Diy AMT

[solhaga]

What is the resistance and how thick alu did you use?

 

[båndsei]

Most of your questions are ansvered erlyer in this tread
The tape I bought was quite expensive.500 Dkr.50mm.x66meter.
alu is 7my.and resistance depends on width thicknes and length.
I aim for app. 4ohms.And usesual I succed.

Bernt.

 

[solhaga]

Yeah, I guess I have to read it from the start again
Thanks for your patience.

Perhaps my next question is answered as well, but I guess that you have tape on both sides of the alu?

7 um can be quite tricky to handle, I´ve tried 6 um once and it was impossible to not accidently tear it apart.

I don´t find 500 Dkr expensive, considering how many tweeters you can build.
I was thinking if any wider tape could easily be ordered so that one doesn´t need to overlap. Or does that pose problem when applying the tape?

 

[båndsei]

I use only one layer of tape.Exept at the ends,where I use 5mm.wide mylartape to guard the alu from tearing in the tool.
In Denmark i buy 7my alu with paper.Paper is easy to remove after handling.
No problem to get wider tape from my Danish supplyer.
And it is useably too.But costs maybe like 2 replacements diaphragms from ESS.

Bernt.

 

[solhaga]

Well, 7 um with sounds great. Could I ask you where you get it?

 

[båndsei]

Sure.
In Denmark it is easy to get some.You buy it in household stores like Brugsen,Kvickly,Irma and it is specified thicknes.Coop Danmark is the mother firm.It comes in 30 cm.wide and 10meter length and used for packaging food.
Hope it helps.

Bernt

 

[solhaga]

Ok, that´s a new one. I always have used pure alufoil, bought at the same kind of stores.
I never considered the paper/alu-foils. I´ll keep my eyes open then.

/Claes

 

[båndsei]

Here is my latest atempt to get low an less distortion.
Pleats 6.5 mm. 28 lines 40mm with ,and 170mm length:

Red is DIY green is ESS Amt.
Later i will show measures and materials from my folding tool

Bernt

 

[solhaga]

Any idea what might cause the higher distorsion around 500 Hz?
If that could be reduced it would seem that your DIY AMT is far better than the ESS AMT, THD wise.

 

[solhaga]

Concerning the Amplitude: any theory of why the DIY AMT has lower output in the lower regions? What are the Bl of the both AMTs do you reckon?

That the response is higher in the highs could be an advantage for my forecoming project: a line array of a number of DIY AMTs.
It could be useful in mitigating the comb effect. Also, no need to bother with getting the resistance high, 1 ohm would be enough if you have four or six in series.

 

[SEdwards]

By 6.5mm pleats do you mean the total depth? So traces of about 5 mm?

 

[båndsei]

This is my latest drawing .

Traces are 5mm and spaces 2,5 mm.on the drawing.Length 170mm, width 28x 7,5mm + 22.5mm.=232mm.
I use a double blade knife with 2mm spacing.Traces are then 5,5mm.
Increase of distortion might be becourse of ressonance at 500 hz.
Diaphragm is "softer" than previus models incl ESS model.

Bernt.

 

[båndsei]

I am not sure what Bl is.
Increase in level at higher frekvensies, might be caused by a larger areal of the diaphragm.
Planars act this way.
Lack of niveau at lower freq.,could be cauce of lower Q (softer suspension)

Bernt

 

[SEdwards]

Thanks for the picture. I was also looking over your Danish thread. I like this direction - I guess that the 20KHz+ diaphragms are more for marketing in the end. In their whitepaper ADAM talk about reaching 12 or 15 KHz with their midrange AMT.

Some Kapton tape arrived in the mail the other day, but I haven't had a chance to play with it much yet.

 

[RayCtech]

Increase in level at higher frekvensies, might be caused by a larger areal of the diaphragm.
Planars act this way.
Lack of niveau at lower freq.,could be cauce of lower Q (softer suspension)

Bernt

Any increase of level at higher frequencies are most possibly breakup / resonances in the diaphragm.

Lack of level at lower frequencies are not due to lower Q, but a combination of diaphragm area, magnet system and diaphragm properties / design and baffle design.

What are the SPL levels used for your measurements ?

I can see that you uses an 200Hz high pass filter in the measurements.

I measured my AMTs at 100 to 110dB SPL at 2.5 meter without any high pass filters. -3dB was then 100Hz and 16 - 17kHz in a dipole baffle (-30dB at 20Hz giving a normal ca. 12dB/oct fall below 100Hz due to the diaphragm area). When I measured the AMT driver in free air (suspended in a wire from the roof) and minimum 2 meter from walls and midway between floor and roof the -3dB in the lower end moved up to ca. 1500Hz.

Using a BLH (back loaded horn) I extended the low end to a -3dB of 60Hz with falling response down to 30Hz (like many smaller loudspeaker systems).
Used as full range without any kind of high pass filters and DC coupled amplifiers the AMT can play with an average level of 80 - 85dB with SPL peaks up to 100 - 105dB SPL.


A flat / linear (overall) and extended low frequency SPL response are a possible goal, but much work and many calculations may be required..

 

[båndsei]

This is a calibrated measurement made at 1 meter 2.8v and a little off axis:

Would you share some informations about dimensions ,depth of pleats ,number of traces, material ,magnet system?.

Bernt

 

[solhaga]

B times l: RayCtech said it better; the magnetic system. In this case the length (l) of the strips in the flux (B) with the current (i) equals the force (f).
Have you made any comparsion between the DIY and the ESS?
Perhaps it doesn´t matter that much...

I was also thinking of a too soft suspension when it comes to the lower SPL in the lows.
One experiment would be to stiffen the folds with some extra strips of tape. Perhaps every other in order to even the overall response? Or perhaps the outermost only.

 

[båndsei]

This is a measurement of the ESS 40x140x4mm. 2.8v. 1meter and a bit off axis:

Magnetic system (engine) too week for the load, (weight)?
I have a lot of measurements and diaphragms,so i have too compare results and decide wich ones to go for.
I think there is a problem in my measurement tool causing the increace in thd at 8khz.
I can show distortion at 2. harmonic and higher too.

Bernt

 

[RayCtech]

Magnetic system (engine) too week for the load, (weight)?
I have a lot of measurements and diaphragms,so i have too compare results and decide wich ones to go for.
I think there is a problem in my measurement tool causing the increace in thd at 8khz.
I can show distortion at 2. harmonic and higher too.

Bernt

When you increase the strength of the magnetic system you will also more easy reveal problem issues with the whole construction. The peaks, dips and spikes in the SPL measurements will be "worse", but it will be more easy to find and tweak the problem issues.

Do not trust and use the THD measurement directly as is for troubleshooting...
The THD result are among other dependent on the measured SPL level..
A reduced measured SPL level (dip) will show up as an increase in THD level.

If you invert the THD curve and overlays the SPL curve you will easily see how the THD and SPL interacts in the measurements...

Some of the peaks and dips and spikes are caused by flexing of the diaphragm pleats itself so the THD peak and SPL dip in the 8kHz area are most possibly an issue with the diaphragm / diaphragm assembly.

I have also observed that mechanical resonances in the magnetic / metal system can be triggered.
This happens normally in the 8kHz to 12kHz range.
When this happens the magnets / metal parts are both physically emitting sound and affecting the electromechanical system. If I remember correctly - I measured Mundorf AMTs to "sing" mechanically and emitting 60dB SPL from the mechanical assembly...
This was the "trigger" for me and I sold the Mundorfs and built my own AMTs.
When I first had detected this mechanical "singing" it was totally destroying any music listening.
Before this discovery I was not satisfied with the fidelity of the Mundorfs and it was this dissatisfaction that started the investigations that revealed the "singing" problem...

 

[RayCtech]

This is a calibrated measurement made at 1 meter 2.8v and a little off axis:

Bernt

For troubleshooting a impedance sweep / plot are one of the best tools.

You can then see resonances etc. as peaks / dips in the impedance plot.

In the SPL plot you will get a SPL reduction (dip) just below (in frequency) the impedance peak and a peak in SPL above. The SPL peak / rise (breakup / flexing) above the impedance peak can add several dB over a large frequency range.

If the aluminum impedance are as low as 1 ohm the amplifier itself may add to the measured "problems" and some of the SPL and THD variance are then caused by the amplifier.. When the diaphragm flexes / resonates the use of aluminum will create a current / voltage that the amplifier must fight. If the impedance are as low as 1 ohm most amplifiers will have serious problems - the higher frequency the worse the problem.
Due to this I use copper instead of aluminum, and I use non global feedback JFET class A amplifiers, and my AMTs are either 8 ohm or 32 ohm.