Shamal
3-way floorstanding loudspeaker 8"+6"+1"+WG
Drivers: Scan Speak 22W/4851, SB Acoustics Satori MW16TX-8, Bliesma T25B + 5" WGCabinets: net volume 35l for woofer, BRSensitivity: ~86dB/2.83V/1mImpedance: nominal 4Ω, minimum 2.9Ω at 110 Hz
Crossover: acoustical LR4, ~330Hz and 1900Hz
Dimensions: H 1000x W 260 x D 350
Earlier waveguide experiments with 3D printed waveguides with various WG sizes and tweeters showed very promising results, so I decided to use this WG/tweeter combo in the loudspeaker. This time I was decided for 19mm Scan Speak R2004 and waveguide with diameter 150mm. But after experiences with T25B and finding out that throat diameter fits T25B, I could not resist and chose T25B.
Also midrange choice was changed from paper to textreme, as this TX rigid cone could be good sonic match to T25B and I was curious to listen to other than paper cone midrange, especially with no signs of cone-edge resonance.
Drivers measurements
Measurements below done at 2.83V, mic 1m distance, SPL calibrated. No smoothing, gating 9ms. Frequency responses 0-180degrees, with 5degrees steps.
Waterfall setting Rise time 1ms, window 9ms, time range 3ms. I am very pleased to see no problems here, especially tweeter+WG assembly I was afraid a bit, sometimes the assembly is not robust enough and can cause rattling and air leaks.
Distortion measurement of the drivers. Voltage 2,83V, mic distance for tweeter and mid 1m, and for woofer 30cm.
T25B+WG horizontals, waterfall and distortion
WG and large roundovers really pay off. 3dB dip at 12kHz is just small wrinkle in otherwise great set of off-axis responses.
This clearly shows that LR4 above 2kHz is feasible and should not cause any problems to tweeter
MW16TX-8 horizontals, waterfall and distortion
As I hoped for, no signs of cone edge resonance. This is very nice up to 2kHz and should fit LR4 2000kHz nicely. Though higher Fc seems difficult.
Overall all distortion components are very low in the intended pass band 300-2000Hz.
22W Revelator horizontals, waterfall and distortion
This looks really good upto 2kHz, just the resonance at 800Hz, but this will not be a problem with crossover ~300Hz. Cone edge treatment would fix this.
Very good distortion results for 8inch woofer. 800Hz peaks are likely caused by cone-edge resonance, but it is not a problem with 300Hz crossover.
FR measurements of each driver in its axis are below. This time no gating aplied, just 1/12 smoothing.
Preliminary crossover modelling
After experience with Ellips-A project and experimenting with LR2 and LR4 crossovers between woofer and midrange, I decided to go here with LR4 as the version for initial listening evaluation. Modelling of LR2 was quite straight and easy, no issues with components values and overall impedance, but LR4 versions took me much more time and itterations to arrive to version below.
Modelling with data 0-180deg, gated 9ms.
Below is the spinorama with NF port and woofer merged to FF. Just out of curiosity I ran preference rating, and quite pleased with such a high numbers.
Photo gallery
