KEF’S JACK OCLEE-BROWN:
driver parameters from the datasheet and look-up tables for the port tuning frequency. I had very little equipment to help me and not too much by way of woodworking tools. But after a few weeks working almost every day, the end product was pretty good – and certainly a huge improvement over the old Sonys.
By the next summer, I was hungry to try again, and designed and built another set of speakers. This time I did a much better job. I built a measurement microphone and preamp, used computer software to design the enclosure box, tune the ports and to simulate the crossover behaviour.
These speakers came with me to the University of Southampton when I started my Acoustical Engineering degree at the Institute of Sound and Vibration Research (ISVR).
During my second year at university, I wrote to every loudspeaker company in the UK to ask for a summer placement
– I remember going through the manufacturers’ list in the What Hi-fi? Buyer’s Guide to draw up a shortlist.
I also mentioned to one of my lecturers that I was looking for an opportunity and eventually got a call from Mark Dodd at KEF who arranged three months working at KEF R&D in Maidstone. At the end of the three months, they asked me to come back the following summer and offered me a permanent job when I graduated in 2004.
My job at KEF is to manage the research and development department in Maidstone. We are a small team of dedicated engineers and we develop KEF’S unique driver and system technology. It’s our responsibility to make sure KEF products sound great.
We do engineering design work on upcoming products, research to develop improved technology that could lead to better sound, and advise KEF’S product planning team on what will or won’t work acoustically in future products. We develop software to help us simulate the behaviour of our loudspeakers and do research into how listeners perceive our products, so we can better understand how to improve them.
It’s my job to coordinate the team, prioritise our tasks and make sure we’re working on the right things to keep KEF producing new and exciting products.
We’re trying to make speakers that are completely captivating, and totally absorb the listener into the music they’re enjoying. Music is truly diverse and, as much as audiophiles like to moan, in most cases produced with deep care and attention to detail.
The job of the speaker is simply to not get in the way, not to show off, or stand-out, or distract. We strive to design neutral speakers that allow instruments and voices to sound natural, that don’t
restrict dynamics or flatten the sound, that resolve the smallest of details, and that spread the sound into your room evenly so you can enjoy the threedimensionality in the recordings.
The starting point for any successful development is the design brief. This is an overview of what we’re trying to achieve – how big it should be, where it’s likely to be used, how it will be used and how much it will cost.
We will set a performance target at this stage too, and this is often done in relation to other KEF products or research prototypes. It’s then the job of the engineering team to come up with a system configuration to hit these targets. This would typically involve a bit of computer simulation, often some early prototyping, and some discussions with manufacturing teams to determine an engineering specification for the parts.
At the end of this stage, we have an overview of the parts required, including the sizes and arrangement of the drivers, the crossover frequencies and the size and location of ports. If it’s an active loudspeaker, we also consider connectivity, as well as the electronics and DSP requirements.
The detailed engineering follows, with every part fully designed and confirmed. This process can take time, particularly for complex parts like a Uni-q driver to be fully designed, simulated and prototyped. The eventual outcome is a number of fully working prototypes that we can assess and listen to.
When we develop a new product, we pick out specific references to make comparisons – the model it will replace, the equivalent model from the range above, and often one of our high-end models too.
Ideally the loudspeaker driver should behave the same for any signal it receives. The trouble is that mid and treble signals require the driver to react extremely fast. The cone must be light and stiff to follow this fast movement or else it will bend and resonance will eventually occur at some point. Typically the best materials, from a lightness and stiffness perspective, are the worst in terms of the severity of the resonances.
In addition there are other factors that have a huge effect on the cone or dome resonances – such as the exact geometry and the characteristics of the attached components, the surround, and voice coil mass and stiffness.