Electrostats vs conventional drivers

And why did you decide that Acoustat installed felt behind the panel for this purpose? Have you taken measurements with and without felt and found a clear spurious resonance that stands out from the rest?
The Acoustat 1/4" thick felt blocks along the centerline of the panels do a pretty good job of damping the fundamental resonance without causing too much trouble in the midrange.
It is obviously a compromise, but one that served them well. Compare blue line(with felt) vs the dotted line(no felt)...ignore the 60Hz noise spike
Acoustat_TapeMod_LF_summary.png
More details here: https://www.diyaudio.com/community/threads/acoustat-answer-man-is-here.183168/post-4691380

My preferred approach is the use of thin acoustic mesh similar to that used in the Quad ESL-63. This mesh behaves very nearly like a perfect acoustic resistor, damping the diaphragm resonance without changing the midrange response.
EDNF_S2.GIF EDFF_S2.GIF
More details here: https://www.diyaudio.com/community/...ots-for-resonance-control.152979/post-1958582


Regarding the low frequency response of isobaric configuration, adding a second identical ESL panel directly behind the first increases the resonance frequency without adding much if any damping. Theoretically, if there was very little air space between the panels the frequency would increase by a factor of sqrt(2) = 1.414. With Acoustat panels there winds up being about 35mm of air between the diaphragms resulting in the resonance frequency increasing only by a factor of 1.35.
Acoustat_Isobaric.png
 
The Acoustat 1/4" thick felt blocks along the centerline of the panels do a pretty good job of damping the fundamental resonance without causing too much trouble in the midrange.
It is obviously a compromise, but one that served them well. Compare blue line(with felt) vs the dotted line(no felt)...ignore the 60Hz noise spike
View attachment 1044001
More details here: https://www.diyaudio.com/community/threads/acoustat-answer-man-is-here.183168/post-4691380

My preferred approach is the use of thin acoustic mesh similar to that used in the Quad ESL-63. This mesh behaves very nearly like a perfect acoustic resistor, damping the diaphragm resonance without changing the midrange response.
View attachment 1044002 View attachment 1044003
More details here: https://www.diyaudio.com/community/...ots-for-resonance-control.152979/post-1958582


Regarding the low frequency response of isobaric configuration, adding a second identical ESL panel directly behind the first increases the resonance frequency without adding much if any damping. Theoretically, if there was very little air space between the panels the frequency would increase by a factor of sqrt(2) = 1.414. With Acoustat panels there winds up being about 35mm of air between the diaphragms resulting in the resonance frequency increasing only by a factor of 1.35.
View attachment 1044004
Parasitic resonances and their elimination are sometimes random. Having tried many damping materials, I abandoned them completely. I already wrote here that the following is happening to them - I put my head, my tail fell, I put my tail, my head fell. If your ESL is focused on the low-frequency region and you are going to fight parasitic resonances there, then be prepared for the fact that along with parasitic resonances, a useful low-frequency signal, which is already weak in ESL, will also go away. Here is an example of how I struggled with parasitic resonances in ESL subwoofers. These are two packages of isobaric construction, each with three membranes and four stators, as you can see, they are NOT parallel and connected by supple natural leather.
 

Attachments

  • DSC_9429.JPG
    DSC_9429.JPG
    199.7 KB · Views: 69
  • DSC_9415.JPG
    DSC_9415.JPG
    218.7 KB · Views: 72
Member
Joined 2009
Paid Member
The Acoustat 1/4" thick felt blocks along the centerline of the panels do a pretty good job of damping the fundamental resonance without causing too much trouble in the midrange.
It is obviously a compromise, but one that served them well. Compare blue line(with felt) vs the dotted line(no felt)...ignore the 60Hz noise spike
View attachment 1044001
More details here: https://www.diyaudio.com/community/threads/acoustat-answer-man-is-here.183168/post-4691380

My preferred approach is the use of thin acoustic mesh similar to that used in the Quad ESL-63. This mesh behaves very nearly like a perfect acoustic resistor, damping the diaphragm resonance without changing the midrange response.
View attachment 1044002 View attachment 1044003
More details here: https://www.diyaudio.com/community/...ots-for-resonance-control.152979/post-1958582


Regarding the low frequency response of isobaric configuration, adding a second identical ESL panel directly behind the first increases the resonance frequency without adding much if any damping. Theoretically, if there was very little air space between the panels the frequency would increase by a factor of sqrt(2) = 1.414. With Acoustat panels there winds up being about 35mm of air between the diaphragms resulting in the resonance frequency increasing only by a factor of 1.35.
View attachment 1044004
So what exactly is "acoustic" mesh? All I can find is mono filament silk screen mesh. Is that suitable?
 
An example of the acoustic mesh can be found on the Saarti website https://www.saati.com/en/markets/consumer-electronics/acoustics.

Another example of a monofilament mesh is bolting cloth - used to sift flour and other fine powders e.g. https://www.macrokun.com/News/screen-printing-mesh-news/281.html. I believe it is bolting cloth that Quad use in their ESLs.

Both the acoustic mesh and bolting cloth tend to be expensive. The cheapest option is the screen printing mesh, which is available in 25 m lengths very cheaply on aliexpress.

The trick is to get mesh of the correct acoustic impedance - the right balance of thread count and thread diameter. A few years ago I developed the following chart based on the Saarti specifications. Note the confusing mixing of measurement units - imperial (threads per inch) for the thread count and microns for the thread diameter - seems to be the industry standard.

The acoustic resistance required on an ESL depends on a bunch of things but usually somewhere between 40 and 100 rayl is about right. If you glue the mesh to the stator, the resistance of the mesh will be higher. If the open area of the stator is 40%, then the resistance will be higher by a factor of 2.5 (1/0.4). The mesh needs to be tensioned to work effectively, other wise it just moves with the air and does nothing at low frequencies.

Acoustic mesh chart.jpg
 

Attachments

  • 1649882861993.png
    1649882861993.png
    7.5 KB · Views: 52
No. rayl is the unit for measurements of acoustic resistance. Acoustic resistance is the ratio of the sound pressure to volume velocity (analogous to electrical resistance = ratio of voltage to current). The rayl is named after Lord Rayleigh who did some of earliest and most significant research on acoustics.

To determine the acoustic resistance from the chart, you need to know both the thread count and the thread diameter. Unfortunately, it seems that the sellers on aliexpress no longer include the fiber diameter in the spec. When I look at other mesh suppliers, I was able to find 160 count mesh with both 58 um and 65 um diameter mesh probably giving a resistance in the range 10 to 20 rayl.
 
Apologies for the confusion caused by missing details in my post.
The 160 count mesh(ie number of threads per inch) I was experimenting with back in 2009 had thread diameter of 64 µm.
I measured the acoustic resistance to be 18 Rayl, so 36Rayl total when used on front and rear stators.

Rather than attempting to hold a high tension on the mesh, it is easier attach it to the stator or closely spaced supports. Like diaphragm resonance, once you get the unsupported dimensions small enough you don't need much tension to keep resonance well above the bass range where the damping is most needed. I haven't tried it, but I bet if you glued mesh to the Acoustat stator frame with its unsupported dimensions of about 1/2" x 1/2" there should be no problem getting uniform LF damping by just gluing it down smoothly.

I had posted some acoustic resistance measurements for Acoustex 032 mesh in a test section having unsupported width of 2".
It didn't take much tension at all to achieve resonance above 200Hz which would provide adequate damping for a full range ESL. With high tension, resonance could be pushed above 1kHz!
Mesh_Tension_Trends.png
Originally posted here: https://www.diyaudio.com/community/...ots-for-resonance-control.152979/post-5978781

Attaching the Acoustex 032 mesh to a perforated sheet with 30% open area, and you can see the increase in acoustic resistance mentioned by golfnut.
Mesh_on_Perf.png
 
Hi,

the double membrane principle was described in a Sony patent from 1983.
It claimed a up to +6dB increase in SPL between fo and ka<1.4.
Their proto utilized a 3-stator-2-diaphragm setup, similar to what ML later implemented in the CLX bass panel.
As mentioned earlier, this setup can only be used to a certain frequency limit due to the increasing phase shift of early high frequency reflections between the membranes.
The resulting phase shift and its influence on the amplitude response can easily be calculated and measured.
The higher in frequency the working range of the panel shall be the smaller must be the inter-diaphragm distance.
If You can keep the distance below a few mm (even better <1mm) there will be no audible drawback.
I implemented such a system into my ESL Mk2 and indeed could measure a up to 6dB increase in SPL just where it counted, with no ripples at the upper audio range.
While a single diaphragm panel of the same size clearly exhibited a drop in SPL due to the acoustic short the double diaphragm panel measured linear down to ~200Hz @4m distance.
The increase in SPL as well as having not to eq the panel (besides the proverbial base resonance notch) resulted in a very considerable increase in maximum output level and sound quality.

jauu
Calvin

If I understand your message correctly, dipole cancellation frequency / roll-off changes for a double diaphragm panel of the same dimensions compared to a single diaphragm one? I would think that the frequency respone stays the same but the double diaphragm gets a +6 dB boost over the entire frequeny range compared to the single diaphragm driver of the same dimensions. Am I wrong?
 
Hi,

the boost gets less with rising frequency.
So it´s helpful just where it´s most needed. ;)

jauu
Calvin

Thanks for your reply, that sounds interesting. But if I'm correct your panels are not small, they look both tall and wide and as capacitance doubles when applying two diaphragms per esl, together with smaller d/s distances like 1mm max. as stated in your previous post, I would expect a rather high total capacity and therefore expect reduced HF range because of LC resonance and to compenstate for raised panel capacitance you could lower step-up ratio of the transformers?

Which is not an unwelcome thing as these will be cheaper and easy to source I guess. (cheap toroid power transformers). If my idea is correct, in your case (a mid / high esl panel with dynamic line array woofer) will boost mid range of the esl panel (compared to a single diaphragm esl) and needing cheaper (smaller) transformers, but this will not necessarily boost output as you need to lower step-up ratio compared to single diaphragm esl?
 
Hi,

no, the capacitance doesn't double because of the second diaphragm.
With a constant charge ESL the stator-stator distance counts .... the diaphragm plays no part in the equation.
The difference in panel capacitance is almost negligable in my panels, about 1nF for the smaller (125x25cm) and about 1.8nF for the bigger (150x40cm) panel.
The associated LC resonances with simple power toroid pairs of 230V/6V resp. 230V/9V were at ~21kHz resp. 19kHz.
Trannies were the same as for the single diaphragm panels.
With the specialized Amplimo toroids the LC resonances were higher, both >>20kHz.
To 'tame' the resonance a primary side resistance of 0R6 to 1R0 sufficed .... of which the output trannies of the Kronzillas made up for 0R5-0R6 already.
The possibility to apply for rather low 1:U values not only allows for the utilization of robust, cheap standard power toroids, but the lower associated voltages improve matters considerably regarding ageing, reliability, losses, dynamics and sound quality in general.

@DiscoPete : I don't know which link You refer to

jauu
Calvin
 
Last edited:
... the capacitance doesn't double because of the second diaphragm.
With a constant charge ESL the stator-stator distance counts .... the diaphragm plays no part in the equation.
The difference in panel capacitance is almost negligable in my panels, about 1nF for the smaller (125x25cm) and about 1.8nF for the bigger (150x40cm) panel.
Adding constant charge diaphragms doesn't increase capacitance, but going from a 2 stator to a 3 stator configuration does.
Can you clarify if you are using a 3-stator-2-diaphragm arrangement as you implied in Post#318? Or are you using 2 diaphragms placed between 2 stators.

The panel sizes and capacitances you mention are indicative of only 2 stators not 3.
But you mention using 2 diaphragms and also gettiong 6dB increase which is only possible with the 3 stator configuration.