Serious issues with tweeter diffraction - how to fix them?

[fjhuerta]

Hi all!

I've been fighting against tweeter diffraction on my Natalies for a long time (maybe some of you will remember this issue).

The cabinets I built have no round overs. I think this is part of the problem. I followed the Nat-P baffle design.

The graph I included shows the diffraction issue I have. It's pretty ugly.

I haven't been able to control it with *anything*. I've tried adding foam to the long side, short side, I've used pillows, everything to try to get the diffraction under control. I haven't had any luck at all... although I've been able to vary the diffraction effect somewhat. I've never been able to find a decent enough compromise, though.

I also measured the off-axis response of the tweeter, thinking that the diffraction issues may go away. They didn't - the 2 KHz "hump" is always there.

So far, I've designed a crossover around the flare issue, with two notches - one parallel, and one series. The transfer function plot looks really strange, but I've been able to achieve 4th Order L-R, and the overall frequency response (woofers added) looks fine. (I, of course, have to add the original Nat-P crossover had to be removed, and I settled for a conventional parallel design).

But I'm wondering if I'll be listening to the "diffracted" tweeter response, and whether this will hurt sound quality.

I was wondering - is this all a result of me not rounding over the cabinet? As far as I've been reading, the offset tweeter position should have helped with the diffraction issues.

[cjd]

Roundovers would very likely have helped.

Have you applied a heavy (wool) felt layer vertically on either side of the tweeter? Say, 1/4 to 3/8" thick, directly next to the actual tweeter dome (so covering the face).

You can find real wool felt at craft stores still, and if you can't find thicker pads you can get the thin stuff and sorta roll it up to be thicker for testing purposes.

C

[fjhuerta]

Roundovers would very likely have helped.

Have you applied a heavy (wool) felt layer vertically on either side of the tweeter? Say, 1/4 to 3/8" thick, directly next to the actual tweeter dome (so covering the face).

You can find real wool felt at craft stores still, and if you can't find thicker pads you can get the thin stuff and sorta roll it up to be thicker for testing purposes.

C

Hi C!

I didn't try wool. I used foam... pillows... everything I thought that could help in absorbing high frequencies. The foam was around 1" thick. When placed directly to the sides of the tweeter, it didn't help at all!

I was thinking yesterday about where to find some wool. I haven't got a clue as to where to start around here, but I'll surely be looking for some.

[ThomasW]

Javier,

Custom gaskets are made from high wool content felt. The mfgr's will occasionally sell it by the running foot.

I can provide a link to a US supplier but don't have a clue where you should look in Mexico.

I use a local supplier I found in the yellow pages. Other's with no local supplier use www.mcmastercarr.com Scroll down to the Raw Material/Springs section and click on the Felt link.

It doesn't need to be very thick 1/4" -3/8" is fine

[fjhuerta]

Thomas, thank you very much for your link!

I saw the felt page, and I found out "discs" and "strips" with adhesive. I know of a place where they sell both of them. I'm quite sure it's "synthetic" wool (whatever that means!) though. I'll buy a couple of strips, just to see if they work.

[Chris7]

That 2k hump almost looks too broad for a diffraction effect. It's almost an octave wide. Also, if it was a diffraction effect, its effect would change off-axis, which you say it doesn't. Could it be related to something else other than diffraction?

[fjhuerta]

That 2k hump almost looks too broad for a diffraction effect. It's almost an octave wide. Also, if it was a diffraction effect, its effect would change off-axis, which you say it doesn't. Could it be related to something else other than diffraction?

I don't know... I'm pretty sure it should have disappeared off-axis, but it didn't. Then again, it did disappear or move around that frequency when I added foam around the short side of the enclosure. So I suppose it's diffraction.

I had never seen such a broad or pronounced peak, though. I measured the other speaker's tweeter, and the result was identical.

Which is pretty strange... I flush mounted the drivers, and followed the plans. I don't understand what's wrong.

I don't suppose the tweeters have some sort of defect, although that would explain why the hump remains off axis.

[cjd]

Um, how does diffraction go away off-axis? It's a wavelength issue relative the distance between the tweeter and other items on the baffle OR the baffle edges...

The thin wool felt I referred to that you could fold/roll up can be found at any fabric store. Not ideal but it does the job for testing (I've measured it)

C

[ThomasW]

Thomas, thank you very much for your link!

I saw the felt page, and I found out "discs" and "strips" with adhesive. I know of a place where they sell both of them. I'm quite sure it's "synthetic" wool (whatever that means!) though. I'll buy a couple of strips, just to see if they work.

You don't want synthetic wool, you want one of the multiple wool blends. The higher the cost the more wool per synthetic ratio.

[TurboFC3S]

I know, apply some serious smoothing to that graph ... put away your test equipment, and convince yourself the problem is gone Maybe start a new project to take your mind off it.

[fjhuerta]

Um, how does diffraction go away off-axis?
C

Thanks C, I'll also get some real wool. Maybe I'll use my wool suits and fold them, just to thest the theory.

I always thought that diffraction had to do with the baffle size / shape - so that when you started listening off-axis, its effects were less audible until, at 90°, they disappeared completely.