U18 D2604

Initial page created, crossover modelling in progress....

As I had the second pair of ClassIllu cabinets available, I was thinking whether to make another ScanSpeak construction or to try other drivers. My curiosity won, and I ordered Seas U18REX/P. This driver looks good though there was not much info available and it looked there might be an issues with this driver around 700-1000Hz, right at the area where baffle has big influence. I also had D2604 limited edition tweeters already in my hands. Both drivers fitted the cabinet and drivers cutouts so the decision to use those two drivers was done.

Cabinet had 16l volume and U18 showed to model well in this volume, port tuning was around 40Hz. This will have to be measured, adjusted and listened to come to the final tuning.

D2604 LE datasheet is below. It has aluminium faceplate, wery linear FR and nice off axis responses. Overall very nice tweeter and I am curious to hear it.

Drivers measurements

Measurement distance was 50cm. SPL was calibrated and adjusted so all graphs here show the real sensitivity at 1m. 

I am very pleased with the measured performance of both drivers.

D2604 has nice linear FR all the way up to 20kHz, consistent off axis behavior nicely above 10kHz, and distortion looks very good, no issues.

U18, well, this is something. Well behaved cone shows exemplary FR and off axis responses. This driver does not require any breakup mitigation means at all. 

Crossover modelling

Since the begining of this project I wanted to go with LR2 filters and hoped for very simple crossover.

Tweeter RLC is there to lower the power response bump, listening tests will show if this is needed.

U18 was very easy to work with and to realize nice LR2 slope, and same was the D2604. Phase tracking is excellent and what I expected from tilted baffle and well behaved drivers. Off axis responses are good, and much better then typical 6+1 2way. LR2 filters are responsible for that, at least partially. If I used straight baffle, symetric LR2 filters would not be possible, and off axis responses would look much worse.  

The sound

It always takes some time to let drivers burn in, to get to know the sound of the drivers and to find the right crossover, tuning, and dampening material amount to be satisfied with the final sound. First version of the crossover rarely play as good as it could, so mild changes are always need. This project was also the case. I needed to pay special attention to trebles frequency response, it is silk dome, prone to mild hissing, so area around 6kHz had to be slightly pushed down, and then I needed to find overall trebles level. The second aspect I needed to solve was the level of midrange and mild peak at 500-600Hz. Both efforts paid off and the result is pleasure to listen to and especially U18 became my favourite midwoofer. It can play reasonably low and I appreciate the bass notes are structured, no hints of one note bass. Midrange is articulate and vivid, they have certain effortlessness. Trebless are smooth and easy to listen, I would say delicate, not attracting attention, and well integrated. 

Photos