Electrostatic design - is a round design practical?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello all,
I am new to the forum here and I have my first question. I am curious as to if it would be possible to construct an ESL panel that is cylindrical. I have seen many designs for a curved panel but none that go the whole enchilada and curve into a cylinder. The major benefit, as I see it, would be an omnidirectional dispersion pattern. I am unaware, however, if this design is mechanically impossible or if there are any major negatives that would make it impractical to construct such a beast.

Can anyone chime in here with advice?

Thanks in advance!
Russ
 
One of the guys here did a cylindrical electromagnetic (Magneplaner-type) speaker- there was an article about it in Speaker Builder a few years back. IIRC, there were problems. Certainly, the acoustic transparency of an ESL panel will cause some interesting issues because of the backwave reflection and the interference between opposed sections.
 
Hello Sy,
Thanks for your reply. I would really like to read that article if you or anyone else reading this could offer me a link or info to hunt it down.

I have thought about the back wave issue and it seems that, given a large enough diameter, one could use a properly absorptive cylinder in the center (something looking like one of Michael Green's Tube Traps) to eat the reflections.

My goal here is not so much to marry myself to an EM or ESL design philosophy but, rather, to come up with a very low distortion, acoustically transparent midrange that is completely omnidirectional and covers a range of roughly 100 Hz to 1,000 Hz

Thanks,
Russ
 
That's right!

sreten said:
Essentially you have to tension the diaphragm vertically.

As is done in Martin Logan panels. A full cylinder would
have problems with the back wave and IMO the dipole
nature of a panel isn't something you'd want to lose.

:) sreten.


I did this ones, It didn't work.
You will loos your dipole effect and it'll be a great load to your amp if you make it with a diameter of 20cm or greater.

Cheers,

Audiofanatic ;)
 
sashua said:
Hello Sy,
My goal here is not so much to marry myself to an EM or ESL design philosophy but, rather, to come up with a very low distortion, acoustically transparent midrange that is completely omnidirectional and covers a range of roughly 100 Hz to 1,000 Hz

Thanks,
Russ

For omni directional you will have to enclose the backwave.

As far as I know for electrostatics this simply doesn't work.

:) sreten.
 
O.K. then...lets stop thinking about electrostatics for the moment. Any other designs within the FR that I mentioned capable of a truly omnidirectional dispersion? By "truly", I mean 360 degrees on every axis...essentially spherical. The most promising existing design I have come upon thus far is the DDD bending wave converter offered by German Physiks loudspeakers. It is a capable design but it is not perfectly omnidirectional and it suffers from a slight "tizziness" in the upper register. there are plenty of other pretenders to the throne, such as Ohm, Mirage and Shahinian but none that even come close to the elusive ideal of the "pulsating sphere".

Russ
 
there are plenty of other pretenders to the throne, such as Ohm, Mirage and Shahinian but none that even come close to the elusive ideal of the "pulsating sphere".
Not to nitpick, but I believe the inventor of the Walsh driver you speak of founded Ohm acoustics, making it the originator, not an imitator. ;)

Also, you mention "pulsating sphere", and I've thought of this too.
I wonder - since it's not a stretch for anyone's imagination, who have pictured a "pulsating sphere" - why not mount a large driver, say, a 12" fullrange type driver inside a bandpass-like enclosure, so it's front-of-cone energy were directed through a tube... to the top of the tube, a balloon like orb.
Not a port, the tube's purpose would be to channel the air stimulations such that they stimulated the air inside the orb, causing the orb itself to be stimulated and emit sound.

Possibly the issue is, it would end up behaving essentially like a passive radiator... all the delays inherent in that. :(

But... maybe an experiment with a large compression driver.
Who knows. Just seems like something someone would have tried by now. Maybe they have. ;)

sashua said:
O.K. then...lets stop thinking about electrostatics for the moment. Any other designs within the FR that I mentioned capable of a truly omnidirectional dispersion?
The Decware Radial Loudspeaker is different too... as mentioned.
Simple, DIY-able, which makes it cool IMO - but it doesn't have the thing that makes the Walsh-driver (or DDD) a "wave-bending" driver -
The variable-thickness diaphragm that resonates differently at different locations on the diaphragm by frequency because of the tapering diaphragm thickness. ;)

The original whitepaper is always a fun read. I've got it bookmarked, and every 6 months or so I stumble across it and it still evokes a smile. Creativity, beautiful. :cheerful:

But anyway...
When you start talking "wave bending", and variable-thickness diaphragms, one driver pops to my mind very quickly:
The Manger driver unit. Essentially, it's a flattened version of a Walsh driver, if you can pardon my oversimplification. It's got a variable thickness diaphragm that resonates in different zones at different frequencies. Because the diaphragm motion isn't really pistonic (even though the motor is), I'd expect it to behave more like a natural sound source, omnidirectional.
I wish I had the money to cough up for a pair of these... I'd love to try them in a car even! :cool:

And of course, if you aren't familiar with their work - another thing that could potentially be very DIY-able to mimick - NXT technology's DML technology.
I was fortunate enough to see some fabricated stimulator-based NXT prototypes at DLC labs, when I was fortunate enough to have received an invite to visit their Detroit facility for a little get-together.
By nature, these distributed mode loudspeakers simply make the diaphram they are mounted to resonate, in a non-pistonic way.
If you like, think of it as controlled 'cone breakup'. :D
And because of this, the sound from one of these is also not directional.

Have you played with dipoles? I'm working on my first dipole project... one reason is to obtain that room-filling, somewhat non directional sound - like the Martin Logans... and I am guessing most of that effect is due to the dipole nature of an ESL.
This may be what you desire...
Speaking of them... have you heard any Martin Logans? The Clarities (my personal favorites) are $2600... which is really a relative bargain, as high-end audio goes (of course, being a hard-core DIY guy, I'd never cough up for that personally).
In fact... one iteration of my dipole project is going to be for arrays of small fullrange drivers per side... 3x3 arrays, mounted on curved baffles, to try to really simulate some of that Martin Logan 'room filling' effect - but with traditional drivers.

Some things to ponder, anyway.
 
As far as a “pulsating sphere” design is concerned, have a look that the MBL Radialstrahler 101 E. To explain the design, imagine an orange or mandarin with its segments. Now imagine peeling just the surface of those segments and arranging them back so they reform the surface of the orange. What you then have is a hollow sphere made up of the segments. The top of the segments are attached to a rod so they can’t move, a bit like a core going through the centre of the sphere. The bottom of the segments are attached to the voice coil of a relatively standard driver. In operation, as the voice coil moves out the segments flex out, and as it moves back in tension is applied to the segments and they contract in. Hope that all makes sense. MBL use different size drivers for the different ranges of frequencies as usual.

I have heard the MBL 101 and can honestly say it is THE most impressive system I have ever heard. I know it’s a cliché but it truly is quite unlike anything I’ve heard before. It doesn’t have the absolute resolution of an electrostatic for example, but for sheer “I am there” factor this was it.

I would be most impressed if somebody had a go at replicating this driver, I can assure you that if it comes within ten percent of the MBL systems I’ve heard you’ll consider it time well spent. For info the “segments” of the Radialstrahler look like some type of metal however I would imaging any type of material capable of consistent flexing could be used.

Cheers,
Pete
http://www.mbl-germany.de/english/start.html
 
I wonder - since it's not a stretch for anyone's imagination, who have pictured a "pulsating sphere" - why not mount a large driver, say, a 12" fullrange type driver inside a bandpass-like enclosure, so it's front-of-cone energy were directed through a tube... to the top of the tube, a balloon like orb.
I reckon you could actually use a ballon. I was gonna post the idea but i re read the thread and you already mentioned it. You just have to find a way to enclose a constant volume of air between the speaker and balloon.
I think using an ESL would be near impossible, If you managed to get two half spheres of perferated metalyou still have to tension a constantly curved membrane.
 
360 degree radiation

In a conversation with Roger West of SoundLab, he told me that in experimenting with full-range electrostats he'd tried many different radiation patterns including 360 degrees, but that the 90 degree arc (front and back) sounded the best. So I'm not sure that a full-range 360 degree electrostat would be the best application of that technology.

I believe that RTR made a faceted round electrostatic transducer array, but as I recall it was only operational above 1 kHz.

An array of drivers could certainly approximate omnidirectional radiation, as embodied in the larger Shahinians.

Note that Walsh-type drivers have a donut-shaped radiation pattern at high frequencies - that is, they aren't truly omnidirectional across the spectrum.

The MBL Radialstrahler is probably closer to a true omni with its small tweeter, but the system would be very difficult to DIY.

Henry Wolcott of Wolcott Audio holds a patent (number 4,850,452) for using a 360 degree spherical waveguide structure to approximate omnidirectional radiation. Here's a link to the patent:

http://patft.uspto.gov/netacgi/nph-...,850,452.WKU.&OS=PN/4,850,452&RS=PN/4,850,452

This looks to me like the most elegant approach available to the DIYer. The commercial product is called the "Omnisphere", and differs somewhat from the patent application.

http://www.wolcottaudio.com/WA_omni.htm
 
The MBL Radialstrahler is probably closer to a true omni with its small tweeter, but the system would be very difficult to DIY.

Actually I'm not sure it would be that difficult to DIY. Matching the performance of the MBL yes possibly, but all it needs is reasonable selection of material for the segments. The segments are driven by a conventional voice coil.

Seeing the extent some go to with cabinets I think it's a very DIY project for some with imagination.
 
How about building a sphere from a matrix of alternating mid and high units? Probably have to either customize the frames or research out some more esoteric units to get it to assemble correctly (possibly some of the square framed mid drivers meant for line arrays and then just chopping up the plastic faceplates of some tweeters to fill the spaces inbetween). Then you'd manufacture some sort of bracing rig for the inside, and it would stand on a metal rod attached to that, passing through the bottom. If done correctly, the whole assembly could be sealed, and possibly even finished in a somewhat civilized fashion if one was clever enough, to result in a "globe" bookshelf. Combine with a dipole bassbox or sub, and you should have some room filling sound.

Just a though . . .

Kensai
 
OK, this sounds like fun.

build a square frame out of 1x1's. sized to fit this idea

take two old drivers that you don't care about. Strip them down to just the motors. technically they should be identical drivers

mount the motors on to the frame so that they are facing each other with the voice coils pointing to the center of the frame.

Now take a standard party balloon and blow it up.

stuff the balloon between the two motors so that the voice coils come in contact with the balloon. glue it with whatever to hold the balloon in place.

wire the coils in parallel, hook up some tunes, and have a good laugh. It might actually sound good too!
 
Do a search on Plasma Tweeters. They are probably the truest iteration of what you are thinking about.

However they have drawbacks, such as frequency extension, very high voltage, very little resources available, and they produce ozone.


But they basically use a plasma flame that modulates at the frequency...in <practically> true 360* nature.
 
I have been digesting all the great ideas you super-geniuses have been concocting. Nearly all of them are known to me but so far only the balloon and plasma ideas are in the right ball park. I am actually using a plasma tweeter in my design (David, you win the stuffed animal!). The problem is that the best all my experimenting has allowed me to hope for is extension down to 1Khz.

The problem is "where do I go from there?" If I am to be true to my project's design principles I will need to get from 1Khz. down into the nether regions where bass loses its directionality and accomplish this by means of a truly omnidirectional radiator. multiple driver arrays are all well and good but if you measure them around the perimeter of the speaker there are dips and valleys that look like a skateboard park. Then, of course, since the plasma is a truly massless design I will also need to match its blinding speed with something equally hellish.

Now do you all see my dilemma? Who among you thinks they have the chops to "think outside the box" (pardon the pun) and come up with a solution?

Russ
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.