Aino gradient - a collaborative speaker project

Elias asked if I have experience with dipoles. The answer is yes, in a way.

The first speaker I played/tuned with was a 8" dipole radio speaker like some of these. I was 10-12 years old.

We had an 1950's Philips tube radio and a turntable, but at Ainola's attic (the house of my grandparents) I found two separate dipole speakers. I attached them to the Philips and to a portable transistor radio (typical plastic box from '70s). I was disappointed of the sound. Obviously it had way too high impedance and a strange response curve.

Old radios with 2-3 speakers in wooden boxes were aperiodic and they sounded marvellous, much better than the new cheap Japanese transistor-portables. Many young guitarists used old radios as electric guitar knobs.

My uncles had made a huge container for a tube radio and a turntable. They had moved the speakers to the front panel of the cabinet/console. These were popular diy-projects during 1960's.
 
Last edited:
lolo, since you are a Shigeru Ban fan, I think we shoud visit also this church in Jyväskylä - Kuokkalan kirkko

kuokkala_church_l260810_2.jpg
 
Jon Marsh is a speaker guru at HTGuide forum Mission Possible. He seems to be a wonderful person and a versatile talent. He has a dipole design called Arvo Pärt (an Estonian composer) He is more keen at box speakers and he likes Butterworth 3rd slopes for xo. Here is his Isiris project (please register as a member)

As 3-way xo I like Duelund and B3. But it might be a little difficult to get those working in this case. B3 is worth trying.

Crossover types in wikipedia
 
Last edited:
There is a crisis in the air here...
My wife has seen my skethes and she sees the outcome of my project like these!

Ok, well I wanted to be an archicect but I had to forget it because I am such a lousy at drawing and in mathematics! My eldest son is fortunately a student at TIT Civil Engineering

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

Biolan_kompostori_kivi-harmaa.jpg
 
Last edited:
Edge is a wonderful neat little program by a marvellous programmer and researcher PhD Svante Granqvist

I simulated the response of my mid driver Beyma 12MwNd

Seems like it is wise to make an ellipsoid little baffle for it.

I have placed an order for miniDSP 4x10, U-MIK and a software plugin
 

Attachments

  • Edge Beyma spl.png
    Edge Beyma spl.png
    799.5 KB · Views: 962
  • Edge Beyma spl ellipsoid.png
    Edge Beyma spl ellipsoid.png
    665.7 KB · Views: 927
edit was closed...

Seems like that with an ellipsoid baffle the
- 6dB is at 200Hz
- baffle diffractions start at 3500Hz

The simulation should be analyzed with the spl curve of Beyma in mind (this was measured in an infinite baffle)

Suits You, Sir!
 

Attachments

  • Beyma 12MWNd spl.png
    Beyma 12MWNd spl.png
    83.2 KB · Views: 862
Last edited:
Edge is a wonderful neat little program by a marvellous programmer and researcher PhD Svante Granqvist

I simulated the response of my mid driver Beyma 12MwNd
Edge is fine when you know what is important.;)
"3" for the Speaker source density is way too few. Make it "20". The cone (or dome) should be filled with small squares. With "3" you only show what the surround contributes to SPL. :p

And DON'T put any baffle around the driver!!! Leo Kärkkäinen of Nokia Research Center can tell you why. Nothing but a magnet mount. It's already too large for the Xover frequency.

Rudolf
 
Rudolf, please - don't be so mysterious!

We Finns are a naion of only 6 million people, but I don't happen to know Mr. Kärkkäinen... Tell us a bit more about this, a link maybe? Do you mean a 12" is too big for M/T xo? But I want to use the high directivity area!

M/T xo is determined mainly by directivity of drivers in mind. The second factor is distortion, but I don't have that info about Fountek. The response is even enough to not be an issue here.

I made new plot swith 21 speaker sources. The response looks similar but cleaner. The -3dB at low end is around 230Hz with the baffle and 250Hz without - not a big difference.
 
Dr. Granqvist has also a smaal polarity simulator, XDir

XO@2500Hz means trouble with vertical polar response and comb filtering near xo frequency

Coaxial arrangement has no lobing issues around xo. The Neo3.5 is too big to be positioned coaxial with Beyma 12" (acts as a diffusor)

miniDSP gives a possibility to do a 48dB/oct xo (LR8) - is this the least bad solution here?
 

Attachments

  • xdir 2800 250mm.png
    xdir 2800 250mm.png
    18.5 KB · Views: 169
  • xdir 2300 250mm.png
    xdir 2300 250mm.png
    20 KB · Views: 124
  • xdir 2300 0mm.png
    xdir 2300 0mm.png
    16.3 KB · Views: 132
... but I don't happen to know Mr. Kärkkäinen... Tell us a bit more about this, a link maybe?
Download http://www.mellowacoustics.com/articles/Disk_in_a_circular_baffle.pdf
Look at Fig. 5-7. The relation between "b" and "a" is explained in Fig.1
Try to understand that there is a relationship between radiation pattern and baffle size.
If you want to learn more about that, go to the download section of my website dipolplus.de and download the second file "How open baffles work ..."

Rudolf
 
Okei, Rudolf, Offene Schallwand (OB)

The Beyma 12" works as a dipole up to 1000Hz and then dips at 1500Hz on-axis, when it loses it's directiviy and dipole function?

The Gradient 1.x has M/T xo at 1500Hz - a little too high?

I guess this means that I need a 4-way system for transitioning from the 12" to tweeter's monopole area. Sounds like NaO Note and LX521...

Those guys use a 10cm (4") upper mid drivers. But because my lower mid is a bigger one, perhaps a 5" for me -the directivity would suit better.

SEAS CA15RLY

Peerless HDS135 a special offer this week!

SB 15MFC