NBDesigns Philosophy & Design

The hardest thing to achieve in speakers, in our opinion - is good bass. As extended as possible, with as good room integration as possible.

The best thing about mass loaded transmission line speakers - is that you can get extended bass from a smaller driver than any other type of speaker, as far as I am aware of.

* The SB Acoustics woofer in the MLTL-6 and Tower 6 - which is nominally a 6", is actually 4.947" - has the highest Mms of any of the drivers I will be using - at 13.8 grams. The Mms includes the mass of the air that the driver moves.

* The Satori woofer I will be using in the Super Tower 6 is nominally 6", is actually 4.875", has an Mms of 12.8 grams.

* The Dayton Audio woofer used in the MLTL-4, Tower 4, and Tower 4T4 is nominally 4", is actually 3.125", has an Mms of 9.9 grams.

*The Sounderlink nominal 4” wideband driver used in the WB4 has an Mms spec of just 2.5 grams.

The mass loading function of the transmission lines I have designed - is a progressive thing. The lower it goes in frequency - the greater air mass inside the cabinet becomes coupled with the woofer cone. I.E. the mass loading varies - and the higher the frequency, the cone is able to work with (much?) less air spring vs sealed and even ported speakers.

The internal baffles in the cabinets form a folded column of air - the section area and length of this column is what Hornresp (8 character abbreviation of "horn response") is doing a 3D acoustic model of; including the quantity and location of polyfil.

Lower Mms of a driver means that it can play music with a greater ability to start and stop the motion of the cone, all else being equal. And the variable air loading in a MLTL speaker can give the speaker a fundamental frequency below the Fs of the driver. A sealed cabinet has a much higher fundamental frequency - in the case of the SB Acoustics 6" woofer, it is about 45Hz above the Fs. And a ported speaker will also have a fundamental frequency about 20Hz above the Fs of the driver.

The MLTL-6 speaker has the fundamental frequency 3Hz below the driver Fs - and the Tower 6 has it 10Hz below.

Midrange clarity was the biggest surprise for me, with MLTL speakers. I think this is because, as mentioned earlier, the woofer has much less air spring/pressure inside the cabinet; and is able to work (close to) how an open baffle speaker does. Especially when you take the Mms of a larger driver into account. The drivers on a recent OB speaker, for example have an Mms of 74 grams (if it is using the Beyma 12BR70 driver).

When a speaker can cover all the frequencies of most music - you get better coherency. The bottom note of a 4-string electric bass guitar is ~42Hz, and on a 5-7 string electric bass is ~31Hz. The lowest note on a grand piano is 27Hz.

The center to center distance between the drivers of the MLTL-6 and Tower 6 (and Super Tower) - is just 5". Compared to many speakers - especially stand mount speaker & a subwoofer - this kind of integration is close to ideal.

Going back to the bass response being the hardest thing to do in a speaker - in an open baffle, the sound waves from the front of the woofer, from the back of the woofer cancel each other to the sides of the baffle. To get the response, you typically need a larger driver. This means, by definition, that it will have a higher Mms - a 12” woofer has an Mms of 74 grams - over 5X more than the SB Acoustics woofer we use.

In a sealed or ported box speaker, the box is there to separate the sound waves from the front and back of the woofer. In a sealed speaker, most of this back wave is suppressed (except for what is reflected back out through the woofer cone). And this is the air spring that the driver (and the amp) have to deal with - starting and stopping the cone motion has to happen with this force.

A ported speaker has output from the port - that is (kind of) indirect from the woofer - it is a resonator that is reacting to the standing waves inside the cabinet. A port is tuned to a narrow frequency band. And it often (always?) has a different phase to the woofer.

An MLTL speaker is using the "actual" motion of the back of the woofer cone, to produce a more broad output, that is by definition in phase with the woofer. So, it is turning a necessity into a virtue, so to speak.

Another important aspect of the lower internal air pressure inside an MLTL speaker - is the reduced stresses on the cabinet panels. And the internal baffles are forming structural braces in the cabinet - that are asymmetrical. They tend to reduce the common mode resonances; and are doing "double duty", in effect.

An important feature of a MLTL speaker, that is related to the mass loading - the motion of the woofer cone is at a minimum at the fundamental frequency. So it is inherently lower distortion.

The main things we look for in a woofer are:
Low Fs
Low Mms
Smooth response up as high as possible - lower breakup modes mean the crossover has less to do.
Higher sensitivity - this usually means 4 ohm drivers as they are up to 3dB higher output for a given power level

The reasons that I really like series crossovers are that they can use smaller value coils - so they have lower DCR & no energy is shunted to ground - all the energy that is available - goes to the drivers. The blending of the drivers is a given - and with fewer parts (in my speakers two coils and two caps and a resistor, plus bypass caps), you can get better quality parts for less money.

How are NBDesigns mass loaded transmission lines different from TL designs that have been built in the past?

We have learned several important things about designing transmission line speakers - the most important change is to locate the driver away from the closed end; by about 1/4th to 1/3rd of the length of the TL. Early TL designs were based on the assumption that in order to get the longest length for a low tuning frequency, the driver needed to be located right at the closed end of the TL. However, this results in a large dip/trough in the response; due to cancellations. This is an inherent problem - and over the decades, and in conjunction with 3D acoustic computer modeling - we now know that by moving the driver away from the closed end, we can avoid the large response dropout - and somewhat counterintuitively, the effective length of the TL is longer than the apparent physical length of the TL - because the sound travels from the driver into the closed end - and then is reflected back out. This essentially doubles the length that the driver is offset from the closed end; making the tuning frequency lower than it would be if the driver was located at the closed end.

Another thing that we learned is that there is no need to have angled panels in the corners of turns in the TL - we thought these were going to make the air flow more smoothly around the turn - but we now know that air forms still “pockets” in the corners where the air can’t/doesn’t flow. We do still need to have radiused surfaces on the inside of the turn/fold of the TL, though.

The mass loading is achieved primarily by “tapering” the transmission line as it gets closer to the terminus opening.

Similar to how we had assumed that we needed angled turning surfaces - tapering the size of the TL does not have to be “literally” tapered. By reducing the size/area of the TL in several steps along the length of the TL, this causes the air to move farther - and as the frequency drops, more and more air along the TL becomes coupled with the back of the driver cone. This literally adds the mass - loading the mass of the air in the TL onto the mass of the driver - and this lowers the tuning frequency of the speaker. In all NBDesigns speakers, the fundamental frequency is below the Fs (free air resonance) of the driver. All conventional speaker designs (sealed/air suspension, vented box/reflex ported, and open baffle, etc) are tuned above the Fs of their driver.

The second method for lowering the fundamental frequency of a mass loaded transmission line - is to place polyfil in the first two segments of the TL - i.e. in the closed end and in the area behind the driver. The 3D acoustic modelling program we use, called Hornresp (short for Horn Response) is able to model polyfil very accurately, and it calculated the precise weight of polyfil needed.

These are generic series crossovers - and as you can see, their driver circuits “overlap” and the components work for both/all drivers. It is important to remember that music is an AC signal - and it always is flowing in both directions, all the time - in that depending on the frequency to positive/top side of the waveform has the current flowing in one direction, while the negative/bottom side of the waveform has the current flowing in the other direction.

The 2-way in the upper left is a 2nd order filter for both drivers. The capacitor is in series with the tweeter (1st order) and the inductor coil is shunting lows away from the tweeter - which adds the 2nd order. For the midwoofer, the coil is in series with the driver (1st order) and the cap is shunting the highs away from the midwoofer, which adds the 2nd order.

The 2-way in the upper right is similar to most of the crossovers that NBDesigns uses in most of our speakers - it has 3rd order filters for both drivers. By adding C2 and L2, these add the 3rd order to each of the drivers. Typically the coils are smaller values vs what a similar parallel crossover would have - so the series crossover will typically have lower DCR (DC resistance) - which lets more power through to the midwoofer.

All the power coming from the amp is going either to one driver or the other. This is how a series crossover can have higher efficiency (lower power losses) and have slightly better dynamics. Also, the “blending” of the drivers is done “automatically” in the circuit. The phase relationship between the drivers still does need to be addressed, but the basic blending of the drivers is controlled by default.

A series crossover uses fewer components for a given filter order, and the coils tend to be smaller values. So, the cost can be lower, and/or the quality of the parts can be higher. We always use air core inductors, and metalized film capacitors - and we highly recommend using bypass capacitors on both/all caps. The quality of the sound of the speaker (as a system) is always going to be better than it would be with lower quality crossover components.

Series crossovers can be 3-way as well - our favorite upgrade for the Sony SS-CS5 is a series crossover, similar to the 3rd order design shown on the lower right. Working in XSim (or equivalent) with a series crossover does take getting used to, as the response for all the drivers is affected by changes anywhere in the circuit - but we think the results of using a series crossover are a better sounding speaker.