PIEZO NXT type panel

[jzagaja]

ARTA can do one shot sine sweep with distortion plot so you don't have to play a Russian roulette with Aurora plug-ins

[Ziggy]

jzagaja..........exactly what have you measured in your frequency response?.......the STARBASS driver?

If so, we have something similar in Australia............sounds like crap, no matter where you put it.

[jzagaja]

That is a question to Pol I even don't know what the STARBASS is

Appended picture contains FEMLAB simulation and actual response with and without added mass. BMR is much "nicer".

[jzagaja]

Sorry wrong picture.

[LineArray]

Hi, these are some current results from my bending
transducer

x-axis from 0 to 10 ms

appprox 1m distance,
horizontally on axis,
vertically approx 10 degrees above axis.

[LineArray]

same setting

x-axis 100Hz to 20Khz , y-axis in 10db scaling
(sorry for bad picture quality)

[LineArray]

same setting

Decay spectrum

frequency from 100Hz to 20Khz

time from 0 - 4,31 ms

amplitude from -26db to 5db

[LineArray]

Since this driver operates modal above 800 Hz,
it would be useful to average over several
microphone positions.

As for all DML speakers a single on axis microphone
position does not tell enough to estimate the behaviour
of the driver. Frequency response varies even with small
changes in angle, which is typical for DML.

When tuning the driver to be nearly flat above 5Khz,
it sounds "overbrilliant" and "nasty".

Auditively the driver sounds rather good by now,
the presence region is not as natural as it should be.

There is still a peak around 5Khz, which can be smoothed
easily using a notch, which i omitted for the measurement.

But i heard a lot of conventional fullrangers which sound worse,
and need more amounts of compensation ...

[jzagaja]

As shown DML with added mass merely alters the distribution of modes because modal density in DML is high and in BMR is low.

[LineArray]

Hi jzagaya,

you have shown interesting investigations.

As far as i understand, there is a main problem which is
common to all transducers which rely on propagation of
bending waves:

Low modal density at the bottom end of operation. Low modal
density of a DML is nearly a synonym for "low quality" in the
frequency range under question.

I can identify the following strategies to circumvent that
Problem in current (commercial, established) systems:

1: Balance the modes by adding mass (and optional damping),
and align the diaphragm to have usable modes for balancing
-> "BMR"

2 a: Avoiding the range of low modal density

2 b: Making the diaphragm (excited panel) large enough
to have high modal density in the low frequency range.

---

2a Leads to multiway systems, incorporating conventional
dynamic bass drivers. A very pragmatical way to go IMO,
which may lead to good results, when implemented properly.

2b Leads to very large sized panels, causing the well known
acceptance problems (at least in the "non audiophile"
rest of the world ).


IMO there would be a third strategy:

Think about increasing the modal density at low frequencies,
without increasing the diaphragm ...

When going this way solely by searching for appropriate
materials, we end up with materials that have low speed
of wave propagation leading to low efficiency.

Kind regards