The Case for Mounting Speakers Diaphragm-out

[jdcolombo]

Hi everyone.

There has been some debate (here and on other forums) about the "correct" or "best" way to mount a cell-phone type speaker in an enclosure: should it be with the diaphragm facing in, or the diaphragm facing out?  I've pretty much always done out-facing, unless the situation was such that I just had to mount the speaker the other way.  My rationale was pretty simple: since the manufacturers test their speakers with the diaphragm out, they must think this is "correct." However, a lot of "sugar cube" speaker are sold with instructions to mount the speaker with the diaphragm facing in the enclosure itself, and I've seen at least one document that claims that "facing in" produces better sound.

Another modeler, Larry Hanlon, who posts on the ESU LokSound Yahoo site has done a series of tests of two such speakers, the Knowles Dumbo (13x18mm) and Donau (11x15mm), and shown via charts and graphs that mounting these speakers facing out results in much better sound - more extended frequency response in the "bass" (not really bass, but more like lower midrange) and smoother response overall.  His work can be downloaded from the Yahoo Group's file section, here:

https://groups.yahoo.com/neo/groups/loksound/files/Micro%20speaker%20measurements/

While technically Larry's results only confirm that "outie" mounting is best for these two speakers, I suspect his results will hold up across the range of cell-phone-type speakers (and he is going to test some more).  Larry also points out something that I've said many times: enclosure size matters - a lot.  And bigger is generally better.
So if you're really interested in "best" sound, maximize your enclosure size (remember, it doesn't have to be just a plain rectangular box around the speaker; it can be a larger box with the speaker offset to one end; an L shape, a cone, a triangle, whatever works).

John C.

[tehachapifan]

Thanks, John! Good info! I think the primary reason I mount the enclosure diaphragm-in is that there's often a nice, clean surface to attach the enclosure to (I typically glue the enclosure to the face rather than do a wrap-around installation to keep the overall dimensions in check)....and I don't have contacts and wires to try to deal with. That said, since I've been having trouble getting good lower tone reproduction in my latest installs that have particularly "bassy" sound files, I will give this a try.

[soo]

So to seal the holes where the wires come out..cca..tenex..bondene?
Wyatt

[MK]

I would use something that seals, like caulking or thick glue.  The stuff you mentioned is more for bonding and too thin to seal a hole.

[jdcolombo]

So to seal the holes where the wires come out..cca..tenex..bondene?
Wyatt

I just use thick CA glue.  Just make sure to get it all around the wires and into the exit hole.  Works just fine.

John C.

[RBrodzinsky]

I've gone to using Bondic to seal the wire holes, and around the speaker/enclosure seam. Subjectively, this has increased the sound (I don't have any monitoring equipment to test).

[Cajonpassfan]

John, thanks for this info. Unfortunately I can't get into Yahoo to view the link, so I have a question for you.

I've done installs both ways, but not necessarily with identical speakers and enclosures so it's apples and oranges here.
Is the sound better because of diaphragm orientation, or because there is more space within the enclosure? That is, if one were to make a slightly larger enclosure to adjust for the enclosure space taken by the speaker, would it still hold true? Or do I have it backwards, since I can't access the link
Otto K.

Edit: rereading the post made me realize I did have it backwards. So the standard "facing out" mount actually has less enclosure volume, given identical enclosures, and still sounds "better"?

[Lemosteam]

Some interesting stuff here.

With 3B printing I wonder if there is anything within this info that I could do to increase lower frequency range within my enclosures (folded horn?)

[woodone]

Working with N scale enclosures it is some time necessary to decrease the enclosure so it and the speaker will fit. What I have found is that has the space in the enclosure decreases so does the sound. With the sugar cube speakers I have not mounted any of them with the diaphragm pointed out has most, the cube speakers have the input wires on that side. So no holes to plug. With the larger speakers I have mounted both ways and not noted a difference, but I have had large enclosures to work with. I have a DB meter to measure with but I don't know it that gives you a good idea has to what you are hearing. BTW If you have a smart phone,  there is an app for a DB meter.     

[jdcolombo]

John, thanks for this info. Unfortunately I can't get into Yahoo to view the link, so I have a question for you.

I've done installs both ways, but not necessarily with identical speakers and enclosures so it's apples and oranges here.
Is the sound better because of diaphragm orientation, or because there is more space within the enclosure? That is, if one were to make a slightly larger enclosure to adjust for the enclosure space taken by the speaker, would it still hold true? Or do I have it backwards, since I can't access the link
Otto K.

Edit: rereading the post made me realize I did have it backwards. So the standard "facing out" mount actually has less enclosure volume, given identical enclosures, and still sounds "better"?

I'm pretty sure that Larry kept enclosure size constant when doing his experiments. 

However, you (and Woodone) are correct that enclosure size affects speaker performance.  I wrote a short article for the N-Trak Steam Annual 2017 about this.
Knowles used to use an enclosure of 1 cubic centimeter (1000 cubic mm) to test the performance of its speakers, and my own experiments indicate that such an enclosure is about optimal, particularly for a 13x18mm.  However, you can get excellent performance with somewhat smaller enclosures for smaller speakers.  For the Soberton 8x12mm, 700 cubic mm seems to be a sweet spot, with not much subjective sonic improvement beyond that size.  But the Soberton sounds pretty good down to about 500 cubic mm.  Lemosteam's 3D printed enclosure comes out to about 576 cubic mm, and I think it works great.  But once the enclosure size for the 8x12 dips below 500 cubic mm, there is a noticeable reduction (to my ears, anyway) in both overall loudness and lower-midrange (e.g., 500hz) response. 

As for the potential difference in enclosure interior air volume as a result of mounting face down (with the enclosure sealed to the front face edge of the speaker vs. face up (with the speaker inset in the enclosure itself), I wouldn't worry much about that.  The difference is actually pretty small.  That's because the speaker back side is open to air movement, so you don't actually lose the full depth of the speaker in a "face up, inset" mounting; all you really lose is the physical space taken up by the back plate itself and the plastic sides; that's a little bit, but not the full depth x length x width of the speaker itself.  And a "face in" mount actually ends up being a little narrower inside, because it is not wrapping around the speaker; for an inset speaker mount, the interior size of the enclosure is the actual width x length of the speaker; but for a "glue it on the face" enclosure, the interior size is a tiny bit smaller.  So these things tend to cancel each other out, IMHO.

The moral of this story is that a bigger speaker is NOT always a better choice for a given installation.  It may well be that a smaller speaker that gives more room for an up-sized enclosure will outperform a larger speaker with an impaired enclosure volume.  Given the choice between an 8x12mm in a 600 cubic mm enclosure and a 9x16 in a 400 cubic mm enclosure, the 8x12 likely will sound better overall.

Also remember that an enclosure does not have to be a box built exactly around the speaker.  The enclosure could be a longer box, with the speaker offset to one side, to provide additional enclosure space.  Or it could be a cone, a cylinder, or any kind of 3D shape you want.  Boxes are easy to build, but they aren't the only shape you could use.

John C.

[jdcolombo]

Some interesting stuff here.

With 3B printing I wonder if there is anything within this info that I could do to increase lower frequency range within my enclosures (folded horn?)

Hi John.

Engineering a proper folded horn or even a bass-reflex (ported) speaker enclosure is complex, and requires knowing a lot of details about the speaker itself - including its resonant frequency, Q, and overall free-air response curve.  I seriously doubt one could design a generic enclosure like this for our tiny speakers that would improve the sound over a properly-sized sealed box.  Theoretically, you could do so for a specific speaker, but I doubt whatever improvement is gained would be worth the engineering effort, given the fact that the tiny speakers we use just don't have much in the way of useable output below about 400hz.  Ten times zero still equals zero . . .

John C.

[wmcbride]

John,

Your threads on sound are always informative and a pleasure to read and apply.

[Cajonpassfan]

John,

Your threads on sound are always informative and a pleasure to read and apply.

I second that! Thank you, John.
Interesting point about a smaller speaker with optimal enclosure vs. a larger one with an impaired enclosure. I always go with the largest speaker available for the space, but I may need to rethink that...
Otto K.

[peteski]

I second that! Thank you, John.
Interesting point about a smaller speaker with optimal enclosure vs. a larger one with an impaired enclosure. I always go with the largest speaker available for the space, but I may need to rethink that...
Otto K.

As I see it (not being an acoustical engineer but knowing just enough of the subject to offer unsolicited free advice), what you describe is a delicate balance game.  The ideal sound transducer (speaker) would have total isolation between the front and back of the sound producing cone.  As in being mounted in an infinitely large and very stiff divider.   That way the cone would have the maximum freedom to move forward and back (like a piston moving in the air) creating sound waves.

Since this is impossible to achieve in a typical speaker system (especially in such small examples we use in our models), the next best thing is to have one side of the speaker cone enclosed in an air-tight chamber (speaker enclosure).  That isolates the front of the speaker cone from the back of it. This way there is no cancellation of the sound waves which would occur if the speaker was not mounted in an enclosure.  If you are interested in more details, a good start is https://en.wikipedia.org/wiki/Loudspeaker_enclosure . But with the tiny size of our acoustic transducers and enclosures, getting any good low frequency response is not very realistic. They just don't have the ability to move enough air to produce bass sounds.

Going back to your statement, the air in a sealed enclosure acts as a damper for the speaker cone. If the enclosure is very small then the speaker's cone movement (produced by the electric signal feeding it) will be dampened, reducing the level of produced sound.   In a larger enclosure with smaller speaker the larger air volume will have less of a damping effect allowing the cone to deflect more (producing louder sound).  But again, in either case, with our small speakers and small enclosures I think that either scenario will be a wash.   This is all much more complicated that what I just stated, but I just presented it as a quick overview.

At least that is how I see it.

[Greg Elmassian]

Total isolation between the front and rear is indeed called an "infinite baffle" and should (within the working parameters of the speaker) give the best frequency response.

This is rarely used though, not because it would be too big (which is actually true), but there is nothing to control the speaker cone excursion at lower frequencies (most cone movement).

So you need some kind of back pressure to keep the cone going so far that the motion becomes nonlinear (extend of spider and surround) or the voice coil bottoming in the magnetic gap or the voice coil leaving the magnetic gap. All of these circumstances cause distortion (if not destruction of the speaker)

So to get desired volume (sound level) people put the speaker in a sealed enclosure (which as you say damps the speaker) to limit the overtravel of the speaker cone/voice coil.

You can also make a tuned enclosure (ported, bass reflex, etc) where you pick a frequency range to "enhance" and then design the "path" from the back of the speaker to the outside such that the back wave comes out in phase (albeit one cycle later) with the front wave, thus almost doubling the sound energy in this frequency range.

Normally in an enclosure like this, you are trying to increase the output in the bass range, where most speakers have difficulty, and it's also the frequency range most amenable to taking twists and turns from inside to outside.


Greg