The aim of this blog is to tell the story of our SA1 speaker; Why we built it, how we developed it and the lessons we learnt along the way. I hope you enjoy it and any feedback is very welcome!
SA1 stands for ‘Small Active 1’ and was intended to fulfill a number of requirements:
- To provide a relatively affordable first step into the CODE sound.
- Have the ability to add bass modules to provide a truly wide band system.
- Provide a super high quality play-back system for those with limited space.
- Provide a compact loudspeaker capable of playing at reasonable volume, whilst keeping distortion as low as possible.
- Create a product that could easily and cheaply be shipped around the world.
- To use as much of our existing components from our more expensive SYSTEM-1 and SYSTEM-2 speakers as possible.
Below is the well reviewed CODE Acoustics SYSTEM-1, which retails at £6,350.
CODE basic design principles
At CODE we’re fundamentally aiming to produce highly accurate sound reproduction systems, capable of playing at high volume with low distortion, whilst producing a neutral frequency response.
If you’d like to find out a little more we have a dedicated blog here:
SA1 is a compact 2-way, fully active sound-system with off-board electronics module; containing four channels of high quality class-d amplification and a Digital Signal Processing (DSP) crossover.
Two compact speaker cabinets are included in the SA1 system with dimensions 260mm H x 150mm W x 140mm D:
The cabinets have an internal volume of 2.5 litres each and are loaded with two 3.5″ Balanced Mode Radiator (BMR) drive units. These are the Techtonic Elements TEBM65C20F-8 (spec sheet link below). We’ll explain the BMR tech a little later in the article.
Click to access TEL-DS-TEBM65C20F-8.pdf
The cabinets are made of 18mm thick deep-route MDF, heavily lined with visco-elastic damping material and Melamine acoustic foam. This produces a pretty inert cabinet, limiting the amount of re-radiated energy.
The vertical edges on the front face are also radiused to reduce edge diffraction.
Our control box houses the DSP crossover module with it’s dedicated power supply as well as the four channels of amplification.
Dimensions are 260mm W x 115mm H x 335mm D.
The user display on the front will show either volume or source. Both options can be adjusted using the buttons on the front plate or with the supplied remote control.
The rear of our control box contains multiple inputs and outputs:
AES /EBU balanced
3 x XLR balanced
<We can adjust the gain of these inputs if you have equipment only capable of line level (RCA) outputs.>
Our control box contains two, 2-channel Hypex NC122MP (spec sheet below) ‘n-core’ amplifier modules. Hypex make arguable the highest quality class-d amplifiers in the world with high power and incredibly low distortion.
The NC122MP will output 125 watts into a 4 Ohm load at 1% Total Harmonic Distortion (THD), both channels driven. If you want to know what this looks like:
Note that for a large portion of the amplifiers power delivery, between 800mW and 80W, THD is below 0.002%, which is pretty awesome! What this gives you is fantastic detail retrieval and just a really pleasant, fatigue free listen.
Our control box contains a 6-channel Hypex supplied DSP module, which is the same one used in our SYSTEM-1 and SYSTEM-2:
This is a very high quality module with dedicated power supply capable of accepting audio up to 24 bit / 192 kHz. It has extensive audio processing capabilities as well as six on board digital-to-analogue converters.
The SA1 uses four DSP channels, leaving two for a left hand and right hand bass module. This is currently a work in progress, but we will provide details as soon as we can.
You might think that designing a big speaker such as our mighty SYSTEM-2 would be significantly more difficult than a compact one, but it’s probably the reverse.
Our award winning SYSTEM-2 prototype at the National Audio Show 2016:
Designing sound systems is all about compromise and this is perhaps taken to the extreme when designing a compact speaker.
Perhaps the most obvious aspect is bass reproduction. We believe that to do this properly you need to move a lot of air gently, which means lots of diaphragm surface area. Something almost impossible to achieve with a compact cabinet.
Compact cabinets mean small diameter drive units, so if you want to move a lot of air you need good linear travel (x-max).
But in a conventional compact 2-way design, asking the mid/bass driver to make large excursions produces intermodulation distortion. Essentially if the drive unit is making large movements to reproduce bass, it makes a hash of the small precise movements to reproduce the mid and high frequencies. This is particularly apparent at higher volumes.
A typical compact 2-way:
So we needed a small diameter drive unit with high x-max, which exhibits low intermodulation distortion at reasonable volume.
This is not a typical drive unit, but fortunately there is a type of driver than can do this …
Balanced Mode Radiators (BMR’s)
A few years ago I was at a friends house and thry had just bought himself a Cambridge Audio sat / sub system. He asked me if I thought it was any good and my initial impression was … yes, that actually sounds very good! One thing that was instantly noticeable was that the sound character was virtually constant as you moved around the room, even approaching 90 degree off speaker axis!
The second thing I noticed was that even at reasonably high volume, the mid and high frequency sound remained remarkably consistent with very little harness. After getting home and doing some research I realised the drivers were BMR’s, which I had read about, but never heard.
Since that time I’d been keen to make a BMR based speaker.
BMR’s were developed here in the UK by a company called NXT, which unfortunately no longer exists. Luckily the expertise remained with the engineers and a reincarnation of sorts has produced Techtonic Elements based in Bedford.
I haven’t got time to cover the detailed technical aspects of BMR’s here, but I’ll point you towards an excellent document on the subject by Naim / Fink Audio Consultancy who used BMR’s in the well reviewed Ovator speakers:
Click to access naim_ovator-s-600_bmr_white-paper_may2009.pdf
BUT very briefly, all speakers ‘break-up’ when the wavelength they are trying to reproduce becomes small relative to the diaphragm diameter:
At this point the speaker has stopped acting like a piston and is distorting. Ideally you just wouldn’t use a drive unit in this range, but with many speaker designs, particularly 2-way’s with low order crossover slopes, you are … and you can hear it! There will be all sorts of phase issues in the crucial 1 – 5kHz range and it just sounds a bit odd. Vocals can be ill defined as your brain struggles to piece the information together.
What BMR’s do is basically counter-balance the diaphragm where it would break up. The common method is to counteract the first break-up node using the mass of the surround (heavier than normal), the second (higher up the frequency range) with a mass on the rear of the diaphram and the third one (still higher) using a foam pad in the centre of the diaphragm:
What this means is you can get a diaphragm to operate over a far wider frequency range than was previously possible with conventional design philosophy.
Once you’ve counterbalanced those nodes the BMR displays quite interesting behaviour:
In fact the above is a motion graphic from the Cambridge Audio website, which you can check out here:
By allowing the diaphragm to bend as above it’s almost like the diaphragm is altering it’s acoustic size; able to act as a woofer, midrange and tweeter all in one.
However, as I always say; audio reproduction is all about compromise. By counterbalancing the break-up nodes you are adding mass to the diaphragm making it less sensitive. What this means is you need power, and lots of it. Luckily at CODE we use powerful class-d amplifiers from Hypex, so this isn’t much of an issue.
You also can’t just put a BMR’s in a box and expect it to give you wonderful sound. BMR’s are complex devices and their phase behaviour once they’re in their ‘BMR’ region is slightly unpredictable, particularly when used in multiples. So it requires someone with a fair amount of talent and experience to extract the best from a BMR speaker.
Luckily we know one!
At CODE we have previously worked with Chris Ellis at CE Electro-Acoustics, who’s an extremely talented speaker design consultant.
He worked at NXT and is one of the foremost experts in BMR technology. We asked Chris if he could help us develop our compact BMR active speaker and luckily he agreed.
Everything falls into place
We were keen to carry-over our top box from SYSTEM-1 (shown below right) with a new front plate as dimension-ally It looked correct for a compact bookshelf speaker, and would reduce development costs and time. We also took the opportunity to radius the vertical edges to reduce edge diffraction.
In case you’re thinking …. why didn’t you just use SYSTEM-1’s top box, there are a couple of good reasons: Firstly the mid-range drive units are £230 Inc Vat each (!) and only have 3mm of x-max. So whilst it makes a fantastic mid / high box, it would become a very expensive and volume limited full-range speaker.
It just so happened that using two of a 3.5″ BMR drive unit Chris developed in our top box simulated particularly well. The plot is not crossover optimised, so don’t worry about the slightly ragged mid response!
We already had our control box from SYSTEM-1 and the only thing required was to correctly match a Hypex n-core amp module to the BMR’s. This was fairly simple as the newly released NC122MP two-channel amp with 75 watts into 8 Ohms @ 1%THD was a good fit.
Dialing the speaker in
We initially tried using both BMR’s running full range, but in the ‘BMR’ range the two units were interacting and blurring the stereo image.
We then tried rolling one BMR off with a first order filter (-6dB per octave), but the BMR’s are so revealing you could actually hear the time related phase issues (group delay).
Finally we ended up with a low pass shelf on the lower BMR and slightly tweaked crossover. It’s definitely not one of the easiest speaker tunes I’ve been involved with, but the end results were most definitely worth it.
It’s fair to say the end results massively surpassed my expectations. It really is a great little speaker (In my humble opinion!).
When I played it the first time for my wife, within 30 seconds of the first track starting, she said ‘I don’t understand, it’s so small, but it sounds so good!’.
The bass is slightly kicked up around 60Hz. By doing this the sound is full bodied and you don’t feel like you’re missing too much in the low bass region.
Timing and punch are excellent and vocal clarity is it’s strong suit. It really does let you ‘see’ into a recording and it’s just very enjoyable to listen to.
Thanks for reading!
This ended up being a bit of a monster blog, so If you made it to the end, well done! Thanks for getting involved and if you have any comments, questions or suggestions please get in touch.
My name is Ceri Thomas and I’m the founder and chief designer at CODE:
Founder and Chief Designer
One Reply to “CODE Acoustics SA1 story”