That arrangement was a suggestion for CharlieM’s specific situation where he was using sheet metal stators and considering trying segmentation. In your case where you have separate panels for each section, it would be better to drive each panel section separately (both front and rear) like the Spectra segmentation or any of the other segmentation schematics in this thread or CharlieM’s segmented panel build thread.
http://www.diyaudio.com/forums/planars-exotics/267694-thinking-about-segmented-wire-stator-esl.html
Yes, the ladder resistor network connecting the segments progressively attenuates and delays the highs more and more as you move down the ladder to additional segments.
Note that for a segmented ESL, the dispersion can never be any better than that given by the first segment. In your case, the first segment is the 9” wide middle panel. So, using segmentation resistors for the outer panels, the dispersion will not be better than an Acoustat 1+1...in other words, still pretty beamy but not nearly as beamy as if you hadn't used segmentation. The Model 3 angled the outer panels to improve dispersion. This approach actually gives better dispersion in the top two octaves than segmentation would with the same set of 3 panels. However, below the top two octaves the segmented configuration will have better dispersion.
If interested, I can post some sonogram dispersion plots of the different configurations so you can see how they compare.
Last edited:
![20150704_162245[1].jpg](/community/data/attachments/444/444637-95efcc0bfce823b5fff7541e6e0bac97.jpg)
Hello James,
It's been a while since you posted anything about your speakers and I'm curious about your impressions; especially as regards treble response with the segmentation arrangement you used. Is there anything you would do differently if building another speaker?
Beautiful work!
Charlie
Sorry for not replying for so long. After so much trial and error (and sanding, and sanding, and sanding...) I was just relieved to get these done.
My impressions are nothing but positive. They sound absolutely fantastic. I am absolutely shocked with the polar response, though. I have a very short seating-to-screen distance in my room (it is wide and short) so my seating area has a huge angular range. To my fortunate surprise most of the seating area gets a near flawless FR. In fact the only issue that you can notice is a loudness variation due to being so much closer to one of the two speakers.
I have attached some pictures of them in place in my family room. I must say that using MiniDSP plate amps was a wise choice. They have more than enough power, and the outrageous amount of EQ and Xover options allows for fine tuning. Using a decent mic I was able to get these things to measure ruler-flat.
Going back I don't think that I would do much differently regarding the segmenting. I certainly made mistakes in machining and coating that were painstaking and required a lot of do-overs.
My impressions are nothing but positive. They sound absolutely fantastic. I am absolutely shocked with the polar response, though. I have a very short seating-to-screen distance in my room (it is wide and short) so my seating area has a huge angular range. To my fortunate surprise most of the seating area gets a near flawless FR. In fact the only issue that you can notice is a loudness variation due to being so much closer to one of the two speakers.
I have attached some pictures of them in place in my family room. I must say that using MiniDSP plate amps was a wise choice. They have more than enough power, and the outrageous amount of EQ and Xover options allows for fine tuning. Using a decent mic I was able to get these things to measure ruler-flat.
Going back I don't think that I would do much differently regarding the segmenting. I certainly made mistakes in machining and coating that were painstaking and required a lot of do-overs.
Attachments
![20151101_105454[1].jpg](/community/data/attachments/477/477967-f20a44aa20e10a66f62e492b74d24bc9.jpg)
![20151101_105632[1].jpg](/community/data/attachments/477/477993-5affc3b3e48ae8ebe05bcf98bc9b5ac5.jpg)
![20151101_105910[1].jpg](/community/data/attachments/478/478040-d6196ead3a9a417f292b0b4ae359883e.jpg)
![20151101_105949[1].jpg](/community/data/attachments/478/478071-3874031475824cfbfc67036147fd01d0.jpg)
![20151101_110140[1].jpg](/community/data/attachments/478/478122-3306e9d0de2812b0eaef83a1f8ce2282.jpg)
Funny:
MartinLogan ?Reinvents? the Hybrid Electrostatic | Sound & Vision
So Martin Logan places a DSP in a hybrid esl for room and dipole correction and it is a "Reinvention". Seems that this has been done many times over, including by me....
MartinLogan ?Reinvents? the Hybrid Electrostatic | Sound & Vision
So Martin Logan places a DSP in a hybrid esl for room and dipole correction and it is a "Reinvention". Seems that this has been done many times over, including by me....
I'm with you... I see more innovations coming out of this forum than any of the commercial ESL manufacturers.
Bolserst, coming to this post through your link, I was pleased to find the curved panel as well. A marvellous comparison!
Are you able to say what the parameters of the symmetrically segmented panel were (attachment 3)? I assume this would have been configuration 2 with the central segment the same width as the other segments. How many sections were involved and what was the total width of the panel?
Are you able to say what the parameters of the symmetrically segmented panel were (attachment 3)? I assume this would have been configuration 2 with the central segment the same width as the other segments. How many sections were involved and what was the total width of the panel?
It is interesting isn't it...not at all how some promote the dispersion behavior of curved ESL.
Although some builders of curved ESLs like Calvin don't try to hand-wave away the truth.
"Fact is, that curved panels are highly directive too and that they only differ in directivity to flat panels in the highest one or two octaves."
http://www.diyaudio.com/forums/plan...rs-sound-systems-white-paper.html#post4102166
Sorry about that, I am usually better at labeling my plots.
All three panels shown in Post#137 are 285mm wide.
The symmetrically segmented panel used Configuration 1 with 30 electrical segments(9.5mm); 59 physical segments.
Not sure if you had seen it before or not, but I had posted some comparisons for Configuration 1 & 2 in CharlieM's build thread.
See Attachment#2 & #4 here:
http://www.diyaudio.com/forums/plan...-segmented-wire-stator-esl-4.html#post4184637
If you are using more than 6 or 7 segments the differences will be subtle and limited to the topic octave.
Last edited:
Thanks Bolserst. I had in fact searched for directivity of curved panels but not come across your exposition. I had guessed that all panels were probably 285mm (you are a methodical experimenter) but couldn't quite believe the section size that my calculation gave. Of course, with 30 electrical sections = 9.5mm your simulation is more theoretical than intended to be practical since you have elsewhere suggested there is little gain in going below 3/4" (19mm) physical segments.
Nor had I seen your examples for CharlieM's project, but I have used your segment calculator and Hülsebos's ESL seg ui to display variations which I have considered so I am familiar with these trends.
Nor had I seen your examples for CharlieM's project, but I have used your segment calculator and Hülsebos's ESL seg ui to display variations which I have considered so I am familiar with these trends.
I think I picked 30 sections for the example because that was roughly what the OP was using in his build.
The segment size would then have just been a fall out of using the same 285mm panel width.
I am trying to recall the context of my suggestion about 3/4" segment size, but am failing.
Can you point me to the posting? Thanks.
The reference to 3/4" segment width was post #10 "Personally, I haven't found the use of segments smaller than about 3/4" to improve dispersion because of capacitive coupling between the segments." in http://www.diyaudio.com/forums/plan...ted-wire-stator-esl-simulator-esl_seg_ui.html
Capaciti also endorsed 2cm as the least width of a segment, #10 http://www.diyaudio.com/forums/planars-exotics/81235-segmented-panels.html
"Forget about a 0,75cm segment. There will be no output from such a segment. 2 cm is the least width of a segment."
These two statements have led me to think that 3/4" (c.19mm) would be the minimum practical size of a segment. In the light of your experience to date, would you now think differently?
Capaciti also endorsed 2cm as the least width of a segment, #10 http://www.diyaudio.com/forums/planars-exotics/81235-segmented-panels.html
"Forget about a 0,75cm segment. There will be no output from such a segment. 2 cm is the least width of a segment."
These two statements have led me to think that 3/4" (c.19mm) would be the minimum practical size of a segment. In the light of your experience to date, would you now think differently?
Aaaaah….ok. The context was with respect to practical issues, not theoretical.
I still think this is a good guideline, although not a hard limit, so would probably add some additional explanation.
If you are considering using segments smaller than about 3/4", be aware that there can be diminishing returns with respect to top octave dispersion. The reason for this is that depending on the D/S spacing and chosen fL, the ladder resistor values may be high enough in value that the capacitive coupling between segments starts to bypass the ladder resistors at higher frequencies. This will effectively undo the attempt to keep the top octave frequencies isolated to just the first few segments.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.