If Q113 were a TV series, the episode about this tweeter would be about secret agents meeting in strange places to exchange materials and information. And then it wouldn't be a lie.
Prototypes. Here you can see the first samples of the Q113 Revolution tweeter in turned aluminum finish.
It is somewhat in the air that the tweeter for the Q113 Revolution speaker must dare to break with convention. Its counterpart, the Q113 Evolution, is based on a principle of refining existing and viable ideas, but the revolution must dare to go a little deeper in an attempt to push boundaries. In this way, Q113 attempts to be the stage for the classic showdown between product development that builds on the existing and a risk-based mindset that tries to tread a narrower and less traveled path to discover something new. And it is certainly not a foregone conclusion which of the two principles will win.
Here is the work of the treasury
The tweeter for the Q113 Revolution is basically designed to do the same job as any other tweeter. It must reproduce the high tones of music from about 2000 and up to 20,000 Hz. The task is critical, because it is precisely in this high tone range that musical instruments have their individual sound characteristics, and the human ear is most sensitive between 1500 and 5000 Hz. When the sound in this range sounds so that you can distinguish the instruments from one another without any doubt and understand what is being said or sung without any particular concentration, the pleasure of enjoying music becomes infinitely more inspiring than if the ear and brain have to constantly correct for errors, distortion, etc. In other words, creating a tweeter is a job well worth doing.
It is also quite cumbersome
A really good tweeter should, in addition to low distortion and smooth frequency response, also have a large working range so that the crossover frequency can be relatively low. We want this for two reasons. One is that the woofer for the Q113 Revolution has a significant boost around 5000 Hz, which can best be attenuated when the tweeter has taken over the work far down in frequency, i.e. by means of a so-called low crossover frequency. The second is the speaker's power response. For physical reasons, the woofer's sound radiation narrows from around 2000 Hz. From distributing the sound evenly, above 2000 Hz it is gathered in a progressively narrower beam bundle. This is what we are trying to address. The solution is again a low crossover frequency, so here is another reason to work in that direction. In addition, we would like to avoid edge reflections from the cabinet and if we can achieve a crossover without too many components, we are well on our way. First, though, we need an idea. It came from an unexpected direction.
The first tweeter with DXT. Norwegian Seas was the first to present a tweeter with the Danish DXT lens.
A man introduced himself on the phone
"I would like to put 50-60 hours into Q113 if I can work in peace", said a voice on the phone one day. The voice comes from Jutland and the conversation becomes the first in a series of agreements between engineers who would like to contribute to the Q113 Revolution tweeter, but as a kind of closed club. The reason is obvious. The man on the phone holds a patent on an acoustic lens and he doesn't want to share the recipe for his patent in public. On the other hand, he is sympathetic to the project, so if he can keep his cards close to his chest, he would be happy to provide us with results without revealing the content of his calculations. I am of course interested. This could be an important component for the particular tweeter I am looking for.
Measurement setup with DXT experiment. Photograph from the Jutland basement. In the center of the picture you see the Scan Speak tweeter with DXT lens.
Secret work in a Jutland basement
It is not exactly in the nature of Q113 to be the scene of development work behind closed doors, but anyone can understand why it does not publish the contents of its patent. However, the proprietor is happy to collaborate with a small group of engineers if it can be done without revealing sensitive details. The small group is therefore allowed to isolate itself and in the following weeks some delightful scenes take place that could have been taken out of a spy movie and which I only participate in via a makeshift e-mail correspondence. A Scan Speak employee delivers a box of tweeters to a carport. From here they are transported to a basement room where they are disassembled. Measurements and calculations are made and experiments are carried out with some new front plates for the tweeters. It is the front plates and their geometric shape that constitute the patent itself, and various models are now being milled in the basement room and acoustic measurements are being made. No one knows what is going on in the basement, but the result comes after a couple of rounds where new tweeters are placed in the aforementioned carport somewhere in Jutland. The result is an e-mail with a drawing of an acoustic lens.
Composed of different parts. In short, we ended up using the motor system from one tweeter, along with the diaphragm from the other, and then a DTX lens was designed.
The patent is called DXT
Diffraction Expansion Technology is the full name of the patent, colloquially abbreviated DXT. In loudspeakers, diffraction is about how sound waves are deflected around e.g. edges and generally you want to avoid the phenomenon. It is much more predictable to work with sound coming from the loudspeakers' membranes and avoid sound radiating from the edges of the cabinet etc. This good control of sound radiation is one of the reasons why we use a tweeter with wave guide in the Q113 Evolution. The DXT lens is designed to take the Q113 Revolution a step further and create a form of constructive diffraction that gives the tweeter a larger and more uniform sound radiation. The first tweeter to use the Danish DXT lens was Seas H1499, which you can take a closer look at here. The starting point for our DXT tweeter is a relatively new type of tweeter from Scan Speak.
Up close. The close-up shows the three edges of the DXT lens, which refracts sound waves at different frequencies and spreads the sound around the living room.
You take a 1st class tweeter ...
And many loudspeaker builders, for good reason, would not go much further, because a really good tweeter is a complex piece of acoustic precision engineering that cannot be successfully rebuilt just like that. The typical self-builder therefore prefers to concentrate his efforts on making the fine tweeter work in its design by means of acoustic measurements and listening tests. That makes sense. Here at Speaker Q113, we'll go further than that anyway, because we're in a business where the small margins can lift our project above the crowd if we're lucky and skillful and seek out the innovation. First, let's take a look at the tweeter that forms the foundation of our new design.
A well-built West German
The tweeter for the Q113 Revolution is based on an almost new model from Scan Speak in Videbæk, Denmark. It is a relatively small unit with a highly efficient neodymium magnet, textile membrane and a rather large and well-damped chamber behind the membrane. The speaker is not a classic dome, where the membrane is dome-shaped, but a so-called ring radiator, where the membrane is made up of two waves that sit around each other. The clever thing about the principle is that the movements of the membrane are not controlled at the periphery as a dome would otherwise be. The speaker's voice coil has a grip further inside the membrane, and thus there is less coloring of the sound because uncontrolled membrane movements do not occur so easily. There is a good grip on the diaphragm, so to speak. The moving mass is 0.35 grams, which is in the good lighter end, and with a natural resonance all the way down around 400 Hz, there is a basis in the speaker's DNA for a very large part of the finished speaker's working range to be handled by the tweeter. Here is a link to a look at the unit on Scan Speak's website.
What do we want?
A tweeter that fits our purpose and nothing else. It's as simple as that, but it's easier said than done. This is especially true when the "development department" is in a basement that you cannot visit, and when part of the project's know-how will not leave that basement because there is a patent involved. It is quite something to sit in Roskilde and follow a work that is taking place far away, between people who do not know each other, but fortunately have the desire to solve the task. And by the way, we have an angle of attack. Good tweeters have in common that they are made to fit many types of speakers. Ours should only fit ours, and with the well-built West Jutlanders as a starting point, the people in the basement set about tailoring an acoustic lens.
Frequency response measured by Scan Speak. We have developed an unusual tweeter with a patented acoustic lens.
It should not be a linear one!
We have previously looked at how you can't just put an otherwise good tweeter into a speaker cabinet and automatically expect a good result. In fact, we simulated what happens to the acoustics of the theoretically perfect tweeter when placed in different places on the front of an enclosure. It's not a small thing that can go wrong. If you want to read the post again, you can find it here. Therefore, we are now trying to create a tweeter that does not measure perfectly in a measurement setup. It should only measure well when it is in the speaker.
The first one did not work
The first report from the basement in Jutland was not encouraging. "The Scan Speak tweeter cannot be made to work with a DXT lens" were the words, but nevertheless a basement study was initiated to investigate how the shape (geometry) of the diaphragm affects the acoustics of the DXT lens. This was life-giving for the project, because although I don't know the explanation, the engineers realized that we should simply ask Scan Speak to build us a speaker with a different diaphragm. And then things started to happen.
Analysis of membrane geometry. The shape of the membrane has an impact on its function with the DTX lens. At an early stage, it was analyzed which geometry works best for Q113.
Low crossover frequency and low distortion
The combination of a flatter diaphragm and the newer tweeter's motor and chamber system became the recipe for the basement's further work with DXT. The first requirement for the tweeter is that it must be able to be used with a low crossover frequency, preferably around 1.5-2 kHz. This is partly because the final speaker will have a greater power response and partly because our woofer has a peak at 5 kHz, which we need to avoid. This is accommodated by the fact that the DXT lens is designed with a certain horn effect that has its maximum just below 2 kHz, where the sensitivity reaches 93 dB. When the electrical crossover frequency is later placed relatively high, this tweeter will have usable output down to 1.5 kHz, but with almost no diaphragm fluctuations and thus with relatively low distortion. In addition, we would like to have a good dispersion of the sound at the higher frequencies, because here ordinary tweeters have certain shortcomings.
The very first samples of the Q113 Revolution can finally be assembled after five months of calculations, debates, reader voting, etc. Now it's time to see if we've made something that works .....
Creates useful diffraction
The special feature of the DXT patent is precisely about sound radiation. As mentioned earlier, diffraction in sound is about how sound waves are deflected around, for example, the edges of the cabinet. A phenomenon we would like to avoid, because it is not possible to control what is going on. Here, DXT takes a brave approach to diffraction by creating sound reflections that can be used for something. If you look closely at the pictures, you will notice that our tweeter has three concentric edges, placed at different distances from the diaphragm and with different angles of the intermediate area on the front plate of the tweeter. I see this as the real strength of the patent. These three edges are strategically placed in relation to the tweeter in the center. Using the edges, you can direct the sound into a larger area of the living room. From the measurements you can see that the frequency response of the speaker is very nice and smooth at an angle of 30 degrees from the center, and even at 60 degrees the response is smooth, albeit with a diminished level, which is actually fine. Scan Speak has subsequently verified the results from the Jutland basement. Things work as they should and Q113 has now reached the incredible point that all components have been calculated, simulated, debated and built. We are about to develop crossovers and soon there will be sound on .... It's pretty crazy!
