderful and inexpensive way to build a subwoofer system. Dayton also offers subwoofer kits and they perform just as well as most commercial units even if they are not as sophisticated in the finish department.  Box enclosures do require extreme bracing. You should see how Magico builds their Q series subwoofers, wild. They are also balanced force like mine. KEF and Martin Logan also make balanced force subwoofers, but continue to use box enclosures. Aside of cylindrical enclosures I would also suggest that like Audio Kinesis, they make their electronics packages outboard. They also need to add high pass filters. 

Great stuff @mijostyn.

It is my opinion that most hi-fi’s I hear are lacking the "gravitas" of live music, the massive bass foundation of music heard in concert halls, and even smaller venues like bars and clubs. It is the bass frequencies that provide the physical sensations that are felt rather than heard in live music. When I hear a grand piano live, it makes the reproduction of that instrument on many hi-fi’s sound like a child’s toy piano.

I have a fair number of recordings of pipe organs made in cathedrals (some of them David Wilson recordings), and sufficient bass reproduction is required to create the "shuddering" sound those pipes make when the bass pedals play those very low notes. The bottom note a 32’ pipe creates is located at 16Hz! The lowest note of a standard 4-string bass (whether acoustic or electric) is located at 41 Hz, and many loudspeakers are incapable of reproducing even that frequency at live music SPL.

Before I built my subs I had a pair of HSU’s original sub (the SW10), which was a single 10" woofer mounted on one end of a Sonotube---a cardboard round tube, like those used in making cement pillars. The tubes made production costs inexpensive (the most expensive part was the real wood-veneered top end cap), and the round shape was effective at preventing enclosure flexing, as you noted. Unfortunately, Dr. Hsu used cheap woofers with foam surrounds, and the SoCal air pollution resulted in those surrounds disintegrating in relatively short order.

Before that I used the woofers in the old ESS Transtatic I loudspeaker, which was a KEF B139 woofer (which David Wilson employed in his original WAMM super-speaker of the 1970’s) mounted in an excellent transmissionline enclosure, which as you know are pretty hard for an amateur to make himself. For my sub builds, I didn’t have access to the necessary woodworking machinery (or the skills and experience to use it safely), so I drew up my sub design plans in the manner I learned in mechanical/architectural drafting in high school, and had a cabinet maker cut the MDF and plywood as specified in my drawings. He had a full wood shop, with a table saw, CNC machine, router, etc., and was very reasonably priced.

For anyone considering adding a sub or four to his system (very highly recommended), take a serious look at the Rythmik DIY kits. If I can do it, most anyone can! By the way, I got the ideas for my enclosure bracing after seeing the interior of the subs Jim Salk made using the Rythmik F12 and F15 sub kits. To see them just do a google search for the Salk line of loudspeakers. Unfortunately Jim has retired, and his custom Rythmik subs are no longer available.

@bdp24 Good for you! Keep it up.

Now, put on a bass heavy tune and turn it up. Put your hand on the sub. Any vibration you feel is audible. Ideally, you should feel none. I went through 3 other designs before developing this one including extreme mass (sand and in my case solid surface material, Corian) which did not perform at the level I was looking for. Two important concepts are the cylindrical enclosure which does not require bracing and balanced force topography, opposing drivers. You can get cylinders in all sorts of materials like Aluminum. You can get 15" aluminum pipe cut to length, figure out a way to finish the ends and plant two drivers in it, one at each end. I'm not a metal worker, but I think you could weld aluminum caps on the ends and cut  holes for the drivers then have the whole thing anodized. I thought about coopering cylinders out of hardwood and turning them round on the lathe but because my system doubles as a theater the subs had to be black and I wanted a more dramatic shape then a plain round cylinder. Size is always an issue for aesthetic reasons. I was thinking of using 15" drivers, but the enclosures would be almost double the size and since I am using 8 drivers 12" is more than enough. Because the drivers brake each other you want a driver with a BL product certainly above 20, the higher the better. This also allows you to keep the enclosure small. Rhythmic drivers are made by Dayton by the way. 

Another important concept for assessing bass performance is AB ing the system with headphones. This takes the room out of the equation and lets you know how much detail is in the recording and what the system is glossing over or adding in. 

By the way: Rythmik sells the factory-built F15HP with a 3 cu.ft enclosure, but for DIY buyers recommends a 4 cu.ft enclosure for greater minimally-greater output at very low frequencies. I did 4 cu.ft., which ended up measuring 24" H x 18" W x 24" D. Those dimensions may be manipulated in any way one chooses to create the 4 cu.ft. internal volume. I chose mine purely on aesthetic grounds.

I also built the Rythmik Audio/GR Research OB/Dipole Subwoofer, again with a double-wall MDF/Baltic Birch plywood construction. I built the OB "frame" in the "W/M" style, rather than the more common "H" style. Siegfried Linkwitz also chose the W style frame for his OB sub. I added a brace across the "open" side of the frame, ’cause why not?

Wow, and I thought my design and build was extreme!

I originally was going to make the enclosure for the Rythmik F15HP DIY kit employing the design Danny Richie shows on his GR Research website: a double-wall box, with the space between the two walls filled with sand.

After considering the resulting weight, I instead built a dual-wall box, the inner layer MDF, the outer Baltic Birch plywood, with no space for sand. I then braced the Hell out of it: a 1.5" x 1.5" BB ply brace every 5" in every plane---front-to-back, side-to-side, and top-to-bottom. The bracing prevents the enclosure from "expanding" in reaction to the low frequencies contained in recordings, minimizing the resonance of the walls and raising the frequency of that resonance to way above the frequencies the sub reproduces.

@baylinor The $250 includes labor, tooling, supplies, electricity and rent if applicable. There is also a distinct danger element. Much of the work woodworkers use to do has been taken over by CNC machines with which we can not compete. For me it is the challenge of doing things that mark everything I do as handmade, things that a machine can not possibly do. 

@brunomarcs I think you envy the wrong tools. I can make anything with a band saw, a lathe and my hand tools. All those other machines just buy me time which is important but secondary. I encourage young aspiring woodworkers to start collecting hand tools, get a band saw and learn how to sharpen. The other stuff comes down the line. 

That's quite a project, thanks for posting.  I really envy some of those machine tools. 

$250/hour! I got in the wrong businesses. Beautiful workshop. Nothing like enjoying the fruit of your labor.

@rhg3 That is an understatement! Those woofers used a total of 230 shop hours. The going rate is $250/hour or $57,500!! Materials totalled $2300. Cost to manufacture $59,800. If they were direct marketed they would sell for $120,000. Most people only need a pair for $60,000 and you need a crossover and amps. Very few people will spend that much on passive subwoofers not to mention I have already told my friends with big systems that I am not making any more. Perhaps a manufacturer will see the post and find efficiencies to get the price down low enough to make them commercially viable. All I want is credit for the design. 

@rick_n ​​@audioquest4life Thank you both.

@blackbag20 They cross to the ESLs at 100 Hz 10th order. The first thing you notice is you can hear every bass note clearly. Many systems including my previous designs have one note bass. You hear bass but you can not make out the note or rather each note is the same. What you are hearing is cabinet resonance. Put on a set of headphones and you can hear each note clearly, that is what these subwoofers sound like. I put on Primus, My Name is Mudd which has an insane bass line and tell people to put their hand on one of the subs. They always then put their hand on the driver to make sure the sub is playing! Then, I get this look of amazement. 

Please give a review sometime ...

All the best recovering, not an easy procedure to bounce back!!

Very impressive.  Whishing you a speedy and full recovery.  Cheers!

@mijostyn 

Not only are you an avid audiophile, but also quite the wood craftsman. That is some serious work you did. Your wood shop is amazing. Congratulations on building an impressive subwoofer system. 

Beautiful! Anything worth doing is worth overdoing :-)