Stupid question about flush mounting a tweeter

[contoursvt]

...how 'flush' is close enough so any difference would be virtually inaudible as far as diffraction and stuff is concerned. I thought I went down 1/8th inch which was the approx thickness of the Vifa faceplate, but after I routered it, I saw that I could have probably gone down another 1/32nd of an inch. Just my margin of error when lining up the depth guide I guess. I know its probably nothing but should I bother going town that tad more or leave it?

I'm also not sure how thick a few coats of paint and some clear might be so maybe thats going to make up for the difference when the paint builds up on the front baffle.

[johnnyrichards]

Compared to the impact the woofer cone has, and the ripple introduced by baffle edges, that 1/32" is inconsequential.

Personally, I like the tweeter to be above the baffle a bit (like yours) for aesthetic reasons.

[contoursvt]

Ok sweet. I will leave it then Was the first time I was playing around with my sears router and now I'm just covered with dust. Damn you Mr. MDF for being so dusty. LOL.

Compared to the impact the woofer cone has, and the ripple introduced by baffle edges, that 1/32" is inconsequential.

Personally, I like the tweeter to be above the baffle a bit (like yours) for aesthetic reasons.

[PWR RYD]

Like Johnny mentioned, I also like to have flush mounted drivers sitting just a bit proud (1/64"-1/32") of the the baffle.

[charlielaub]

It's a complete fallacy that you have to "flush mount" the tweeters. What you have to do is make sure that the diffracting edge around the tweeter is non concentric about the tweeter dome. The worst situation occurs is when the diffracting edge is the same distance from any point on the dome edge: this is what you get with the standard round flange. While flush mounting can remove this diffracting edge, there is really too much emphasis put on ye old flush mount. There are other ways to skin this cat.

Let's say you don't have a router or don't want to bother to flush mount. Do the following:

• find some appropriately thick material (e.g. thin wood the same thickness as the tweeter flange)
• surface mount the tweeter
• cut a hole in the extra material that you will apply equal in size to the tweeter flange OD
• glue it on around the tweeter flange.

This will break up and distribute the edge diffraction, will diminish the effect on frequency response (e.g. the ripples will be much smaller), AND you will reduce the diffracting effect of the woofer surround because the tweeter will be raised WRT the woofer.

-Charlie

[Ludo]

It's a complete fallacy that you have to "flush mount" the tweeters. What you have to do is make sure that the diffracting edge around the tweeter is non concentric about the tweeter dome. The worst situation occurs is when the diffracting edge is the same distance from any point on the dome edge: this is what you get with the standard round flange. While flush mounting can remove this diffracting edge, there is really too much emphasis put on ye old flush mount. There are other ways to skin this cat.

Let's say you don't have a router or don't want to bother to flush mount. Do the following:

• find some appropriately thick material (e.g. thin wood the same thickness as the tweeter flange)
• surface mount the tweeter
• cut a hole in the extra material that you will apply equal in size to the tweeter flange OD
• glue it on around the tweeter flange.

This will break up and distribute the edge diffraction, will diminish the effect on frequency response (e.g. the ripples will be much smaller), AND you will reduce the diffracting effect of the woofer surround because the tweeter will be raised WRT the woofer.

-Charlie

Charlie, Maybe I missed it, but what shape should this "extra material" have. Any pics of an example? Louis

[charlielaub]

Charlie, Maybe I missed it, but what shape should this "extra material" have. Any pics of an example? Louis

You can use any shape you want. The goal is to distribute the pathlengths (in the radial direction) from the source (the dome) to the diffracting edge (the edge of the new material) among as many different lengths as possible. For instance, the unmodified round flange of the tweeter has a single pathlength for all directions, thus it is a "worst case". If you make the shape in to an elongated rectangle (e.g. longer in one direction than the other) and the tweeter dome is towards one end, there is a more broad distribution of pathlengths. But you could use a star shape, or something irregular in shape - it doesn't matter as long as there is a broad distribution in path lengths.

It's not diffraction that is bad, but rather too much diffraction happening at the same pathlength to the listening point (e.g. the same path of source->diffracting edge->listening point). Think of the holy grail: the sphere. Don't think that there is not diffraction happening - as the sound wave propagates away from the source and around the sides of the sphere diffraction is going on. It's due to the fact the the diffraction occurs along the entire curving side of the sphere that the diffraction signature is so smooth - it's like many infinitesimally small little diffracting edges all adding up bit by bit. This spreads out the diffraction effects so that they do not reinforce at any one frequency, and the resulting transition (in the frequency domain) is very smooth.

As long as you distribute the path lengths, the diffraction effects will be minimized. Note that path length distribution will vary as you move off axis!

Keep in mind that significant diffraction of the sound waves will still occur at the cabinet edge (assuming the typical rectangular speaker box) so you also should try to distribute the pathlengths to the cabinet edge, too. The approach I have described in my post is addressing the diffraction from the tweeter flange itself, or from the new material added. The cabinet edge is yet another source of diffraction to be dealt with.

Jeff Bagby's Excel based diffracton modeler is helpful for checking out how to position the tweeter on the cabinet and how to shape the extra material to minimize diffraction issues as much as possible.

SL has a page on diffraction from the cabinet edge, showing measured frequency responses for a few different baffle shapes:
http://www.linkwitzlab.com/diffraction.htm
A similar thing applies to the tweeter edge itself.

-Charlie

[Ludo]

You can use any shape you want. The goal is to distribute the pathlengths (in the radial direction) from the source (the dome) to the diffracting edge (the edge of the new material) among as many different lengths as possible. For instance, the unmodified round flange of the tweeter has a single pathlength for all directions, thus it is a "worst case". If you make the shape in to an elongated rectangle (e.g. longer in one direction than the other) and the tweeter dome is towards one end, there is a more broad distribution of pathlengths. But you could use a star shape, or something irregular in shape - it doesn't matter as long as there is a broad distribution in path lengths.

It's not diffraction that is bad, but rather too much diffraction happening at the same pathlength to the listening point (e.g. the same path of source->diffracting edge->listening point). Think of the holy grail: the sphere. Don't think that there is not diffraction happening - as the sound wave propagates away from the source and around the sides of the sphere diffraction is going on. It's due to the fact the the diffraction occurs along the entire curving side of the sphere that the diffraction signature is so smooth - it's like many infinitesimally small little diffracting edges all adding up bit by bit. This spreads out the diffraction effects so that they do not reinforce at any one frequency, and the resulting transition (in the frequency domain) is very smooth.

As long as you distribute the path lengths, the diffraction effects will be minimized. Note that path length distribution will vary as you move off axis!

Keep in mind that significant diffraction of the sound waves will still occur at the cabinet edge (assuming the typical rectangular speaker box) so you also should try to distribute the pathlengths to the cabinet edge, too. The approach I have described in my post is addressing the diffraction from the tweeter flange itself, or from the new material added. The cabinet edge is yet another source of diffraction to be dealt with.

Jeff Bagby's Excel based diffracton modeler is helpful for checking out how to position the tweeter on the cabinet and how to shape the extra material to minimize diffraction issues as much as possible.

SL has a page on diffraction from the cabinet edge, showing measured frequency responses for a few different baffle shapes:
http://www.linkwitzlab.com/diffraction.htm
A similar thing applies to the tweeter edge itself.

-Charlie

So are you proposing that adding some irregularly shapped material around the tweeter, with the same thickness as the tweeter flange, is better than flush mounting the tweeter? Call me skeptical but I'd expect that flush mounting would be the better way to go and the slight, 1/16th of an inch or smaller, discontinuity caused by not making it perfectly flush would be negligible.

[fastbike1]

Charlie never said it was "better" than flush mounting. He said that you didn't have to flush mount (i.e. more complicated). Charlie's suggestion ends up with the tweeter being effectively flush mounted, just with simpler holes.

So are you proposing that adding some irregularly shapped material around the tweeter, with the same thickness as the tweeter flange, is better than flush mounting the tweeter? Call me skeptical but I'd expect that flush mounting would be the better way to go and the slight, 1/16th of an inch or smaller, discontinuity caused by not making it perfectly flush would be negligible.

[Jeff B.]

So are you proposing that adding some irregularly shapped material around the tweeter, with the same thickness as the tweeter flange, is better than flush mounting the tweeter? Call me skeptical but I'd expect that flush mounting would be the better way to go and the slight, 1/16th of an inch or smaller, discontinuity caused by not making it perfectly flush would be negligible.

It depends on how you are using the tweeter. If you are mounting the tweeter in a small minispeaker baffle where is close to equidistant from three edges of the cabinet, then Charlie's method would actually smooth the diffraction compared to simple flush mounting.

Jeff B.

[Ludo]

Charlie never said it was "better" than flush mounting. He said that you didn't have to flush mount (i.e. more complicated). Charlie's suggestion ends up with the tweeter being effectively flush mounted, just with simpler holes.

Ok, so he is proposing this as an alternative to flush mounting then? It seems much more complicated to me so if there is no benefit vs flush mounting, even a mediocre job of the latter, then I don't see the point.

[Ludo]

It depends on how you are using the tweeter. If you are mounting the tweeter in a small minispeaker baffle where is close to equidistant from three edges of the cabinet, then Charlie's method would actually smooth the diffraction compared to simple flush mounting.

Jeff B.

Interesting example; so in this case Charlie's approach would potentially be an improvement over flush mounting, but how does one go about coming up with the best shape, without potentially making things worse? Not using your software, I'm not sure if its able to model this level of detail?

Louis

[JimHRB]

Good article by Zaph on the subject.

http://zaphaudio.com/mtg-surface.html

[dlr]

It depends on how you are using the tweeter. If you are mounting the tweeter in a small minispeaker baffle where is close to equidistant from three edges of the cabinet, then Charlie's method would actually smooth the diffraction compared to simple flush mounting.

Jeff B.

Sorry guys, I have to disagree with several points, all of it from empirical testing. Not picking on you Jeff, just posting all of my thoughts here.

The basic idea of surface mounting, then adding a flush-mount add-on is a good idea. I've done that using a rectangular piece of cork for a while with a Morel square flange tweeter (MDT-44). But as stated, the baffle edges are still a factor. If the tweeter is still placed equidistant to the edges, there's essentially no improvement to the most important diffracting points, the baffle edges. Maybe I've misunderstood your statement here, Jeff. A surface-to-flush piece will do little for the main baffle edges no matter where the tweeter is located on the baffle.

I still found that the rectangular piece had a significant diffraction signature. I used felt on this piece that improved both the offset piece and the baffle edges. But right up front let me say that there's no data that I've seen that documents audibility of the area of impact from such an added piece, since it's much higher in frequency that that of most baffle edges, so I'm basing my comments on measurements that do at least document the physical acoustic changes.

I also think that in most cases it would be better to flush mount to the baffle since that usually helps with acoustic offset, though it isn't large. That may be the least important overall. A tiny bit proud, that's measurable, but probably not audibly significant.

The diffraction that will change related to an adjacent driver will change little. The diffraction from the curved frame is actually not significant and changes radically with angle. It's already very distributed just by the geometry and proximity to the tweeter, plus on most listening axes, it's more reflective. The most important diffraction is related to the cone and that will change little between flush and non-flush or more forward mounting due to the wavelengths involved. It goes to the point I've made frequently, the most important diffraction in a system is often from a large driver close to a tweeter. A single piece of felt between drivers is the first and on occasion the only felt I apply, especially if the tweeter has been offset on the baffle.

Better than flush mount? I have to totally disagree on that one. It can at best be no worse with regard to diffraction, but if we're nit-picking, it's worse due to added acoustic offset that is for most configurations already placing the tweeter in front of the next driver.

dlr

[Jeff B.]

Sorry guys, I have to disagree with several points, all of it from empirical testing. Not picking on you Jeff, just posting all of my thoughts here.

The basic idea of surface mounting, then adding a flush-mount add-on is a good idea. I've done that using a rectangular piece of cork for a while with a Morel square flange tweeter (MDT-44). But as stated, the baffle edges are still a factor. If the tweeter is still placed equidistant to the edges, there's essentially no improvement to the most important diffracting points, the baffle edges. Maybe I've misunderstood your statement here, Jeff. A surface-to-flush piece will do little for the main baffle edges no matter where the tweeter is located on the baffle.

I still found that the rectangular piece had a significant diffraction signature. I used felt on this piece that improved both the offset piece and the baffle edges. But right up front let me say that there's no data that I've seen that documents audibility of the area of impact from such an added piece, since it's much higher in frequency that that of most baffle edges, so I'm basing my comments on measurements that do at least document the physical acoustic changes.

I also think that in most cases it would be better to flush mount to the baffle since that usually helps with acoustic offset, though it isn't large. That may be the least important overall. A tiny bit proud, that's measurable, but probably not audibly significant.

The diffraction that will change related to an adjacent driver will change little. The diffraction from the curved frame is actually not significant and changes radically with angle. It's already very distributed just by the geometry and proximity to the tweeter, plus on most listening axes, it's more reflective. The most important diffraction is related to the cone and that will change little between flush and non-flush or more forward mounting due to the wavelengths involved. It goes to the point I've made frequently, the most important diffraction in a system is often from a large driver close to a tweeter. A single piece of felt between drivers is the first and on occasion the only felt I apply, especially if the tweeter has been offset on the baffle.

Better than flush mount? I have to totally disagree on that one. It can at best be no worse with regard to diffraction, but if we're nit-picking, it's worse due to added acoustic offset that is for most configurations already placing the tweeter in front of the next driver.

dlr

Dave, I would never accuse you of nit-picking, you wouldn't do that.

However...... regardless of the improvement flush mounting makes to the acoustic offset, that, in and of itself, has nothing to do with diffraction, so let's leave it out as reasons go in a diffraction discussion.

I admit my post was vague and deserved to be called out. My only real point was the oversimplification that being equidistant from several edges is the worst scenario for diffraction. An oblong baffle with the tweeter off-center would randomize the distances more and be smoother. You are right, doing this on a rectangular baffle doesn't help much.