• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Baffle Edge Diffraction Take A Look at a Horizontal Speaker

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
967
Likes
396
Location
Yerevan "Sydney Born"
An interesting conversation developed in the Directiva thread on baffle edge diffraction and I thought we could continue it here.

I can't say I understand all the technical aspects of baffle edge diffraction but it seems a few points(from Heismann acoutsics page) has given me the basics as well as Mr Peabody's effort in the Directiva forum.
1. Give as much as baffle as possible
2. Give as little as baffle as possible
3. Give unsymmetry to the drivers
4. Use a waveguide
5. It seems chamfering also helps(general advice)

The drivers in the design below are offcentered with very little baffle to the left plus to the right there is quite a a lot of baffle.

For education sake it seems the designer has taken in consideration the basics but is there a way to tell how he has worked it out? And if it is effective? or if its precisely calculated? Is the baffle to the right enough to eliminate the problem? Is the baffle to the left little enough to eliminate the problem? Or just has he used the basics? I am not worried about baffle step loss but more about comb filtering and other interferences that DSP cannot mitigate.



1620393821401.png
 

Attachments

  • 1620393579826.png
    1620393579826.png
    323.7 KB · Views: 87
  • 1620393687360.png
    1620393687360.png
    52.2 KB · Views: 106

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
It's incorrect to say "eliminates the problem" as what that design does is "smear" diffraction over a larger range of frequencies. As for lobing .. that would depend on where and what is used as a filter network.
 
OP
Trdat

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
967
Likes
396
Location
Yerevan "Sydney Born"
It's incorrect to say "eliminates the problem" as what that design does is "smear" diffraction over a larger range of frequencies. As for lobing .. that would depend on where and what is used as a filter network.

So is that smearing of diffraction fairly effective? If I had a design with a waveguide on that type of design would that further direct frequencies away from the baffle?

Regarding lobbing, the centre to centre distance between tweeter and midwoofer is very close it shouldn't be a problem right?
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
So is that smearing of diffraction fairly effective? If I had a design with a waveguide on that type of design would that further direct frequencies away from the baffle?

Regarding lobbing, the centre to centre distance between tweeter and midwoofer is very close it shouldn't be a problem right?
The driver layout is very similar to my K&H O300D and the Neumann KH310. Both speakers have symmetrical waveguides for mid dome and tweeter, as far as I can tell. Both speakers are excellent wrt price.
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
Your intuitions are correct, but the waveguide is by far the best solution. Filleting corners with a radius less than 1" (2.5cm) doesn't do very much.

Troels Graveson has a speaker called the PMS (Poor Man's Strad) which has a wide baffle, I like this approach in a passive speaker since it places the baffle step much lower than normal. Baffle step and diffraction are sort of the same phenomena.

If you want to see the effects of baffle geometry in 3d sound radiation, play around with VituixCAD. Basically, diffraction doesn't have a huge effect on the diffuse field / power response of the speaker, so you would think it is unimportant, but the diffracted waves are essentially secondary sources with slight delay. In other words, they can be very bad for intelligibility and subtle cues, but they will not kill tonality overall.
 
OP
Trdat

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
967
Likes
396
Location
Yerevan "Sydney Born"
I did a quick vcad baffle sim with a tweeter with an sd of 5.7.

One line is what it would look like if the tweeter was in the middle. The other two are in the proposed spot except one has a 44mm edge radius.

What does it exactly mean by 44mm edge radius?

But yes, it seems offset make a big difference plus the 44mm edge radius is even better. Thanks for sharing.
 
OP
Trdat

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
967
Likes
396
Location
Yerevan "Sydney Born"
Your intuitions are correct, but the waveguide is by far the best solution. Filleting corners with a radius less than 1" (2.5cm) doesn't do very much.

Troels Graveson has a speaker called the PMS (Poor Man's Strad) which has a wide baffle, I like this approach in a passive speaker since it places the baffle step much lower than normal. Baffle step and diffraction are sort of the same phenomena.

If you want to see the effects of baffle geometry in 3d sound radiation, play around with VituixCAD. Basically, diffraction doesn't have a huge effect on the diffuse field / power response of the speaker, so you would think it is unimportant, but the diffracted waves are essentially secondary sources with slight delay. In other words, they can be very bad for intelligibility and subtle cues, but they will not kill tonality overall.

So essentially we are reducing the amount of secondary sources and there delay and this layout on the baffle has tackled it fairly well it seems.

Yes, I am looking at modeling the concept but with a waveguide which will increase the tweeter and midwoofer c-c but it is still within Kimmoto's 1.4 time wavelenght rule. I can round off the edges, I can also put the mid woofer in a small waveguide like the above mentioned Neumann KH310 just not sure if its worth the effort.
 

headshake

Active Member
Joined
Apr 20, 2019
Messages
297
Likes
238
What does it exactly mean by 44mm edge radius?

But yes, it seems offset make a big difference plus the 44mm edge radius is even better. Thanks for sharing.
Like 617 said, 1" of rounding does not do much. I rounded it to an amount that actually does change the 1khz on up at every axis. If you mess with vcad, you will see the change as you increase the edge radius- 44mm was just a number I picked. Anything over 30mm does nice things.
 
OP
Trdat

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
967
Likes
396
Location
Yerevan "Sydney Born"
Like 617 said, 1" of rounding does not do much. I rounded it to an amount that actually does change the 1khz on up at every axis. If you mess with vcad, you will see the change as you increase the edge radius- 44mm was just a number I picked. Anything over 30mm does nice things.

Great. Got it. I will get onto VCad.
 

alex-z

Addicted to Fun and Learning
Joined
Feb 19, 2021
Messages
910
Likes
1,685
Location
Canada
The Parts Express C-Note kit is a prime example. Despite having a centrally located tweeter and minimal roundover, it doesn't suffer much from diffraction because the waveguided tweeter is crossed over around 3000Hz. So that 1000Hz bump from baffle diffraction that is shown in a sim is dramatically reduced.

It has other issues like port resonance and cheap crossover, both is other is a pretty great bookshelf speaker.
 

hex168

Senior Member
Joined
May 29, 2020
Messages
396
Likes
338
Look at off-axis as well. One effect an asymetrical layout can have is to make the off-axis response different to each side, which makes the speaker less amenable to EQ.
 
Top Bottom