Last week I had a very special experience. Multi-site speaker development! Bruno Putzeys borrowed some Altec speakers from a friend. It contains an Altec 421-8H woofer in a cabinet, and the horn on top is an RCF N481 unit inside an Altec 511 horn enclosure. The goal of the speaker development was to do a one-day attempt to make a speaker for the Triode Festival, together with Guido Tent and Peter van Vegchel. The constraints of the owner were summarized by the sentence "surprise me".

Unfortunately I was recovering from a cold, and as Buno lives 1.5 hours driving from my place I couldn't join the team. Therefore we've decided to send measurements to my place via e-mail, and I would develop the speaker off-site, trusting the observations of the people on the spot! First, the speaker was heavily mechanically modified, as there were a lot of prominent panel vibrations. After that, Peter performed some measurements, and sent me the files, and I tried to develop a filter (based on 2nd-order acoustic Linkwitz-Riley response). After 3 hours of iteration, the speakers were transported to Eindhoven, to Guido's place.

The next day I was able to re-measured the units, and found some difference in phase. I re-did the design. For those who are interested, the filter and the characteristics are shown below. The circuit is a simple 2nd-order crossover, with some attenuation to align the efficiency of the units.



The response shows a lot of peaks and dips, which is part of the intrinsic characteristics of the units. The horn and cabinet have a lot of panels, which reflect and diffract, causing these peaks and dips. The first 12 dB, the response follows a Linkwitz-Riley characteristic, and after that the units are filtered very steep, meaning they hardly have acoustic overlap outside their crossover region (at about 600-700Hz). I suspect this to be a good idea with units that have such a different off-axis response due to their enclosures. As the horn is a bit prominent, we've attenuated it a bit more than flat on-axis.



The acoustic phase response shows the Linkwitz-Riley characteristic of the filter even better.



The impedance is tube-amplifier friendly, which is a necessity for a Triode Festival.



The measured response:



The high-end built of the prototype:



I must say I'm pleasantly surprised about the final result. Where in the beginning the whole sounded thin, the overall result is very live-like and homogeneous. It doesn't have the transparency and imaging capabilities of a high-end speaker, and treble is not "airy", but its timbre is very musical and enjoyable, and makes the music very involving. I could happily live with this type of sound at home (but not with the form factor!). All 'n' all this was a very funny project to do, with a surprising outcome. I'm curious about the reaction of the owner.