GAYA2-Final, finishing the unfinished after 15 years

[ctrlx]

can you post the impedance curves on one graph, i.e. bare box / 100g / 200g etc?

 

[JanRSmit]

Tried to do so in Arta-Limp but can only do 2
Is there another way to do combine all?

 

[JanRSmit]

@ctrlx Here is a collection in VituixCad of all 10 impedance curves.

Note, the wiggles in low frequencies is from the working table when i touch the computer mouse to click ;-)

 

[ctrlx]

heres an example of what im thinking of
curves are separated out and no smoothing is used
we can see what the material and amounts used do to the imp curve

 

[JanRSmit]

Here is a VituixCAD project file with the impedance data enclosed as overlays, and the separate ZMA files also:

In the W000 impedance the primary modes of the cabinet are well visible, and are very quickly damped. What remains at around 400-500 Hz is most likely panel resonances of the test-cabinet. The TIB measurement in W000 shows this also:

 

[JanRSmit]

heres an example of what im thinking of
curves are separated out and no smoothing is used
we can see what the material and amounts used do to the imp curve

check my post with VituixCAD file.

 

[JanRSmit]

A quick analysis of the distortion measurements using STEPS (at 2.83V, measured at 30cm) , shows an issue in the range 200-600Hz with a peak at ~ 390Hz.

A assume it is a mechanical thing, cabinet resonance or something like that. Typically most pronounced in the uneven harmonics. It looks like it correlates with TIB measurement and with the impedance measurement.
To be continued.

 

[JanRSmit]

heres an example of what im thinking of
curves are separated out and no smoothing is used
we can see what the material and amounts used do to the imp curve

@ctrlx , can you tell more about the yellow curve how that was achieved?

 

[ctrlx]

yellow is long fibre wool (LFW)

red is Bonded Acetate Fiber (BAF)

note that bubble wrap beats both, above 1KHz

 

[fluid]

issue in the range 200-600Hz with a peak at ~ 390Hz

That looks like the basket resonance. You could try to put a support piece under the magnet to damp the mass spring by connection to the enclosure.

 

[JanRSmit]

@fluid, what makes you think in that direction, may i ask?

 

[JanRSmit]

yellow is long fibre wool (LFW)

red is Bonded Acetate Fiber (BAF)

note that bubble wrap beats both, above 1KHz

Thanks for the explanation.
BAF is indeed quite poor, we used to put the blankets with the side facing the driver, to get a bit better damping.
Long haired wool, aka baileys long haired wool i never used. I did trst carding of the wool we used, but that was quite poor in damping.

 

[Sean P Custer]

Forget about the speakers I love your coffee table !!! did you make it? lets see some more pics of the coffee table please..

 

[fluid]

@fluid, what makes you think in that direction, may i ask?

There was a big fuss in an ASR review of the Sointuva speaker where a distortion spike was found like that. Same sort of frequency region but slightly different as it was for the 6.5" woofer.

Information here and in surrounding posts
https://www.diyaudio.com/community/...ifi-woofer-speaker-builds.352063/post-7058058

 

[krivium]

Other source of info about it:
https://linkwitzlab.com/frontiers_2.htm#N

 

[JanRSmit]

I took some time off to check.
checked the bolts etc.
It is the cabinet itself. I have also a tweeter mounted in the box, and that when connected as source shows similar strong exitation when tapping on a panel.
The cabinet is only ment to test the wool and how much of it.
So not a big deal, other than to watch for it when mounted in the final enclosure.

 

[JanRSmit]

The four sides are quite similar in size/shape. I will glue some stiffening to those sides in different ways so to get different and higher resonance frequencies.

 

[JanRSmit]

Unlike the sointuva case in mine it is not an air leak by the way.

 

[JanRSmit]

Just a question: has anyone tried to mount a driver by its maget?
It is what @lrisbo suggested in the sointuva case, and with my mechanical/shipbuilding background i looked at it and concluded it is not trivial to construct shunt to some mass without disturbing resonances.

 

[hifijim]

I have looked into it, and I think it is difficult to do correctly.