Dipole Rear Wave, Round 2

[matevana]

Thanks for all the comments on padding the rear wave of a dipole. I wanted to start a new thread because I went off in a different direction.

Using my Diminui Dipole Project test baffles, I added a threaded rod to the back of the baffle, centered just above the full range driver (Peerless 830987). I then machined a hardwood block about 3/4" in thickness, roughly the same surface area as the back of the driver. The block can be moved along the length of the rod, and is ultimately held in place by a nut and lock washer on either end. Essentially this creates a deflector behind the full range driver, at variable tuneable distances.

My thoughts were twofold. (1) It is easier to work with the rear radiation up close, before the wave disperses fully. (2) The typical figure-of-eight pattern is altered so that the wall receives less of a direct reflection, thus reducing the need for an optimal minimum distance (1m) from the rear wall.

Initial Impression: I have only had an opportunity to play with the test baffle for about 24 hrs. It does show some promise. Sitting in the listening position and having a helper tune the deflector by varying the distance has an effect on both imaging and reverberation. It seems to have a positive effect on any siblance that may have been produced as a result of improper placement. The sound is somewhat less diffuse yet still retains the positive dipole qualities. As with everything else, there seems to be a sweet spot, in my case when the deflector is placed about 4 inches behind the full range driver. Anything closer seems to approach monopole in nature. Anything further allows more of the back wave to reach the rear wall directly (too close in my setup) and reduces the benefit. The baffle seems less sensative to relative distance from the back wall, allowing closer than optimum placement.

I appologize for having no test equipment, other than my ears. I am a musician and believe I have a good feel for accurately reproducing the soundstage, as unscientific as that may be.

Please comment. You have all been extremely helpful over the past few weeks!

Ed

[LineArray]

Hi Matevana,

i guess there are some problems which go together:

The front and rear radiation of your (any) widerange driver is
not symmetric, due to magnet and spider in the (rear) way and
the cone shaped asymmetric.

BTW:
How is the cutout shaped ? A very open conical cutout to
the rear or a thin baffle near the wideranger is desirable
to avoid cavity effects. The full cone area should be open
even to the rear. (I know some baskets open only 1/3 or 1/2
of the area ...)

Even when placed in a simple open baffle there will be a drop
in the highs radiated to the rear. That drop may be associated
with some bumps and notches: It is usually not very smooth.

That rear drop (with irregularities) may appear at above 1-2
Khz with a small to midsized fullranger.

Since your driver is very small that effects may shift up to
some Khz, which is a critical range of hearing determining
presence and thereby the estimation of distance of the
stereophonic phantom sources.

Angular dispersion (to the rear) is varying too with frequency
in that region.

Since the driver and the surrounding structures are very small,
those irregularities mentioned might as well affect the frequency
regions where fricatives ("sh", "s", "f" to mention only the voiceless ones)
of the human voice and cymbals have high energy.

As a further consequence you have ratio of directed vs. reflected sound
energy in the presence to brillance region fluctuating with frequency
and additionally the reflections are rather early because of your small
distance to the rear wall.

I fear if you want to tune your deflector setting auditively,
you will end up with different results depending on the recording
used for tuning. In the worst case your speaker will only sound
acceptable with your favourite recordings that you used for tuning.


There are some unsorted suggestions from my side, you may see
whether some of them meet your taste:

- Make the rear cutout as open as you can

- Shape the deflector more as a diffuser, e.g. attach a cone to
it with the tip pointing to the driver

- The cone need not be massive it may be made like a nautilus
from open celled rubber foam e.g.

- You can enlarge the rear block to a (asymmetric?) disc and
stuff the cylinder shaped slit with damping material, soft and nonreflective
material near the driver, more rigid but also open celled material outside

This is going towards a "resistance box" with cardioid dispersion pattern.

The rearside disc could also be perforated - keep the structures
itself well damped and rigid - so the resistance box could be built like
a round or egg shaped "cheese box", preferably with the center offset
from the driver cutout.

A radical solution would be a stuffed tube used as an enclosure turning
the wideranger into a monopole, that tube could be open if long enough
and damped sufficiently, it can also be bended to save space.

But i feel you want to maintain as much openness as possible, so a
cardioid might be better suited for your room setup as a dipole is.

After applying those or similar changes you will most probably
have to correct the tonal balance of you system.

Do not let yourself distract from that: A better solution concerning
damping and more stable dispersion may sound worse at first glance
concerning tonal balance. To be "fair" whith each solution found
you have to compare them using a tonal balance as similar as possible.

Since you have two speakers you can modify them alternating and
listen in mono, giving both a similar and nearby position.
That keeps you from beeing fooled from your memory, since all
settings need some time to be mounted.

Fine tuning using stereophonic material can be made afterwards.

Use of - even simple - measurement equipment can help you
seriously while moving forward by correlating results with
your auditive impression. For watching the tendencies even
an uncalibrated cheap microphone is helpful as long as you
do not even try to get ruler flat plots.

Try to make the system behave consistently under different
angles and listening distances, smooth changes with angle
are allowed, abrupt ones are undesirable.

In my experience pure "sweet spot optimization" is not the
way leading to an enjoyable speaker.

Kind Regards

[LineArray]

This is a simple "rear muffler" made from felt (oustide) and
soft polyester wool sewed into a coat from a very open
fabric, to prevent material touching the cone and
giving some nicer optics.

I used it to tame the irregular transition region of the
rear radiation mentioned above. This way the rear
radiation looks more like low pass filtered in a controlled
fashion.

This could be a cheap start for your setting too.

Of course in the "Dipol 08" design i use filler tweeters,
to maintain dipole operation up to the top end.

[matevana]

Oliver, that is GREAT information. Before I read your reply, I experimented with a small preformed styrofoam cone, with the tip placed behind the driver. Probably not the best diffuser material, but it also showed promise.

Also a great point about tuning by ear. I have a Behringer DEQ2496 with measurement mic and I will be using the RTA function to perhaps get a better picture of what's going on, along with the SPL function to ck levels at varying frequencies.

Your rear muffler seems like a great idea, and seeing the picture really helps me visualize how that might work. I realize that you are using additional filler drivers in that design, but did you find the muffler to be effective?

[LineArray]

They are effective in making a more smooth rolloff to the
highs on the rear side and tame the "dirt" coming from the
rear at higher frequencies, yes.

Frequency response plots taken from the front looked a bit
nicer, spectral decay looked nicer as far as i remember, but
in my impression there are two sides of the medal:

With my setup the mufflers led to a subjective degradation
in fine dynamics but as that speaker shown is a fullrange
with subwoofer also the mid to upper bass region was
affected slightly.

Maybe using more sophistcated measurement equipment,
some non linear behaviour could be found. Auditively
i believe, there is some kind of compression.

You might get away without that with your speaker,
since you have a 2-way.

Since my distance to the rear wall is 1.2 to 1.6 meters
depending on the room, the amount of "dirt" whithout the
mufflers is close to be inaudible and i mostly prefer to run
them open, which is easier when there is a tweeter delivering
a proper rear output in the presence to brillance region.

With closer distance to the rear wall a cardioid dispersion
should be preferable.

I guess in your setup the art will be not to overdo it,
i would go for "more volume and less resistance" of the
semi open "muffler chamber" - however you will design it -
in case the sound gets "compressed".

You will be the only one who is able to find a suitable
compromise. Maybe an "improved diffuser" sounds better
to your ears. Report if you like ...

[LineArray]

I wonder how a cone out of pumice stone would do
as a diffuser with absorbent component.

[matevana]

Originally Posted by LineArray

I wonder how a cone out of pumice stone would do
as a diffuser with absorbent component.

LOL, and why not pumice??? It's hard, dense and can be honed to various shapes. It also has those nice holes which may add surface area as a diffuser. To date I have tried a stryofoam cone, a baseball hack sawed in half, a tennis ball half, and my personal favorite, a ceramic serving spoon. The spoon (convex side toward the driver back) seems to have the perfect dimensions and is being held in place by a gooseneck. The gooseneck microphone boom has the added advantage of being able to angle the curve and experiment with reflections. The baffle is now infinately more usable at close distance to the rear wall.

[LineArray]

Hi,

yes my suggestion of pumice was meant seriously ...
though i never tried it for those applications.

Have you got a picture from your serving spoon
arrangement ?

[matevana]

Originally Posted by LineArray

Hi,

yes my suggestion of pumice was meant seriously ...
though i never tried it for those applications.

Have you got a picture from your serving spoon
arrangement ?

Oliver,

Sorry about the picture quality but you can get an idea how the convex shape of the spoon diffuses the frequencies from the full range driver.

The tripod and alligator clamp was the rig i used to test various materials placed behind the 3" driver, with varying outcomes.

Ed

[LineArray]

Hi Ed,

interesting arrangement. I guess the serving spoon
will not have expected a carrier like this.