Electrostatic AMT?

[mige0]

Originally Posted by DavidJanszen

Have you written one for Heil's magnetostatic AMT, Michael? It would be great to get a feel for the membrane forces for that case.

A good idea, David!
But what I see as a problem is that for magnatostatic AMT we do not have reliable data for
- mass
- B*L
- pleat dimensioning

Manufacturers (Mundorf, Beyma...) only provide veeery sparse data.

How should we compare, any suggestions?

Michael

[DavidJanszen]

Originally Posted by mige0

But what I see as a problem is . . .

I was thinking it would just be written to derive the Lorenz force and the foil temperature rise from the essential parameters of the magnet's magnetization or magnetic moment and physical dimensions, magnetic gap width, the current, the foil thickness, dimensions, density, and conductivity, and the pleat depth and gap. I think it seems okay for this purpose to simplify by assuming the static field is at a perfect right angle to the current flow, and also to ignore the dynamics and assume DC current flow. Might be worth searching to see if anyone published such a thing already, though, which I have not done. Seems like it would be a great 2nd semester physics project for a smart kid.

[mige0]

Oh boy !

[Keith Taylor]

Hi All, I think our understanding of the issues involved in AMT's in general, and the ES version in particular, is advancing thanks to your input and stimulus.

When we read the Heil patent he has no problem with f figures of 10, yet in the working version this is halved to around 5, Why?

Could it be that the pleats cannot be creased sharply enough to cram them into the space available?

Could it be that the creases are rounded as if wrapping around posts that do not exist creating the "dead air" situation?

Are there noise generation issues due to turbulence or the pleats contacting one another?

Maybe the motor force is inadequate?

For any, or all, of these reasons retreating to a lower f makes sense. It also makes sense in the ES version, but not without problems. Chief among them is the way we end up with an excursion potential that we cannot use due to membrane tension. By varying the angles of zig zag pleats we can change the f number. I was about to write "average f number" in the belief that it was high in the bottom of the pleat opening and low at the top. This would be true when we consider the pleat depth as being fixed, and note the way the width varies as we move our width measuring ruler up and down the depth dimension.

Another way to look at it is as a sequence of air layers being progressively less compressed as they approach the open air, but with the same f number as boundaries. Not certain what it all means!

For anyone who has doodled with serpentine shapes, as in AMT's, it becomes obvious that at each end we have a "half strength" pleat. The final opening is bounded by a pleat that moves and one that does not move, so reducing the spacing seems in order. The Heil woofer, called Transar, was an example of a half strength AMT, and an example of zig zag construction; actually sawtooth. Endlessly fascinating!

Keith

[DavidJanszen]

I forgot to mention that the effective area of the planar ESL can be varied by simply taping over areas of its surface. Between changing the area of the ESL and the area of the solid surface, the effects of the f number and pleat depth can be investigated quite easily. Driving a planar ESL in the first place will also validate the signals and bias.

[Keith Taylor]

When you think about it the standard EM AMT designer is changing his f number by "concertinaring" the pleats. It is just that he has to operate within the constraints of a frame; so to reduce f he has to remove the membrane and snip off a few pleats and reinstal it. Thus any distiction between zig zag and parallel pleats becomes rather blurred. It gets down to a question of how sharp the folds are. Of course the ES case calls for parallel pleats or an appropriately shaped stator or other mechanism to achieve an even force over the pleat area when they depart from parallel.

Keith

[DavidJanszen]

Originally Posted by Keith Taylor

When you think about it the standard EM AMT designer is changing his f number by "concertinaring" the pleats. It is just that he has to operate within the constraints of a frame; so to reduce f he has to remove the membrane and snip off a few pleats and reinstal it. Thus any distiction between zig zag and parallel pleats becomes rather blurred. It gets down to a question of how sharp the folds are. Of course the ES case calls for parallel pleats or an appropriately shaped stator or other mechanism to achieve an even force over the pleat area when they depart from parallel.

Keith

You might say that triangular gaps reduce the acoustical impedance while also reducing the vibrating area relative to parallel pleated gaps. For ED AMT's, there is probably a well characterized relationship published somewhere between the pleat geometry and the spectrum and SPL per unit area. Also, in designs where the magnetic field runs fore and aft while the current runs side-to-side, the force will be reduced by a factor of the cosine of the angle of each pleat away from parallel, e.g., a 90 degree included angle will produce .707 times the force, equivalent to halving the area.

[Borat]

AMT is a clever name used to obscure the real motivation behind the design.

motion transformer !
accelerating air !

WOW !

not really ...

pleating is BAD. they HAD to do it in order to organize the field around the ribbon element in an efficient way.

you see for good field u want it not to have to jump across a wide gap. now ribbon is large in two dimensions but zero in third. they needed a way to orient the field along the third dimension in which ribbon is flat - and they did just that with AMT.

then they had to think of a way to sell it - so they came up with the clever marketing name of "air motion transformer"

in ESL you're already having the field in the right place. it doesn't need any fixing.

i analyzed cons and pros of various geometries such as ESL and AMT before writing this:

Plur

the design challenge is to match the shape of the field to the shape of the diaphragm. ribbon speakers don't do this which is why AMT is a solution to that problem. my PLUR technology is another ( more elite ) solution to that very same problem.

thing is as i said ESL doesn't have this problem to begin with so you're solving a problem that doesn't exist with an electrostatic AMT ...

[kenpeter]

OK then, heres a weird variant....

Supposin you let the each pleat touch both neighbors? Not short, just touch...
They would all need one dielectric surface oriented to insulate those meetings.

So anyways, you now got two caps fighting for posession of a surface area.
The pleat peels away from one cap and onto the other, until charge per area
are equalized. But then there is this peel effectively sliding back and forth...

Does this direct contact method avoid the non-linearity of field inbetween???
Could it be done with less Voltage since the distances involved are less???
Would Stiction or Van der Waals mess this up, or are those in balance???

[kenpeter]

Does the reduced distance now make the motor strong enough for the purpose???
Does this solve the problem of high frequency rolloff with increasing pleat depth???
I mean, even if the pleat is some absurd 30 inches deep, you got only a small part
of the pleat peeling from one surface to the other at any time. Or have I somehow
misunderstood the implications???