ESL Spacer Orientation

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Working through my first design for wall mounted segmented ESL's and new to the Forum (glad to be here!). I've worked out the desired segmentation calculations and am contemplating the implications of diaphragm spacer locations. One thing I find counter-intuitive is that I see a lot of images of vertical line source speakers using horizontal spacers following the 100*d/s rule. I would anticipate inner modulation problems with the tweeter segment being mechanically forced to follow the large displacements coming from either side. I'm thinking that aligning the support insulators vertically between segments would reduce this by isolating the tweeter. Any thoughts or experiences to share on this?
 
Hi ArneAudio

Perhaps you misunderstand. In a segmented design, at high frequencies, only the membrane close to the central segment is driven. As the frequency decreases, a greater fraction of the membrane is driven, and at the lowest frequencies, the whole membrane is driven.

There is no doubt some intermodulation distortion, but its very much less than you get with conventional speakers.

I'm not sure what you have in mind, but a wall mounted ESL does not usually work very well. The ESL produces sound from both front and rear, in opposite phase. If a surface is placed close to the rear, the surface will reflect any output back straight out the front of the ESL, and will cancel any output. You need to make sure that (ideally), the ESL is at least 1/4 wavelength from the wall at 100 Hz, that's about 0.8m.

regards
R
 
Greetings Golfnut,

Experience is certainly the best teacher. Let me hum a few more bars. My understanding is that a moderate size ESL speaker (1m^2) will have a pronounced resonance in the 120-150 HZ range and most designs will cut over to a traditional bass driver around 300 HZ to suppress that resonance. The alternative is to dampen that peak and extend the bass range below this resonance by applying cloth directly to the back stator. In addition I'm ok with applying acoustic tile directly behind the rear stator. I'm thinking that if I can get the design flat enough and up against the wall that this might allow the reflection and original image to converge. If anybody has tried this I'd love to hear how it worked.
 
So just curious here....

Would intermoduation distortion be less on vertically spaced, vertically wired, segmented panels?

My thinking is this..... If vertically spaced; the diaphragm itself will be segmented vertically along with the wires. Since the outer segments of the diaphragm will not be producing higher frequencies (since the wire segments themselves are producing less and less higher frequencies as they move out from the center), those frequencies cannot physically be modulated by lower ones..... due to the physical spacer that blocks vibration/sound from moving horizontally across.

If you use horizontal spacers and vertical wires, then the higher frequencies in the center can vibrate the diaphragm all the way out to the lower frequencies at the edge (and thus be modulated).


Hope that all makes sense, even if it is wrong (or insignificant) :).
 
Last edited:
Hi Bengel
I don't think it will make any difference. At high frequencies, the membrane just does not resonate because it has so little mass.

I think of it this way...

At any frequency, the Q of a resonance is 2.pi.f.M/R, where f is frequency, M is the mass, and R is resistance (representing losses). The beauty of the ESL is that the mass of the membrane is negligible and the resistance, due to acoustic resistance of the air is very large, so the Q is extremely low - which is to say that any sort of membrane movement decays away very very fast. The membrane has to be actively driven to move at all. This same magic is one of the reasons ESLs sound so clear - they do not suffer from the high-frequency breakup problems of conventional drivers.

This situation changes at low frequencies. At low frequencies, once the wavelength is comparable to the width of the speaker, the air is able to slosh around between the front and rear of the speaker - think about the additional airmass attached to a bedsheet when you shake it. This has two negative effects (i) the radiation resistance falls, and (ii) the airmass attached to the membrane increases. This means that the Q rises rapidly, and we do see significant membrane resonances - entirely due to air mass.

To damp the resonance at low frequencies, we have to provide additional acoustic resistance - and thats what the screen printing (and dustcovers) cloth does. And it does not have to be very much. The acoustic resistance due to air at high frequencies is about 860 rayl, at low frequencies it falls away to stuff all, and 40 rayl cloth is sufficient to bring the Q down to 2 or so.

regards
R
 
My previous segmented panels used vertical welding rod conductors with horizontal spacers and my current segmented panels use vertical wires with vertical spacers. I didn't notice any sonic difference between vertical versus horizontal spacers with respect to resonances and/or distortion; which in my mind confirms Golfnut's assertion above.

I opted for vertical spacers in my current panels solely to minimize the panel's stray capacitance. That is; the portion of the wires covered by the horizontal spacers added to the panel's capacitance but didn't contribute to the sound output.

By using vertical spacers and omitting wires where the spacers lie, practically all stray capacitance is then eliminated. In this case the total wire area produces sound (except for the small area where the wires are anchored at the panel ends, which does not change).
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.