ClassIllu

I like the overall simplicity of 2way standmount loudspeakers and if they are designed well they can offer also a lot of good sound. I still recall how much I enjoyed Ghibli and Jurko and even this time I wanted to keep overall look and dimensions. 18W Revelator or 18WU Illuminator would be better choice, but these need volume >20liters, and I wanted to keep the volume requirements on lower side, below 20liters. I also wanted to use less expensive driver. Experience with Xavian XN270 with 18W/8535 in closed box made me to consider Classic line midwoofers so I chose 8545K00. 8535 in closed box was not bad but I wanted more bass and extension to lower frequencies.

8545K00 models well in volumes almost upto 30liters, but from my experience it is better to stay lower with the volume to obtain good bass sound quality. And as this project is standmount and usually smaller is better so I chose 16liters based on modelling, with tuning around 35Hz.

Classic line midwoofers are known for certain smooth, warmer and sweeter sound character and I was little bit afraid of this initially. After some listening I can say Cllasics are little bit darker compared to Revelators and Illuminators, but overall sound is still very enjoyable, has energy and makes sense overall and that I cannot say about all drivers I had worked with. No wonder they are still on a market, I can easily the reasons like firm tight bass, good midrange and nice and easily distinguishable look of their cones. I already mentioned Ghibli project which used Satori midwoofer, I still recall all the qualities, but this scan speak 2way seems to have better bass and midrange quality that is not far behind Satori midrange qualities. Anyway, I would need side by side listening to be 100% sure with my assessment. The bass character reminded me 32W Revelator in closed box, not with the extension to the lowest octaves, but with the energy, texture and certain vivid character. 

My positive impression from 18W might also be caused by the fact I cross over this driver at 1650Hz LR4 acoustic slope. Other implementations usually used less steep crossover slopes and crossoed it higher, and very importantly, there is a dip around 1-2kHz, and that all might contributed to the sound character this driver is known for.

Drivers measurement

After few experiences with measurements and crossover modelling without using of gatting, I decided to follow the same way of working here as well, therefore all measurements for crossover modelling are not gated, and with smooting 1/12. There are drivers measurements with 5ms gating just for the reference and comparison.

One thing I immediatelly noticed was that 18W has very nice, consistent off axis responses with quite low directivity. On the other hand, mild dip between 1-2kHz complicated things. When I did crossover modelling this dip was not possible to compensate with low order crossover. Fortunatelly all things just fell in together nicely and acoustic LR4 showed to be optimal for nice off axis response and 18W FR shaping around 1kHz.

Crossover modelling

Initial crossover modelling looks as following. After the very first listening test this sound really good and promising.

Diffraction signature is there, though it is not as severe and bad as it could be. Anyway, this was expected, and the only way to deal with this is to pay attention to 2-6kHz area and set the level based on longer term listening tests. I could easily model crossover for declining power response, but that is not way to go. It would lead to 2-6kHz area being down by ~3dB and that is too much, resulting in the sound without sparkle and life. So it is not that easy and I always do listening sessions trying this and that in crossover, listening to various songs and genres. There is no easy and simple solution, as always it is about balancing the compromises.

At the beginning I tried LR2 crossover, but this was not much successful, it looked feasible but phase tracking was not good. It just confirmed my initial thoughts that natural response of this driver is declining above ~5kHz and that there is phase rotation involved.

It showed that electrical 3rd order / acoustic 4th order was what was needed. Crossover point is 1650Hz and tweeter is at ease thank to LR4 slope. Additionally, midwoofer crossover gain peak at 1kHz helped to compensate natural 18W dip at the same frequency. This would not be possible with LR2 crossover. 

There is one RLC to compensate tweeter Fs peak and one RLC for midwoofer to flatten ~700Hz bump. Generally 18W classic driver is not easy to work with due to FR irregularities in 500-2000Hz.

Vituix six-packs are below, on the left with no gating + smoothing 1/12, on the right gating 5.3ms + no smoothing. 

For a 6"+1" 2way this is very nice off axis response, mostly thank to very consistent off axis responses of 18W driver. LR4 filter at quite low frequency helps to that as well.

The level of 2-6kHz area had to be confirmed by listening tests and I tried several capacitor values in tweeter section tocome to this final FR and to get balance in the percieved sound. Just going by power response, I would tend to supress that area by 1-2dB, I tried it, and it led to too soft sound, string instruments lost certain sharpness, overall sound was too relax and laid back. I usually test my impressions with the regular visitors knowing the sound of my loudspeakers.

Polar maps for horizontal and vertical planes are below.

And the graph with the raw responses of the drivers and responses with the filters.

Photos