Electrostatic AMT?

[mige0]

Originally Posted by Keith Taylor

Michael, with my construction, where membrane contact is via the posts adjustment of tension is not something I have much control over. If tension is too small we loose contact and it stops working. I would not disagree that your spreadsheet is in need of some debugging. I could be missing something, but when I reduce the audio voltage to zero we are still getting a huge SPL figure!
Keith

Though I have not done any debugging at this point . there is a new - and now almost completed - version available I have uploaded today (091001)

Maybe you still are a little bit confused that I claculate *two* SPL figures – let me explain:


There are several brick wall SPL limits with ESL's

1.) SPL limitation due to limited Signal and Polarisation Voltages ("Perc of breakdown Voltage")
these limitations are addressed so that *if* you enter too high voltages for a given gap the corresponding fields turn RED. The maximum one can safely (and reliably) apply is determined by the break down voltage of air (arcing).

2.) SPL limitation due to limited gap width ("Full Gap excursion SPL")
this is a pretty easy to understand constant as it is the same as the mechanic x-max of dynamic speakers. The *SPL* you can reach depends on frequency (for the same max excursion)
You have to have the power to reach this limit though!!

3.) SPL limited due to current drawn ("Pow demand (100% Signal Voltage)")
Usually you have a certain amplifier or step up transformer wattage limit. If you push voltages too high or make areas too large – then you unavoidable will run into this limits.
These limits are addressed in that the fields "Pow demand (100% Signal Voltage)" turn into RED

4.) SPL Limit from too high Audio or Polarisation Voltage ("100% Signal Voltage SPL") with respect to the lowest frequency ("Low Frequency")
It is possible to stay within all former limits and nevertheless get too much excursion with respect to what your gap allows. This may occur in the lower frequency department where radiation impedance becomes very low.
To address this limitation the fields " Perc of Full Gap excursion" turn into RED


To sum up
The two numbers of SPL in my spreadsheet are for
1.) the purely mechanic determined SPL limits of your design and
2.) for the operational determined SPL limits

Originally Posted by Keith Taylor

I could be missing something, but when I reduce the audio voltage to zero we are still getting a huge SPL figure!
Keith

As outlined – the pure mechanic X-max SPL limit is there even when you turn your voltages down

Please report back if it doesn't turn out right for you

Originally Posted by Keith Taylor

An AMT tutorial
In the electro magnetic version the pleats are arrayed parallel to one another in space. That space has a magnetic field, where the flux is running parallel to the pleats. Any forces, and subsequent motion due to current being passed through conductors on the pleats will be at 90 degrees to the field. The thing to realise is that each pleat is subject to forces quite independantly of conditions in neighboring pleats. To do any useful work (move air) three sides of the pleats have to be sealed to the neigthboring pleats such that the opening alternates between front and rear. The seal can be as floppy as you like because it does not have to perform any "station Keeping" function for the pleats.

In an ES AMT things are quite different in that the forces are between adjacent pleats, and, in the case of bias; exist continuously whether audio is being applied or not. OK, we know that in the perfectly centred case bias induced forces will cancel. It is just that we cannot rely on it and have to resort to tension. Apart from the necessity of tension the inter-pleat spacing method prevents us from achieving the the air displacement that a floppy seal allows. There are pockets of "dead air" abutting the posts.

Keith

I'm well aware of the principle and of the mechanics involved – I currently use my second generation of AMT's – knowing them inside and outside and even a little bit more

I'm fascinated by your design because it seems o be possible to overcome the lowish SPL limits of general ESL's.

Key here is pleat "depth to width factor" nothing else!
We will see if this is eaten up by the additional radiation impedance load once your prototype is up and working – really hope not...

Originally Posted by Keith Taylor

A new idea
In the drawing is depicted a stator type ES AMT where the stator has a special shape, and is acousticaly transparent. It could be perforated metal, wire or wire mesh. The membrane pattern is zig zag but the forces over the pleats are more even than if the stators were flat. There should be a lower percentage of air that cannot be moved compared to parallel pleats.

We come to the question of how to view it acousticaly. One way is to note that our "box of air" has become a "wedge of air", where for a given ammount of motion we have a pressure gradient in the pleat opening. For the situation where motor force is marginal perhaps we have a "half way house" between a flat radiator and a pleated one. It could also be viewed as a micro horn. The shape of the stator could mimic the catenary? curve that the pleats may assume to supply higher forces at the edges.

I figure that if you were to put some water in a bath and provide a flat surface below the water you could do the following--. Tape a pair of flat boards (wooden rulers) together at one end and lay them on the underwater surface. Add a polystyrene ball as an indication of fluid motion and move the boards. I think we should see the transformation from high velocity at the apex to a reduced velocity at the exit. Likewise the pressure should be transformed if we had seals top and bottom.

It won't have the acoustic energy per area of a parallel pleated design but breaks the nexus between pleat spacing and motor force? If we look at what is happening at the narrow end of the pleat opening during a compression the motor is not burdened with being a vacuum pump as well as a pressure pump to the degree it would be in a normal AMT. Construction should be simpler than the all moving design with some interesting scope for varying the spacing by bodily moving the stator triangles and altering tension without loosing membrane contact. It seems to lend itself to a more midrange design.

Keith

Nice idea !
But allow me to repeat that I think you are too much worried with the "aerodynamics" – this is quite different with waves...
Worst scenario you get is what could be subsumed as "added air mass" – and as outlined – a ten times the membrane mass let your SPL drop by 20dB – very easy to predict...

Michael

[DavidJanszen]

I think your new idea is very clever. Although it presents some daunting fabrication challenges, it would seem more apt to produce waves that can travel out.

[Calvin]

Hi,

I'm fascinated by your design because it seems o be possible to overcome the lowish SPL limits of general ESL's.

Everything has imits.....and sufficiently high SPLs can be reached with general ESL design. AMT does nothing else but reducing size. That this won´t come without compromising on other parameters should be obvious.

jauu
Calvin

[DavidJanszen]

Originally Posted by Calvin

Everything has imits.....and sufficiently high SPLs can be reached with general ESL design. AMT does nothing else but reducing size.

There is what would be an advantage under some conditions to reducing the radiating area for a given SPL -- wider dispersion. Also, simply making the baffle more compact might be a good thing. Alternatively, getting more SPL from a given area may also be appealing, although it would come with the same high capacitance that the folded up area would have if unfolded.

Originally Posted by Calvin

That this won´t come without compromising on other parameters should be obvious.

I think most people grasp the fact that trade-offs exist, but not all are aware of what they are at the onset of a brainstorm. Even experienced engineers should regularly allow themselves to postpone considerations of practicality, at least to some extent. Amateurs and experts operating in this mode can and do get intriguing and potentially useful basic ideas. Then during the process of evaluating those ideas, they apply or gain expertise that may or may not lead to alteration or rejection. Any expertise gained will benefit future thinking, and may put a person partway down a road that no one has traveled before. This is the basis for much original work.

[mige0]

Originally Posted by Calvin

Hi,


Everything has imits.....and sufficiently high SPLs can be reached with general ESL design. AMT does nothing else but reducing size. That this won´t come without compromising on other parameters should be obvious.

jauu
Calvin

jauu Calvin, do you find at least my spread sheet useful to some degree ?

Michael

[mige0]

Thinking about the impacts of the ESAMT design and reading a lot valuable sharing’s in ESL threads, and also extensively consulting my own spreadsheet – I found some good reasons not so obvious for me at first to continue with your invention Keith.

1.) I have come to the conclusion that stability in the sense of collapsing pleats is a non-issue. Compared to a standard ESL there is basically no difference in the condition of stability versus film tensioning. Worst case scenario IMO is that you end up with doubling the tension to keep stability - thus most ill result would be doubling the lowest possible resonance frequency. No big deal IMO.
2.) Thinking about additional air mass that loads the ESL when folded to an ESAMT I think this can be seen as a net positive effect. For the upper frequencies it does no harm as is obvious when we look at large single panel normal ESL’s that don’t suffer from large areas in the upper range *from principle*. The beauty of ESL’s actually is that the additional load due to greater radiating area is balanced by the stronger motor (as long as we can afford the higher power demand of course, and also keeping in mind the pleat depth lamda /4 limit). Practical limits like directivity (beaming) are even in advantage for folded ESL. As for the lower frequencies the higher load – per front area – should have the effect to more efficiently dampen the F-res peak – again, equal front areas compared to each other.
3.) Once folded ESL gets in size towards full range speaker there is a benefit in that no additional spacers are required to insure film stability over large radiating areas.
4.) Last but not least - to put out more SPL per front area is a benefit by its own of course

To sum up I see no big drawbacks in the upper frequency range (besides the lamda /4 limit that we do not have with standard ESL’s) but a serious advantage in the lower frequency range.

To get way better loading at the low end is an huge advantage IMO, as ESL’s have their weakest spot there.

As for possible ill effects of flow resistance when the pleat depth to width ration increases, my guts feeling form what I did in sectoring horns and what I see from phase plugs for compression drivers I would really be surprised if these effects seriously would enter the picture below - say - a factor of 10.



Michael

[DavidJanszen]

Originally Posted by mige0

Not exactly the case, as you can increase signal voltage *and* polarisation voltage – which balances with the breakdown voltage.

I'm not sure about your wording about balancing the breakdown voltage, but you are right about Coulomb's Law not being the problem that I wrote it was. I even know better from both experience and arithmetic, but had a brain fart. Sorry everyone. Please do not pose the obvious, embarrassing questions.

To review correctly, force is proportional to the product of the bias and drive voltages, and inversely proportional to the square of the gap width. Along the lines of what Michael pointed out, it can be arranged that the bias and drive voltages are both doubled for a doubling in gap width, whereby any increase in gap is entirely made up for. There are other factors, but this holds true at least when considered in isolation.

[DavidJanszen]

Originally Posted by mige0

to make things simple for you guys, I started to put together a spreadsheet

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.

[Keith Taylor]

Michael, I agree with much of what you are saying. We really need to explore f factors of 10 or more otherwise we might as well give up because we are stuck with a situation where we have a spacing and excursion capability that we cannot use because the resonance is too high. I have a feeling air mass loading is not going to be the problem that spring loading obviously is. I also believe that closing the pleat ends is going to lower the resonance.

When we consider guideline suggestions for normal ESL span to spacing ratios of around 100 (Sanders); 10 is a modest figure. My next move will be to build a deliberately oversize pleat depth to try and answer two questions.
Is air mass loading a problem?
Where is the upper cut off frequency?

The "AMT tutorial" was not aimed at anyone in particular; just thinking aloud!

The mention in an earlier post of zig zag pleats allowing a continuum from flat to pleated should have had the qualification Provided the pleats are acoustically small In the realms of impracticality, if we take a flat radiator powered by means that are irrelevant, and segment it into small pleats with hinges, we could concertina it between flat and fully closed. This would change the f factor from 0?, 1? to infinity. Any dissagreement?

Keith

[DavidJanszen]

Originally Posted by Keith Taylor

My next move will be to build a deliberately oversize pleat depth . . .

I think it will be instructive to try approaching a solid surface with a planar ESL panel while it produces pink noise. Maintaining the panel parallel to the surface will illuminate one set of things, and trying it at different angles will illuminate another, and varying the area of the solid surface will produce additional results.

If you don't have one, and can not get one easily or cheaply enough, one can be made that is good enough for this test using two small panels of perforated sheet metal as stators, some strips of plastic, and some contact adhesive. Cut one corner from each stator to allow a clear connection to the opposite panel. Insulate and reinforce the membrane with PET tape along the edge where you bring it out for the bias connection. Tape the outsides of all edges with at least two layers of PET tape to protect yourself from the signal while handling the panel. (If you have not yet felt the signal during your AMT trials, trust that you have enjoyed the preferable alternative.) Use a pair of washers in contact with the protruding bit of membrane to make the bias connection and clamp them together with a clip lead or some better method of your own devising.

If your film is biaxially oriented, you can probably tension it adequately using hot air after gluing it to the spacers. If not, I read of an interesting method some time back where the film is taped down over a small table with a big hole in it that has a bicycle inner tube laid upon it, and the inner tube then inflated. The stator with its spacers already coated with adhesive is brought down onto the now tight membrane, and voilà.