Infinite Baffle An infinite baffle design is defined as an enclosure that contains a greater volume of air than the Vas of the driver. An infinite baffle system can easily be applied to an automobile. This is accomplished by mounting the speakers on a board and using the trunk of the vehicle as the other walls of the enclosure. It is important that there be no leaks that would allow air to move from the front to the back of the cone. Look for speakers where the Qts is greater than 0.6.

One of the most simple designs is closed-box. This is simply a box with a hole in it where the transducer is mounted. A lot of hi-fi cabinets are closed box designs. A rough rule of thumb is that the bigger the box, the lower the frequency response of the system. This is only true up to the point where the transducer can no longer produce lower frequencies. Many old speaker designs used closed cabinets, as the frequency response was fairly good, despite almost no design work needed. It is possible to predict the frequency response of a closed-box, but our applet will not do this.

Closed-box systems are designed around one variable, box volume. Box volume is a function of the driver parameters and the system Q, Qb. To design a system with minimum peak or droop in the passband, Qb must be 0.707.

The designer has the choice of setting Qb and solving for the box volume, or setting the box volume and solving for Qb. There is also the choice of assigning values to both of these variables and solving for one of the driver parameters.

Two engineers by the names of Neville Thiele and Richard Small developed an analysis method that may easily be used to design vented cabinets well. These days most speaker manufacturers will provide a set of parameters for their speakers known as Thiele-Small parameters from which cabinets may be designed. All those symbols on your speaker brochures such as Q_es, X_max, V_as and so forth are Thiele-Small parameters.

This is quite similar to a ported enclosure, except the mass of air in the port is replaced by another cone. This second cone is called, variously, a passive radiator, an auxiliary bass radiator, or a "drone cone". It is, quite simply, a speaker without a voice coil or magnet. Sometimes, the rear spider is also omitted, but PR's with spiders are highly preferred.. As with ported enclosures, PR enclosures are lined but not stuffed. Although not always suitable as a general-purpose port replacement, PR's offer some unique advantages of their own, especially in the frequency range around and below the tuning frequency. For more detailed information on PR's, see the PR FAQ and the DIY PR design tutorial published by Lambda Acoustics, the most serious vendor of PR's for the hobbyist.

A bandpass enclosure is, by definition, simply a sealed enclosure with an acoustical filter in front of it that serves to limit the upper-end of the driver's frequency response. This natural limiting of the high-frequency response of the system makes the selection of mid-bass drivers critical. If your vehicle cannot fit larger midbass drivers (such as a 6 1/2" or larger) then a bandpass enclosure is probably not the best choice for you. Using a bandpass enclosure with insufficient mid-bass reinforcement will lead to sluggish, sloppy, muddy, impact-less low frequency response. In short--it will sound like a soggy pancake hitting a cardboard box.

Once adequate mid-bass reinforcement has been selected to complement the sub-system it will be necessary to add additional electronic filtering to further limit the upper frequency output of the enclosure. Contrary to popular belief, a bandpass enclosure (of any type--single reflex, dual reflex, series-tuned, etc.) does require the use of an electronic crossover to achieve optimum performance since the acoustical low-pass filter is not a very effective filter. What proponents of "crossover-less" bandpass enclosures neglect is that there is a considerable amount of high frequency output (called "out-of-band noise") that can get to be quite annoying.

A bandpass design is a special type of ported box. It contains two or more chambers, any of which may be sealed, ported to the outside, or mutually ported with another chamber. The driver is mounted on a baffle between two of the internal chambers, so the sound can only get to the outside through a port. As with ported enclosures, bandpass enclosures are lined but not stuffed. As their name implies, bandpass designs exhibit both low- and high-frequency roll offs of 18 dB/octave.

The transmission line system is a waveguide system in which the guide reverses the phase of the driver's rear output, thereby reinforcing the frequencies near the driver's Fs. Unlike sealed, ported and bandpass systems, transmission line systems are theoretically non-resonant, and therefore are capable of producing very clean and uncolored bass - if done properly! The transmission line system affords uncolored sound due to the lack of system and cabinet resonances, provides extended bass support, and all at moderate efficiency.

Transmission lines tend to be larger than the other systems, due to the size and length of the line required by the design. Theoretically, the length of line should be 1/4, 1/2 or 3/4 of the wavelength of Fs, however shorter lengths will work if stuffing is used within the line to increase its effective length (sound travels more slowly through the stuffing). Designs that use the shorter (1/4 wavelength) lines generally require more care and attention in getting the stuffing right - but you end up with a smaller box - and greater SAF!

Usually, only drivers which have low Qts (0.25 - 0.4) , Qes (0.3 - 0.4) and Fs values are suitable for transmission line systems.

Think of a TL as a pipe containing a driver in one end and with the other end open. A TL has no tuning frequency (Fb) like a sealed or ported box. "Tuning" a TL is a simple matter of making the line length 1/4 wavelength of where the driver begins to roll off, so that the rear wave can reinforce the front wave. The line may be tuned to either the driver's F3 (-3 dB) or F10 (-10 dB), depending on whose design methodology you use. Similarly, unlike a port, which acts as a Helmholtz resonator, the parameters of a TL are independent, the cross-sectional area being determined by the driver's Sd, and the length being determined by the driver's F3 or F10, as previously noted. Some TL's (notably the Focal Daline series) are a hybrid, utilizing a small enclosure which vents to the outside via a more conventional TL. Even in a Daline, though, the cross sectional area of the TL where it joins the box is much larger than a typical port, usually 1-1.5 times the SD of the driver. TL's are usually stuffed, often with stuffing materials of varying density. Unlike sealed enclosures, the stuffing in a TL is used to reduce the speed of sound through the line. There is no significant increase in efficiency over a sealed box. Response rolls off at a 12 dB/octave slope, just as in open-air mounting or a sealed box.

To contruct a transmission line system, you first need to start with a suitable driver. Suitable drivers have a fairly high Qms (3 to 6), a fairly low Qes (0.30 to 0.40) and a correspondingly low Qts (0.30 - 0.40). Experienced transmission line builders normally stick to a few brands and models of drivers that are known to work well in this type of application.

Choosing the length of the line:
Secondly, you'll need to choose what length transmission line you're aiming for. The theoretical line length will have to be 1/4, 1/2, or 3/4 the wavelength at the driver's Fs. A short list of sample lengths is given below. The lengths are based on a figure of 341m/s for the speed of sound:

•                 line length (metres)
  Fs            1/4       1/2       3/4
  ======================================
  50 Hz         1.71      3.41      5.12
  45 Hz         1.89      3.79      5.68
  40 Hz         2.13      4.26      6.39
  35 Hz         2.44      4.87      7.31
  30 Hz         2.84      5.68      8.53
   

These lengths are in most circumstances much too long to use in any practical speaker design. However, if the line is stuffed with wool or other damping material, shorter lengths can be used. For example, for a 100% stuffed line (wool), the corresponding lengths would be:

•                 line length (metres)
  Fs            1/4       1/2       3/4
  ======================================
  50 Hz         0.62      1.24      1.86
  45 Hz         0.69      1.38      2.07
  40 Hz         0.78      1.55      2.33
  35 Hz         0.89      1.77      2.66
  30 Hz         1.03      2.07      3.10
  

The 100% wool damping has the added effect of damping any resonances that may develop within the line itself.

Based on the above figures, it makes sense to choose a line length that is a bit longer than those given by the second set of figures, then tune it to the correct frequency by adding wool or another similar damping material. until the line is tuned to the correct frequency.

Stuffing the Line:
Fill the 1st 1/6 of the line with long hair wool .5 lb per Ft^3, the next 1/3 with a mixture of 50% wool 50% Acousta-Stuf at the same density, the next 1/3 Acousta-Stuf at the same density, then leave the last 1/6 empty.  Measure the output of the driver using a mike less than 3 inches from the cone, then measure the frequency respons of the port with the mike in the port. The port will usually have significant out put for several octaves. The trick is to damp (stuff) the system until the port output compliments the driver output. As you add stuffing you are attenuating the higher frequencies and lowering the lower frequencies. Wool absorbs higher frequencies, Acousta-Stuf is better at bass frequncies. An example of the measured port and driver outputs is shown below:

And below is the actual measured output of one transmission line system (click on image to see the full size version of the graph):

Tapering the Line:
For the best results, the line's cross-section should taper down from its start point at the driver to its finish at the port. A good method to do this is to start with a cross-section that's twice the driver's radiating area, then tape this down to a port that's 70% of the driver's radiating area. The line can be folded a few times to fit it in a decently-sized box - this will also tend to reduce the line's tendency to resonate at other frequencies and color the output at the port. A sample transmission line showing one method of incorporating the line in the design is shown below:

Bends and end effects:
For a 90 degree bend in the line, the line length is effectively reduced by 0.4*r, where r is the radius of the bend. For a 180 degree bend, the effective length is reduced by 1.0*r. If the end of the line is free, the effective length is increased by 0.58*r, where r is the radius of the opening. For a flanged end (the normal case), the effective length is increased by 0.82*r. If the opening is placed close to a surface (like the example above), the effective length is increased by 1*r.

Classic Transmission LineRules of Thumb
--------------------------------------------------------------------------------


make the line length ~ 1/4 the wavelength of Fs (1/4 wavelength will be the F3 point)
taper the line from 1.25 - 2 Sd down to Sd at the port
after building it, stuff the line -- increasing or decreasing the stuffing till it sounds right. Add if the line is too "boomy", reduce if its too lean. The stuffing increases the length of the line (by slowing the speed of sound down to .7-.9) and (hopefully) damps out higher frequencies.

These rules seemed to work fairly well and were based on the available lit plus the dissasembly of a number of Fried, Rogers & IMF tlines.

Generally drivers somewhat between the extremes (around 0.4 Qt) seem to be ideal for these. Such drivers require some reinforcement in the lower mid/upper bass to counteract the baffle loss and hence benefit from the partial and slight hornloading provided. (more info)

Note that, there is no such thing as an ideal Qt.
Recommended Qt's are:
0.35 for bass horn speakers, fr/Qt should be above 100 Hz
0.7 for closed box designs
0.3-0.7 for bass reflex
0.9-1.5 for TML at fr below 30Hz.