"Dipol 08" Baffle Dimension and List of Crossover Parts

[LineArray]

The real crossover in its "shelter", some wires
still missing ...

[CLS]

Hi LineArray,

Thanks for your insight. I'll do some experiements when there's chance.

Adding rumble signal to overcome the 'hysteresis' of the drivers is new to me. I didn't even think of that. Maybe it will work nicely. But there's a 80Hz HPF for your array, isn't it? Then the rumble would be largely attenuated. How do you plan on getting a good balance between over modulation and the effectiveness?

[LineArray]

I did not mean that "rumble" generator seriously
to 100% , but it was not just kidding too ... just
playing with options how to overcome a problem if
it cannot be solved at its roots.

In active speakers you could integrate such a
circuit even for all "highpassed" drivers, causing
a very small subsonic vibration, so the drivers have
always something to do ... even the tweeters.

Best would be to have drivers with little mechanical
loss and also amplifiers which work perfect for very
small signals.

There are some drivers on the market which have very
dominant damping by back-EMF and little mechanical loss.

On the other hand it is fairly impossible to make good
drivers without some amount of mechanical damping
in the membrane and the surround, because you need
it to control vibrational modes.

So it is kind of a typical deal with different tradeoffs.

But i feel mostly dome tweeters with heavily impregnated
surround suffer from that nonlinearities at low excursion.

Some (pure) ribbon tweeters may be good in that respect,
as they get a large amount of damping from radiation and
air friction, so usually those need no dampening coat.

In dipol 08 i only made that observations mentioned
with the tweeters, but i would say it is a very minor
issue.

Dome tweeters with heavily impregnated surround and/or
ferrofluid in the magnetic gap are candidates which
should be inspected for distorsion at low excursion
and shift of parameters depending on signal level ...

This is why i no not use ferrofluid tweeters.

Best

[LineArray]

Originally Posted by CLS

Hi LineArray,

...
How do you plan on getting a good balance between over modulation and the effectiveness?
...

If such a circuit could be shown to have a "healing"
influence you would do it like with any compensational
measure ... measure low level distorsion and increase
the "residual rumble" until there's no improvement
to be detected ?

Signal could be bandwitdh limited noise, maybe a
5-15 Hz sine would even do for prototypical
experimenting ...

[LineArray]

This datasheet refers to FR 125 S, the new version
FR 125 SR does not differ significantly, but has
a changed basket design with identical dimensions
for mounting.

This datasheet is somewhat more detailed than the
single page version mostly found in the net.

FR data is rather smoothed, like usually in manufacturers
datasheets ...

http://www.meniscusaudio.com/images/...5-data-v3s.pdf

[LineArray]

These are the tweeters used (2 tweeters per panel):

Monacor DT 25 N

DT-25N - Monacor Neodymium HiFi tweeter 80Wmax 8Ω   - Europe Audio

DT-25N Neodymium-HiFi-Hochtöner, 80Wmax, 8Ohm MONACOR / IMG Stage Line günstig kaufen schnell bestellen



A comparable one may be

Audax TM 025 F17

But the tweeter panel's crossover would have to be
changed for this one ...
(lower Re, higher voltage sensitivity)

[Elias]

Hello Oliver,

Nice project!

Let me ask some questions to understand the design more:

1) How and where the tweeter panels are placed with the main panels? Got any picture with the tweeter panels installed?

2) How high the 6 fulrangers play with the cross over? I see a treble boost. Does it mean there play the top range as well?

3) Have any measured acoustical responses of the lower and upper triples separately with the cross over?

4) And the tweeter measured responses with the cross over?


The long list of questions indicates I may have missed the essentials of your design

- Elias

[LineArray]

Hello Elias,

1) Maybe the attached picture helps ...

2) Yes ... the tweeter panel is only for correction
of the polar response to the rear.

Under free space conditions or in very absorbent
rooms the contribution of the tweeter panel is
less important. In usual living rooms it is
advantageous ...

The upper trio of fullrangers gets dominant above
4Khz, as you can see from the voltage transfer
curve of the filter i have posted.

3) I have no "isolated" responses - in fact there is
no such thing as the lower trio of fullrangers
is in connected in series with the upper trio
... maybe have a look at the wiring diagram
i posted in the beginning.

There is a simulation of upper 3 vs. all 6 drivers, at
http://www.dipol-audio.de/dipol08_simulationen.html
but with idealized drivers and without "treble boost".

4) Same as 3) but i may find a diagram of
the voltage transfer function for the tweeters ...


Kind Regards

[Elias]

Thanks it's more clear now.

How did you end up with the driver spacings, is there a formula or is it based on iteration of vector summation with simulator?

Using such a sparse arrray it is all against the common gospel of center to center spacing and all that

However I don't doubt its functionality.

Originally Posted by LineArray

.
The frequency dependent voltage shift between the
driver groups is very smooth, so you cannot really
detect which fullranger reproduces which frequency
range from usual listening positions, the array sounds
as being centered subjectively around the 2nd driver
from above as a consequence of the combined effects
of distance weighting and power tapering.

This is quite interesting. I took another look at the voltage transfer functions and it seems upper and lower triples are identical up to 3kHz. The 3dB difference point seems to be at 7kHz. At first it sounds a bit high, but then the fullrangers are starting to beam and it might lessen the comb filtering.

What happens if you move vertically from the floor level up to the top of the array, where is the perceived point of radiation? Does it follow you as you move or is it fixed spot withing the array i.e. is there a tilt in the wave front? Let's say the distance to the array is about 2m.


- Elias

[LineArray]

Originally Posted by Elias

Thanks it's more clear now.

How did you end up with the driver spacings, is there a formula or is it based on iteration of vector summation with simulator?

...

Iteration with simulator starting with an equidistant
array as a base .... golomb rulers were the
initial inspiration for modifying the spacing ...

Originally Posted by Elias

...

Using such a sparse arrray it is all against the common gospel of center to center spacing and all that

...

Yes, the speaker breaks all the rules in the book.

But it is a kind of "bending" the rules, not just
ignoring them. As the orthodox close CTC spaced line
array rules cannot be obeyed using mid sized
fullrangers anyway, no matter how close you mount
them ... a point which often overlooked i feel.

Originally Posted by Elias

...
What happens if you move vertically from the floor level up to the top of the array, where is the perceived point of radiation? Does it follow you as you move or is it fixed spot withing the array i.e. is there a tilt in the wave front? Let's say the distance to the array is about 2m.
...

It does not follow in the way a long vertical ribbon
line source follows your ears when "knee bending".

At a distance of 2m you feel some "soft centering"
around the upper drivers, resembling presence
and brillance coming from there and the sound
sources being more dense - it is always the combined
filter + spacing effect which makes up the subjective
centering.

Many people have been "knee bending" in front of this
speakers - you do not experience "gaps" or discontinuities
although those are there, but the spectrum wiggles smoothly
dependent on vertical angle - balanced enough
to statisfy even experienced listeners.

Compared to other speakers they are "statistically
balanced", no extremes occur at certain angles.

If you listen too close and sit on a low sofa there is
a tendency of slightly missing sparkle - but at listening
distances from 1,5m on sitting and standing gives the
same subjective balance.

The comb filtering is much influenced and mitigated
by the non equidistant spacing ...


Kind Regards