how can sealed subwoofers produce lots of low fast bass?

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
it has often been said that sealed subwoofers have limited bass in the low frequency region compared to ported subwoofers. Many high end sealed subwoofers such as velodyne and krell have managed to achieve producing lots of low bass (more bass than ported subs and with superior transient response) despite the small sealed cabinet.

how can they achieve this?
Not exactly,

A sealed enclosure is analogous to a second order high pass, and will go really deep, provided you use a really large enclosure. The bass cutoff is gentler but without as much spl in the pass band.
A ported enclosure is analogous to a fourth order high pass. The bass cutoff is much steeper, but the spl is higher in the pass band. Actually this is much easier to draw than to explain.
Many of the Velodyne subwoofer models use motional feedback and a few in conjunction with a mass-loaded passive radiator. If the driver has a very high linear excursion range and high power handling capacity, the frequency response in a given enclosure can be lowered by the use of an appropriately designed filter that lifts the low frequencies upto predetermined levels. Many such alignments are possible: a few examples are, The Bassist from Marchand Electronics, Linkwitz alignment, Multiple Feedback Filter as described by Jeff McCaulay in Electronics World of October 95 (Geoff could correct me about the exact issue, just in case my memory has failed).

Another approach is to design an enclosure too small for the given driver so that the system resonant frequency rises considerably, and then through electronic means operate the driver only below that resonant frequency. Rod Elliot calls it EAS (Electronically Assisted Subwoofer). A company called BagEnd call it LFE (Low Frequency Extension). Rod's project on the said subject can be referred to in his site. Another good reference would be a recent article in Electronics World by Graham Maynard, I think (again Geoff or someone else should come to my rescue). Power handling and Xmax of the driver are of paramount importance in this application.

Many other approaches are possible, but they all invariably require comparatively larger enclosures.

If you want a very small-size enclosure with a sealed subwoofer, and still want a lot of low frequency response, relative to the high frequency response, then you need, most likely, a low crossover point and maybe equalization. Moreover, you most definitely need to increase the amount of power by a great deal, especialy with very small boxes, as a decrease in size always results in a decrease in efficiency. And last, high excursion is a must, unless you are using many drivers.

The Sunfire (the original one designed by Bob Carver) needed THOUSANDS of watts and a large amount of excursion (and equalization?) in order to over come the small enclosure and driver sizes, respectively.

If I remember correctly from reading the white pages on the Sunfire(they took them down when the redsigned the site :( )The driver is a 13inch driver with a 2inch p-p movement. There is also a passive radiator with 1-1.5lbs of weight on it(please correct me if I'm wrong). These coupled with Bod Carvers tracking down converter and a 2700watt amp the size of a candy bar(yes I said 2700watts), give the Sunfire its power.

On the other side I have heard one of the BagEnd 18inch sealed subs and they sound amazing. Get some info on them if you can because they really know what they are doing.



You are right about the Sunfire. I think there were two versions of the Sunfire-True Subwoofer and one used a 10" driver. You are right about the amplifier's power rating.

The BagEnds use Low Frequency Extension and go down to about 10Hz. Less Phase Shift and quick transients is what makes them brilliant. In comparison almost all other types of Subwoofers would sound 'smeared'.

May be its time to try out a design using this principle.
I have always been partial to those accelerometer-based
motion feedback designs, although the examples I found on the web haven't been particularly spectacular. Certainly less so than one might imagine, but I'm confident that it could be quite a successful project.

The most general explanation is this:
You cannot have small size, high efficiency and wide bandwidth at once. You can have two, but not three.
These designs usually sacrafice efficiency or bandwitdh for size. In a subwoofer neither matter that much if you have a spare kilowatt amplifier or want to xover at 60Hz.
As far as I know, the Bag End strategy is the same as what Carver is using. Quite simple, really--use a too-small box on purpose. This causes the sub to be well into its rolloff by the time you expect it to reproduce any sound. The trick is that the response is now totally predictable, a nice, even slope. At that point, you use a nice, even, matching boost to compensate and give the sub flat response. It takes a prodigious amount of power to keep this up for long, which is why Carver has a 2700W amp in his subs. If I recall correctly (it's late, and I'm way behind on sleep), the power requirements are x4 for every octave you drop with a normal sub. Now, on top of that, you've got to account for the amount of power needed to put muscle behind the boost.
It's really only an extension of the idea of taking a normal sub and boosting as it starts dropping out. The difference being that you allow the driver to have a normal-sized box, which allows it to be flat in the pass band until it begins a normal rolloff. The Bag End/Carver subs bypass flat response entirely, manipulating the frequency response curve from the word go.
I've done the second strategy before with stereo subs, each with four KEF B-139s. Boost kicked in as the KEFs rolled off, somewhere (help me here, Geoff) around 40-45Hz. Sounded great at lower volumes, but as you turned it up, you quickly exceeded the Xmax of the drivers and it got very, very muddy sounding. A buddy of mine has Kinergetics subs which, if I understand correctly, do the same thing. They, too, sound muddy as you turn up the volume. The quantity is there, but not the quality
It would be more practical now with Titanic/Shiva/NHT class drivers, since they have a much larger excursion. However, I've given up on the idea and intend to go with servo loops on my current subs, hopefully optical, if all works well. I'm having trouble finding the parts I want. There are other options, of course. Dedicated accelerometers (nifty, but expensive), various kinds of pressure sensors (not as flat in response as I'd like), and the old standby--piezo units (lotsa phase shift). If anyone is interested, I'll report later.


Oh, sorry, forgot to answer the other part--ELF stands for Extended Low Frequency.
diyAudio Moderator Emeritus
Joined 2001
Quote GRollins: "There are other options, of course. Dedicated accelerometers (nifty, but expensive), various kinds of pressure sensors (not as flat in response as I'd like), and the old standby--piezo units (lotsa phase shift). If anyone is interested, I'll report later."

Please report. Also, I have come across mention of the capacitance method of measuring cone excursion, based on the principle that as distance changes between plates, so does capacitance. You might consider that as another option.
The problem with capacitance is that it's inverse square law, i.e. real non-linear. Maybe there's some clever configuration that will get around that, but I wasn't planning on following that option.
A possibility that I didn't mention was to build a miniature voice coil by putting a couple of layers of windings on a drinking straw. Epoxy a small magnet to a standoff and the standoff to the cone of the driver (total mass no more than a gram or two--no worse than the accelerometer chips). Yes, you could put the coil on the woofer if you wanted; advantage--lower mass, disadvantage--then you'd have to get the signal off of the madly jumping cone. Flip a coin and follow whichever strategy makes you happy.
I've also got a few other things I've thought of now and then, but the optical one is still my goal.

Just a clarification: While we are talking about picking up information on the excursion of the cone (for feedback purpose), it is not possible to measure the pressure changes within a (closed) box to obtain and use similar information. I thought I'd just mention this, since this question arose recently in response to which John Watkinson wrote a recent column in Electronics World.

It is also possible to have a second voice coil and pick up information to correct the response curve of a sub-woofer, or for that matter a condensor mic placed very close to the dust cap so that at maximum excursion, the cap does not touch the mic.

I don't know about websites, but there have been a number of articles in Electronics World over the years covering motional feedback designs. One of the more recent ones that I can remember is Russel Bredon's 'Roaring Subwoofer' in the Feb '97 issue.

One common feature of all the articles I have read is the difficulty in obtaining adequate stability in a motional feedback system, which has resulted in the burning out of the amplifier, the speaker or both at least once during the development stage. The most reliable feedback method seems to be the use of a dual-coil speaker with the second winding being used to provide the feedback signal.


Sounds like you have thought about your servo system quite a bit. What sort of servo system are you thinking about?

Also, you keep mentioning optical feedback devices. Are you thinking of DIYing the sensor, or do you have a commercial device in mind?

What sort of Xmax are you dealing with?

The drivers I'm using are the Dayton Titanic 12" (twelve of them). At present, they're in ordinary T-S boxes, so they're good down to upper 20's or so--good enough to get by on, but not deep enough to suit me. I got addicted to subs that dig into the teens years ago; now anything higher than that sounds like sub-lite, if you catch my drift. I've got plenty of SPL capability--far more than I need. Each driver has a rated excursion of an inch or so, if I recall correctly. I don't keep stuff like that in my head, but the specs will be on the website:
As for the optical part...let me see if I can beat this thing into submission first. Then I'll start a thread like I did with water-cooled amps, and the SOZ w/current sources. If optical doesn't work, I've got several fallback positions. Trust me, I'll get a working loop or perish in the attempt. It's just that I've got so many irons on the fire and so little time.

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.