I have, for over a year, been secretly developing a new AudioSmile Supertweeter to replace the current version. [B]Part 1[/B] The new design started for two reasons. Firstly, I often found the current model sounded best toed out from the main speaker to make it fire off-axis. In some cases it also sounded better 'upside-down' like the image below with Kensai. This way the sound comes to the listener more diffused and doesn't intrude on the balance on the main speaker. So I got thinking about a design with two ribbons at angles to each other. A trial with dual ribbons gave even wider and taller soundstage than the single unit. [img]http://www.audiosmile.com/images/audiosmile/Product-KensaiST.jpg[/img] I decided that the shape of the new supertweeter had to be able to work in both an upward and also forward firing position. It strongly depends on the speaker and room which position sounds best. I got stuck into AutoCAD and came up with this basic design. The drawing shows controls sticking out from the end-cheeks but in practice they will be recessed into the cheeks allowing the supertweeter to be stood on end and a channel in the cheeks and large face allow the cables to neatly hide. [URL=http://s608.photobucket.com/user/tenson_uk/media/STCAD1.jpg.html][IMG]http://i608.photobucket.com/albums/tt169/tenson_uk/STCAD1.jpg[/IMG][/URL] [URL=http://s608.photobucket.com/user/tenson_uk/media/STCAD2.jpg.html][IMG]http://i608.photobucket.com/albums/tt169/tenson_uk/STCAD2.jpg[/IMG][/URL] Secondly, I got thinking about making a highly adjustable filter circuit that could be tuned to match almost any speaker. It's certianly not easy to do and I can see why many designs have limited options. Many days simulating and testing came up with a circuit that has 6 volume levels in combination with 3 filter settings - combining these it's possible to choose 18 different filters. A problem was hit when trying to implement level adjustment because a resistive divider would always present a low impedance to the amp. Eventually a method was found using resistors in parallel with different capacitor values such that lower level settings present a higher easier to drive impedance. [URL=http://s608.photobucket.com/user/tenson_uk/media/ST-Schematic.jpg.html][IMG]http://i608.photobucket.com/albums/tt169/tenson_uk/ST-Schematic.jpg[/IMG][/URL] Each of the small groups needs switching in and out, both signal and ground to keep them out of the way when not in use. So it meant having a lot of PCB pads to be soldered to rotary switches. [URL=http://s608.photobucket.com/user/tenson_uk/media/STPCB1.jpg.html][IMG]http://i608.photobucket.com/albums/tt169/tenson_uk/STPCB1.jpg[/IMG][/URL] PCB back from production. It's extremely satisfying when you populate a new PCB, and find it works first time when testing! :) [URL=http://s608.photobucket.com/user/tenson_uk/media/ST-PCB-2.jpg.html][IMG]http://i608.photobucket.com/albums/tt169/tenson_uk/ST-PCB-2.jpg[/IMG][/URL] Stay tuned, more to come!