Widerange dipole with Karlson Couplers ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Some pics and quick measurements ...

From left to right:

A) Karlson Coupler from 60mm PP tube, length about 37cm
B) Burst decay, rear of driver covered with several layers of cloth, coupler slightly stuffed
C) FR and THD (Distance about 1m, 15 degrees vertical above axis
D) Group delay with device as shown
E) Group delay using a closed rear chamber

Within 5 degrees the highs are quite flat, more than 15 degrees off axis
causes noteable falling response >2Khz, radiation in highs is pretty narrow.

Lower cutoff frequency about 300Hz.

What about using 2 of such devices mounted in opposite direction and
adjusted (dipole path length) to have efficient dipole radiation down to
<=300Hz ?

Could be a pretty compact and efficient widerange device ...

Driver used experimentally is

http://www.tb-speaker.com/detail/1208_03/w2-748sg.htm

Of course some tweaking will be needed to make it acceptable quality ...
 

Attachments

  • K-Coupler_Length37cm.jpg
    K-Coupler_Length37cm.jpg
    359.8 KB · Views: 640
  • K-Coupler_01_BurstDecay.JPG
    K-Coupler_01_BurstDecay.JPG
    318 KB · Views: 607
  • K-Coupler_01_FR.JPG
    K-Coupler_01_FR.JPG
    233.8 KB · Views: 576
  • K-Coupler_01_GroupDelay.JPG
    K-Coupler_01_GroupDelay.JPG
    104 KB · Views: 555
  • K-Coupler_00_GroupDelay_RearChamberClosed_15DegAboveAxis.JPG
    K-Coupler_00_GroupDelay_RearChamberClosed_15DegAboveAxis.JPG
    100.9 KB · Views: 541
Last edited:
Hi Radugazon,

cannot answer right now,

comparison between

"closed rear chamber only"

and

"closed rear chamber + coupler" should be possible

Output of naked driver at 300Hz is close to nothing
of course ...

Cutouts have been done "by estimaton" using metal
shears ("L" and "R" type) and sanding the slot a bit
afterwards ... no rocket science necessary.
 
Rear of driver wrapped in cloth roll

Rear wrapped driver vs. rear wrapped driver + front coupler

No stuffing of coupler like in first post, now shows more resonant bahaviour

Pics:

A) On Axis at 1m distance

B) Same but, 15 degrees off axis (above and beside the slot)

Blue : coupler
Red : No coupler
Black: Gain

The more balanced output above lower cutoff in the first post,
is mainly achieved by stuffing.

Refinement of the slot contour may also be possible, but i guess
there will be no balanced midrange without a certain amount
of damping.

Maybe a resistive pressure chamber can give smilar effects using
a smaller diameter tube.

In current arrangement - cross section of tube equal to
membrane area - there is no gain at HF and the top end
is significanlly reduced off axis.

Effects at LF do not look too different from baffling the driver,
but the lower cutoff is steeper.
 

Attachments

  • CouplerVsBare_RearCloth_1m_00Deg.JPG
    CouplerVsBare_RearCloth_1m_00Deg.JPG
    183.8 KB · Views: 136
  • CouplerVsBare_RearCloth_1m_15Deg.JPG
    CouplerVsBare_RearCloth_1m_15Deg.JPG
    186.8 KB · Views: 108
Last edited:
The first say 10cm are "unslotted" maybe top end
would improve, using a slot starting at the throat
of the device like in known HF designs.

What to make of it (instead dropping into the waste bin) ?

When used in dipole configuration maybe for annoucement at
railway platforms ...


Could be used in situations where a 25x25cm baffle for
dipole mids has disturbing looks.

Maybe optimize for smooth midrange >250Hz and use a
dedicated tweeter/supertweeter design (e.g. back to
back domes or dipole planars).

At least it has low profile and only little effect
on radiaton of any HF transducer mounted close to it.

3-way using a cardioid bottom woofer, this device+tweeter
elevated using a low profile "transparent" stand or "flying" ?
 
Last edited:
Well...some basic thoughts :

On the non stuffed plots, there's a resonance at 450 Hz, is your tube 19 or 38 cm long ? This would be too beautiful to be true. You have to try with a longer one to see what happens.

Look at this 180 kg wonder made in Germany :

An externally hosted image should be here but it was not working when we last tested it.


And as you are here enjoy the interactive view.. We can suppose
that there is a 1" inside, so high passed around 1500 Hz, but the tube looks very long, probably 1 meter or more, and the slot is relatively shorter than yours. It's also inclined at 12 degs. If there is some logic in the high end, maybe this design is to consider.

I have seen also in a 1960 optimist paper that the opening "approximates an exponential variation", maybe you have to make the things seriously with a template.

About the directivity, if you have such an increase off axis (on the side ?) it's then a polar in V, or a ring if we think in 3D. Not good...

Last thought, you have to try with an other pipe on the back of the driver as in your original idea. With different lengths, this could be a way of modulating the resonance, and the compound horns have some interesting properties.
 
FWIW, some measurements I took of the Transylvania Power Company "The Tube", 1" diameter, 5.3" long. Drivers are Selenium D220Ti, P.Audio BM-D440 and BMS 4550.

An externally hosted image should be here but it was not working when we last tested it.


Such device don't really have much gain like horns do, but they are very uncolored and natural sounding IMO.

Getting good HF to be reproduced and dispersed might be hard in tubes of larger diameter.

IG
 
Last thought, you have to try with an other pipe on the back of the driver as in your original idea. With different lengths, this could be a way of modulating the resonance, and the compound horns have some interesting properties.

I would have thought of 2 devices with own drivers first.

Although that "impedance compensation" idea using different
tubes is nice, the dipole path length might be too long, possibly
causing side lobes.

I thought of "opposite" but shifted mounting like this:

=============Driver
...........Driver==============



Concerning the "Färber" device

i cannot see any different behaviour here in principle:

http://www.faerber-acoustics.com/assets/Uploads/widgets/equilibrium-acoustic-measurement-big.jpg

Yes where is "on axis" (in case of my test device) ?

There seems to be a deviation of the main lobe from the
tube orientation around 15 degrees for mid frequencies
(to the slotted side), surely depending on the shape of
the slot.

The main lobe seems not to occur under same angle for
all frequencies. Main lobe for the top end seems more in
line with the tube.

My test tube is 37cm long, lambda/4 resonance seems to show
"effective acousic length" around 19cm, of course the slot
makes the pipe shorter effectively due to lowest resonance.

For better LF performance it has to grow in length, sure.
 
Last edited:
FWIW, some measurements I took of the Transylvania Power Company "The Tube", 1" diameter, 5.3" long. Drivers are Selenium D220Ti, P.Audio BM-D440 and BMS 4550.

...

Such device don't really have much gain like horns do, but they are very uncolored and natural sounding IMO.

Getting good HF to be reproduced and dispersed might be hard in tubes of larger diameter.

IG


Thank you for dropping the measurement, i tend to concur with your
estimation. In which frequency range do you use your devices ?

Does "uncolored" also refer to the lower end ?
 
Last edited:
Thank you for dropping the measurement, i tend to concur with your
estimation. In which frequency range do you use your devices ?

Does "uncolored" also refer to the lower end ?

I used it crossed over at both 3kHz witht the Selenium D220Ti and then 1.5kHz with the BMS 4550. Both configurations worked well, especially the 1.5kHz, but that's also because the BMS driver is really good. It was also about matching the K12 underneath.

You are correct that the highest frequencies have their main lobe on axis with the tube. However, listening with a bit of tilt towards the slotted side seems to subjectively improve dispersion. I don't know if I have data on this anywhere. I suspect a larger tube might have similar propreties, shifted down proportionally.

IG
 
If there is some logic in the high end, maybe this design is to consider.


http://www.faerber-acoustics.com/assets/Uploads/widgets/equilibrium-acoustic-measurement-big.jpg

If you compare that even to my unstuffed version out of
"poo-poo tube" - as we say here at the countryside - you may
find by far more pronounced ripple even >1Khz at the
"High End" version.

I am not very ambitious concerning my sloppy experimental design,
but accidentally the shown "High End" version's measurement uses
comparable 1/6 dB/octave smoothing ...

That thing performs grotty to say the least.

IG's measurements also strikingly show, that one can do better.

I fear they made the thing that long, because they needed
delay and are very proud on having their design "phase accurate".

Instead of compensating any delay electronically - if
necessary - they made that extra long coupler, slotted only about
half of the length.

That seems contrary to all well performing K-Coupler style tweeters
on the market.

Imposing materials often seem to hide the view on the inner
logic of a certain design.
 
Last edited:
You are correct that the highest frequencies have their main lobe on axis with the tube. However, listening with a bit of tilt towards the slotted side seems to subjectively improve dispersion. I don't know if I have data on this anywhere. I suspect a larger tube might have similar propreties, shifted down proportionally.

IG

Yes i see, tilting it down a bit or at least directing it somewhere
below your head while listening seem's preferable.

Thank you for sharing your estimation and experience.
 
Farber certainly did go for an acoustical center alignement. A longer un-slotted portion likely also provides a better load on the diaphragm. freddi has tried a Farber-style tube for the HF and liked it IIRC.

Have you though of using a Klam assymetric projector for midrange? In a way, they are akin to a stubby K-tube.

IG
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.