acoustical filter for tweeters ?

[Eldam]

hi,

is it possible to make an acoustical filter for an 1'' aluminium dome tweeter which is too brillant and cause listning fatigue?

it has a metal grill and I think of something with cotton wool or a remover make-up pad fixed with blue tack or adhesive tape ? Maybe foam ? The higher frequency are near the center of the dome so do i place a cotton wool or cotton shet just upon the grill in the center ? I tied with different sucess...

I think after reading some newsletter about my model that I have some ringing at 5K Hz (a bump between 4 and 6 Khz) or higher and wanted to make an electrical notch on the existing fourth order crossover...but don't know how and have no tools for measurement...

bad ideas ?

 

[xrk971]

Just a guess here - try a disc of larger porosity reticulated open cell foam (looks like a motorcycle air filter sponge foam) and adjust thickness to suit taste. Make sure foam doesn't actually touch driver dome. Adding a resistor in series or L-pad lets you dial down the tweeter. Adding a resistor in parallel also has an effect near resonance. See the thread on X-overs, this post in particular: http://www.diyaudio.com/forums/multi-way/189847-introduction-designing-crossovers-without-measurement-2.html#post2587249

 

[tinitus]

split/thin toilet paper hanging in front of tweeter works 😉
(attenuation may depend on quality, layers, thickness, etc)

since its an alu tweeter, its probably a spike, and needs more 'caretaking'
but try the paper trick ... you might be lucky

 

[Eldam]

Tinitus,

The tweeter is already a big **** (or the crossover?) ! and sure toilett paper is too thin for that !

i prefer to be positiv and try the xrK971 way (thank you for the answer)

Anybody here on DIYA try acoustical filter and do measurement after to where the attenuation is?

 

[tinitus]

you don't know without trying

 

[Pano]

Yeah, I done it with felt trying to attenuate some highs on axis while leaving the off-axis alone. It does work, but it wasn't easy to get right (for me).

Let us know what you find.

 

[speakerdoctor]

Have you considered a trap filter in the xover to pad down the spike? that's what's normally done.
Another thing to consider is thick, needled felt. About 1/2 inch thick or so. That is what Spica used about both the tweeter and bass/mid. The felt pad can be seen in the pic at the link below. It's located just below the grille.

Spica TC-50 / TC-50i Datasheet, The Spica Speaker Enthustiast

http://speakerdesign.net/felt_amelioration/feltssenter.html

 

[Eldam]

Hummm I have to take the smarter decision on that sunday... just 3 toilett paper sheet left !

 

[Cal Weldon]

Keep in mind that any fabric type material will have an increasing effect as the frequency goes up so you might want to consider a shelving network, notch filter or L-Pad.

 

[Eldam]

Yes Speakerdoctor that's what I talk when I wrote about notch

But don't know how and exactly where the bump is (no tools). i thought about a larger resisitor at the entry of the crossover (genuine is 10 homs // 1 uf polycarbonate) but I risk to reduce the cut and it doesn't avoid the bump. maybe a notch to reduce all the tweeter curve with a // self and a cap in serie of the + polarity? I think to the 2969 ISO curve... The tweeter begin at 2600 Htz.

i know there were a second version of Boston Lynfield 500L which have the same tweeter and midrange and charge than mine 400L and solve this issue. I don't find the shematic of the crossover on Internet.

 

[Eldam]

Cal, it's not a fabric dome but an aluminium dome.

 

[Eldam]

Hummm i'm scared with the idea to show this picture cause americans are in the habit of doing lawsuits... but the people of Boston acoustic are nice guys.

So after asked some spare to the lady next door (I think it's over for a futur dinner !) I took this picture.

And the shematic (4 order tweeter + second order midrange)

 

[Eldam]

Another thing to consider is thick, needled felt. About 1/2 inch thick or so. That is what Spica used about both the tweeter and bass/mid. The felt pad can be seen in the pic at the link below. It's located just below the grille.

very interisting Speakerdoctor, on your link the bump is 4.5 /5 kHz maybe like mine ! I have to try too.

The cotton wool and hadesive tape have a strange effect : more transparency, a little more definition (because less reflexion?) but the bright is always here... it's just a little warmer in low value....

the felt pad is a lot used in France by Mulidine manufacturer which like acoustical filter behind woofer and around tweeter.

So two things to try : reticuled foam and felt pad ("needled" felt is felt with little holes if I translate ?) and why not toilett paper after all... if it doesn't work I will use it after crying !)

 

[weltersys]

So two things to try : reticuled foam and felt pad ("needled" felt is felt with little holes if I translate ?) and why not toilett paper after all... if it doesn't work I will use it after crying !)

The problem with all the above (and you can add T shirt cotton material) is they all have the greatest attenuation at high frequencies.

For example, say 2 thickness of cotton fabric placed in front of the tweeter cuts 10 dB at 16,000 Hz, it may cut 5 dB at 8,000 and only 2.5 dB at 4000 Hz.

If the annoying peak is at 4000 Hz, the fabric has done little to remove the problem, and the attenuation of the upper frequencies may make the problem seem worse.

A passive notch filter (or smoother tweeters 😉 )is really the only solution short of a narrow band equalizer.

 

[Eldam]

For example, say 2 thickness of cotton fabric placed in front of the tweeter cuts 10 dB at 16,000 Hz, it may cut 5 dB at 8,000 and only 2.5 dB at 4000 Hz.

A passive notch filter (or smoother tweeters 😉 )is really the only solution short of a narrow band equalizer.

It's interesting yours values with the fabric because it's proportional and the curve is diving (I really don't know if there is a bump at 4500 because the crossover of my 400L is different of the 500L one & two serie).

Ok Cal, know understand what you wrote ... sorry : my bad english !

I considered change the tweeter but after it's another speaker ! And i don't want to waste these beautifull speakers... I will sell them to youngers ears who like accurate treble most of time.

The passive notch is difficult for me to design and set up by laking of experience with that and tools for real measure... i prefer sell them...(but I can't, it's like "an old maîtresse") and go for a tweak with active filter or someting like Maggies 3.7 (but too expensive for me here in France) or new Kef (I like the plain sound of old Kef but don't know with the knew aluminium line.... hé, comme les Boston Lynfield !

I considered ErigStik design with B&G planar but surmise it's too clear and bright (for my perszonal taste i prefer A plain and matt sound with details an dtransparency which is hard to do like for example the real good Hansen The Prince V2... tweaked Scanspeak Revelator tweeter).

So thank you guys for the mecanichal tips... I will test all of them. Did you know that the fabric grill of the Lynfield are made with Lacoste Polos' textile ? Written on the owner manual !

 

[freddi]

you could try something like EV did with their hard -dome - a piece of felt or foam with a round hole smaller than the dome's diameter - maybe Art remembers the size? - would not tackle a peak but might sound cool 😀

 

[Eldam]

Thank you Freddi for the input,

the dome is a 1". So I fix a round sheet of foam or feld in front of the dome (I can do that with the solid grill).

That the inverse of the feld around the tweeter for the 4 to 5 kHz bump, so I have to try both solutions and the the both in the same times...

 

[Pano]

For example, say 2 thickness of cotton fabric placed in front of the tweeter cuts 10 dB at 16,000 Hz, it may cut 5 dB at 8,000 and only 2.5 dB at 4000 Hz..

Yes Art, that's the basic problem I ran into with my felt experiments. 🙁

 

[Greebster]

Had tried this before and from memory was something like a flared conical or flat flared V using needled felt with the outer part covered in reticulated filter foam with a porosity of ~20/inch about ¼" thick. Felt shouldn't be too thick. Then cut the peak of the V rounding it out so no contact with the dome. Now if the felt is thick enough this cutout will leave a small round or elliptical hole in the center of ~½" directly over the dome... think that was one way, working with the density at oblique angles managed this task best. Still a little fuzzy in my head, twas a long long time ago in a galaxy far far away 😉

 

[Eldam]

So, i decided to experiment and follow the advices upon,

I try to find feld and open reticuled foam without spend too much money.

First I have tried toilett paper and why not like St Thomas believe with experiment!

i tried with cotton wool and it was special : more details but the sibilance was always here and the sound was a little covered but above the bad sibilance i want to kill.

Here i tried with beautifull three layers toilett paper for delicate a...udiophile that I named for the experiment : "Can Peak Atenuator?"

first The paper was upon the tweeter but not on it, just behind the grill cover.

i test with two cd : Musica Nuda live at Fip Radio France : the reccording is naturaly too bright, the voice of the female singer goes very high and the micro was often satured by that (I was at the live) and another cd with the soft but strong Ben webster.

Well, the attenuation is too soft and somewhere above the glitchy peak, near null effect, and sibilance already here (the tweeter is : 2600 cuted to 35 000 Hz).

next try : the three layer "Can Peak Attenuator?" on the grill dome tweeter (nearer and with adhesive tape around : not open around like the cover grill test).

After feld and fabric, and after foam... to be continued