PIEZO NXT type panel

[mikesbmw]

Hi Oliver,

There is truth in what you are saying about trying to compensate for more then one listening position (if they are not close to one another). This cannot be done. It is also true that compensation will, most likely, be mainly beneficial for low frequencies.

The corrections I do, are for one listening position. All the other positions are for easy listening only and for those it doesn't really matter if there are deviations.

Do you have any measurements for the different materials used? Most of the materials tested (in this thread) are done by ear (which together with the brain is the weakest link).

I'm not saying that one can simply use DSP to correct for everything, but in these simple and crude DML setups I very much doubt that the frequency plot would be within +/- 5dB between 200-15Khz.

[LineArray]

I do not have systematic measurements on different materials
under same conditions.

But the following called "A1" is from stamped aluminium sheet-
quite similar pattern like your prototype ...

And the one called "B1" is balsa wood.


As i said mounting conditions and measurement conditions are
not comparable, but may give a tendency.

[LineArray]

In my experience there are two possible reasons for
early (and steep) rolloff in highs:

- moved components of the exciter have too much mass

- membrane (core ? ) or glue is soft and thick, so there is
a compliance between the "listening side" surface of the
membrane and the voice coil, which is too high

Of course both factors can go together ...

[mikesbmw]

Interesting graphs!

I would not have guessed such a nice frequency response from both materials. Especially the aluminium looks nicer then expected.

After doing some further reading on BMR speakers I see I'm thinking of them acting like a piston too much eg. standard loudspeaker. I totally "forgot" about the propagation of sound through the panel material and then leaving the material at different intervals. Introducing lobbing, distortion and unfocussed sound. Excuse the layman's terms, I am no scientist

I'm currently at page 42 of this thread and I see that many of my questions and wanderings are the same as many of the contributors of the thread... I'll shut up a bit for now and read the rest of it first

[LineArray]

Originally Posted by mikesbmw

Interesting graphs!

I would not have guessed such a nice frequency response from both materials. Especially the aluminium looks nicer then expected.
...

Hi mikesbmw,

that ALU prototype must also be burried somewhere in here ...

The AL sheet had little patterns like pyramids stamped on it, so it
does not sound like a flat sheet. Sonically it was OK up to the midrange
region. But it had rather sharp lobing, which caused severe changes in
presence and brillance region when you moved your head a little.

I tried to heal that with some kind of "phase plug" or diffuser. That
helped but the underlying problem remained.

The problem might have been, that the wave propagation on the
material is very fast at high freqiencies (large bending wavelength) and
the sheet was rather small (approx DIN A4). With smaller bending
wavelengths on slower materials the lobes get finer distributed over
listening angle, so they are in best case not detectable - auditively -
anymore as you move around.

As you can see, there is ringing even at the high frequencies, although
the sheet was clamped at its edges using highly damping materials.

So: The graph looks better than it sounded actually !

(And i bet, in that case the problem was even unhealable with DSP ...)

Kind Regards

[el`Ol]

Oliver, don't you think your CSDs on top are a bit too fast to qualify as DML? In experiment with balsa (small sheet, no end grain) I got ~6 ms.

[LineArray]

There is a damping suspension ...

What do you mean exactly with "not sure if qualified as DML" ?

I assure you, the sheet is not able to move pistonic.

The motion is purely modal.

It is not ringing which qualifies as DML.


Kind Regards

[LineArray]

In fact all manufacturers of good "modal" transducers
state that the right amount of damping is essential
and one of the most tricky parts.

The ideal would be surrounds (of the panels, membranes used)
which are free from reflection over the entire frequency range.

The Göbel website e.g. illustrates that in a demonstrative
manner.

[el`Ol]

I suspect the high frequent parts of your wave are dampened away long before they reach the edge of the diaphragm, so you have conventional bending wave transducers, not DMLs, at least for the treble.

[el`Ol]

Originally Posted by LineArray

The ideal would be surrounds (of the panels, membranes used)
which are free from reflection over the entire frequency range.

Wouldn't that prevent a modal operation? This is more what the Manger is intended to do.