This will be pasive crossover version of previous project with OctoDAC+nanoSHARC:
https://pkaudio.webnode.cz/ellips-a/
First of all, measurements were done in a very large room, speakers were lifted 1,6m above the floor. I also used thick damping material on the floor to mitigate floor reflections.
This provided very good conditions for the measurement, therefore I did just apply 1/12 smoothing and did not do NF merging. After many experiences with merging, I started to look at the smoothed measurements as well. Measurements were done for 0 to -180deg, in 5deg steps, but for better visibility I show just 0-90deg with 10deg steps. Vertical measurements copied from horizontals (Vituix feature).
I had a lot of time to plan and model crossover for this pasive version, and now I am very happy to see it was worthwhile.
Horizontal polars are nearly perfect.
Spinorama graphs, especially power response with no dip at the crossover, and monotonically rising DI look good. Good to see that all features (driver arrangement, crossover slopes, WG, roudovers, etc) all locked in, work in accordance and result in perfect graphs.
From active version I know there is a potential for excellent sound as well.
Crossover parts and cabinets should arrive ~next week...
I visit ASR forum from time to time, to see what loudspeakers are the subjects of worship over there. It seems these Ellips-P could stand among the best measuring loudspeakers, it would be interesting to see how this project performs on Klippel NFS...
https://pkaudio.webnode.cz/ellips-a/
First of all, measurements were done in a very large room, speakers were lifted 1,6m above the floor. I also used thick damping material on the floor to mitigate floor reflections.
This provided very good conditions for the measurement, therefore I did just apply 1/12 smoothing and did not do NF merging. After many experiences with merging, I started to look at the smoothed measurements as well. Measurements were done for 0 to -180deg, in 5deg steps, but for better visibility I show just 0-90deg with 10deg steps. Vertical measurements copied from horizontals (Vituix feature).
I had a lot of time to plan and model crossover for this pasive version, and now I am very happy to see it was worthwhile.
Horizontal polars are nearly perfect.
Spinorama graphs, especially power response with no dip at the crossover, and monotonically rising DI look good. Good to see that all features (driver arrangement, crossover slopes, WG, roudovers, etc) all locked in, work in accordance and result in perfect graphs.
From active version I know there is a potential for excellent sound as well.
Crossover parts and cabinets should arrive ~next week...
I visit ASR forum from time to time, to see what loudspeakers are the subjects of worship over there. It seems these Ellips-P could stand among the best measuring loudspeakers, it would be interesting to see how this project performs on Klippel NFS...
Attachments
Last edited:
Nice work
I imagine you would be told the speaker is too big to be lifted onto an NFS without a team of movers. My guess at the analysis, slight directivity error and avoid ceiling reflections and/or use a thick rug 
Oh for certain you would be told you have a slight directivity error... and you would be told that the predicted in-room response is almost acceptable, but it does not follow the correct slope.My guess at the analysis, slight directivity error and avoid ceiling reflections and/or use a thick rug
Very nice job Pida. I really liked the active version of this speaker, I would love to audition it... too bad it is a 12 hour plane ride away from me.
I spent some time, again, optimizing the crossover, this time not to achieve or improve the curves, but to minimize the capacitor values in midrange part. I was pleased it paid off and I managed to lower needed capacitors by 22+10uF. The overall FR, phase tracking, etc remained almost identical to previous version I started with.
Attachments
yes, Superior 22uf costs 85eur, 10uF 45eur.
I like the sound character of Superiors, it brings clear improvement over Bennics etc. There are other even more expensive choices, but I do not like the prices, and audibly, they are not clear improvement over Superior, rather different, little bit. If it was just one capacitor for tweeter crossover it would be bearable for the pocket, but midrange crossover with Fc ~300Hz is demanding for overall needed capacitance.
I like the sound character of Superiors, it brings clear improvement over Bennics etc. There are other even more expensive choices, but I do not like the prices, and audibly, they are not clear improvement over Superior, rather different, little bit. If it was just one capacitor for tweeter crossover it would be bearable for the pocket, but midrange crossover with Fc ~300Hz is demanding for overall needed capacitance.
Last edited:
After 2days listening and mild crossover tuning, the following curves seem to be the right balance. Of course it will take yet some time to confirm that.
I started with +0.5dB (compared to graphs in the attachment) in >2kHz (easy change of the first capacitor in treble crossover from 1.05 -> 1.1 uF), this sound very good, though after ~2hrs listening trebles seem to attract too attention. It does not cause listening fatique though. So I tried to go to 1uF, which brought the area >2kHz down by 1dB, the overall balance shifted to lower midrange and bass, all the details of the string instruments are still there, classical music got more weight, though some sparkle starts to missing and I felt it overall sounded little bit treble shy. So it seems 1.05uF and the graphs below will be the right balance.
Midrange is what really shines here. 18WE is veeery expensive and I would like to say there are other, less expensive drivers with the same or even better performance. Of course I did not try and listen to all midranges, but I listened and worked with a lot. So far 18WE is for me above all of them (Revelators, Illuminators, Satori P and TX, ATC middome, Accuton ...just to mention few). I would like to try Purifi midrange some day.
I started with +0.5dB (compared to graphs in the attachment) in >2kHz (easy change of the first capacitor in treble crossover from 1.05 -> 1.1 uF), this sound very good, though after ~2hrs listening trebles seem to attract too attention. It does not cause listening fatique though. So I tried to go to 1uF, which brought the area >2kHz down by 1dB, the overall balance shifted to lower midrange and bass, all the details of the string instruments are still there, classical music got more weight, though some sparkle starts to missing and I felt it overall sounded little bit treble shy. So it seems 1.05uF and the graphs below will be the right balance.
Midrange is what really shines here. 18WE is veeery expensive and I would like to say there are other, less expensive drivers with the same or even better performance. Of course I did not try and listen to all midranges, but I listened and worked with a lot. So far 18WE is for me above all of them (Revelators, Illuminators, Satori P and TX, ATC middome, Accuton ...just to mention few). I would like to try Purifi midrange some day.
