Discovery10

Making 3way floorstanding loudspeaker with 10" woofer with Discovery drivers have been on my wish list for some time. Cabinet was supposed to be easily built, though I wanted to make it more pleasing for the eyes then large shoe box, so I added black plinth.

The woofer was 26W Discovery 4ohm. I used it already few years ago in similar project and was curious to hear it again. I modelled closed and BR alignments. After Ellips-A project I wanted to try closed box again, 26W modelled well, so I came with 65L cabinets. I considered both scenarios, closed and ported, so plinths were created with the holes for 70mm ID Jantzen ports. I started with closed, but after few listening tests and tuning I installed the ports. Closed had a lot of potential, it sounded almost good, but it was too dry. I missed the body of the instruments. Installation of the ports and damping material adjustment brought the effect I was looking for. The bass and lower midrange was still well defined, though less dry.

18W Discovery as a midrange was obvious choice. 3rd party measurement data promised easy response and easy crossover work.

For the tweeter I knew I wanted to use waveguided tweeter. Quite recently I tried TXN tweeter in Jantzen WG and measurements looked really good. I was also very curious about the sound of this driver. Yes, it is much expensive than Discovery drivers, but my curiority won.

Cabinet wa supposed to be quite easy to build, this time I did not want to make it more expensive going with CNC baffles. Nor I wanted to make sparated baffles, which would bring better off axis performance for both midrange and tweeter. 

To mount tweeter to WG I had to remove faceplate from the tweeter first. This was easy. Tweeter was press mounted with MDF plate to the WG secured with four M4 screws. To install three dome protection elements I needed to measure the hole pattern in the original faceplate and draw the model in Solidworks. When done, I created the drawing and left the drilling of the holes to the professionals. The result is amazing and looks really good.

And the sound? I have to say I was little bit nervous at the beginning. What would be bass quality from this 26W woofer, how about midrange quality? Would both midrange and woofer keep up with TXN? And TXN itself, would it be good, and as expected the character somewhere in between silk domes and Be domes? Hmmm, I was ready to replace TXN with DN - dome if needed. Fortunately it was not, though it took me several listening sessions to get this tweeter sounding right. At the beginning, the overall sound was too forward sounding, with the balance moved to upper octaves and "tsss tsss" character. I had to suppress mild bump at 10kHz by ~1dB. I also needed to fine tune the area around 2kHz and 3kHz. After these adjaustments I started to actually like what I hear from the tweeter. There is one aspect of DIY, you can try and lisen to any driver you want, but drivers alone and also drivers in the loudspeaker, with crossover, sound all different. Sometimes it takes longer time to adjust the ears and the brain to the new sound and even more time to come with the right voicing. Fortunately I found the tweeter sounding as I expected, there was certain hard dome character, the clarity, structure and detail in the trebles, and also top end air and sparkle. TXN does not sound like Bliesma Be, it sound calmer and quieter, smooth and does not attract attention. This character seemed to match 18W Discovery character well, and the result is the sense of transparency and clarity in midrange and trebles. It was not the sense of ultimate transparecy like with some more expensive drivers, it was more to easy to listen side. I am curious to hear TXN + 18W Classic or 18M Revelator in the 3way, maybe some day.

Measurements for the crossover

Measurements were done from 90cm at 2,83V.

Crossover modelling

Crossover is LR4 for all drivers.

To address midrange off axis bunching at 2kHz I pushed down that area a bit. It showed to be a good step, and thanks to the bunching the power response does not exhibit typical mild dip at Fc. Narrower baffle or R30 of the baffle edges would solve the problem, but this was not part of this project.

Photos