sound decoder ?

NW Modeling List nw-modeling-list at
Thu Oct 29 08:48:32 EDT 2015

The topic of bass speaker enclosure fabrication by amateurs (meant in the true, non-derogatory sense of the word) has been covered for at least 5 decades that I’m aware of—even back in the ‘60’s one could get copies of A. N. Thiele’s seminal work providing the basic equations for bass reflex enclosure design.  (That paper is still worth a look—Thiele was a true pioneer in the field, and his paper is revered.)  In the ‘80’s I built a successful set of “Klipsch horn knock-offs” using components and directions from Speakerlabs, who also published other enclosure designs for their speaker elements.  Electro-Voice still has a pamphlet available (on-line, see references below) that provides enclosure dimensions for their bass speaker elements.

The reason for attempting a ported (bass reflex) speaker enclosure design is to use the extra complexity of the design to extend the bass response of the speaker toward the low frequency end of the spectrum without increasing the volume of the speaker enclosure.

But there are a couple of things that make applying a bass reflex design challenging in a model:
1)  Design of a beneficial bass reflex enclosure is hugely dependent on the exact electrical and mechanical characteristics  of the speaker element.  Most manufacturers of speaker elements for model use aren’t publishing (and maybe don’t even know) the characteristics of their elements that are absolutely necessary for good bass reflex enclosure design.  Even worse, production variances even for model speaker elements that are nominally the same part number might be large enough that a bass reflex enclosure would have to be individually tuned to a specific speaker element purchased from a local hobby shop.
2)  For the speaker elements and frequencies we deal with, even in O-Scale and certainly in HO and smaller scales, the fabrication precision for the enclosures would have to be better than what could easily be achieved with model aircraft plywood or styrene sheet.  (Now that I’ve written this, I suspect this would be a great application for 3D printing!)
3)  If you enclosed a bass reflex speaker in, say, an HO-Scale tender, then the tender enclosure itself would further “load” the speaker element.  This additional loading would only be beneficial by pure serendipity (and more likely antagonistic to good sound reproduction), unless the tender enclosure was modeled acoustically (see Reference 3 below to begin to understand what that might mean), and the design for the bass reflex enclosure developed to compensate for (or be synergistic with) the effects of the tender enclosure.  I suspect this particular challenge is why the various model speaker companies do not offer ported (bass reflex) enclosure designs for their speaker elements—they would have to design them for specific tenders by specific model manufacturers in specific scales!  Not likely to happen!

That’s a daunting list of challenges, and therefore it’s not at all hard to understand why nearly all model speaker enclosures are closed, rather than ported.

A few references that interested people could investigate:, for those who just have passing interest;
file:///C:/Users/botte/Downloads/Bulletin10A.pdf<file:///C:\Users\botte\Downloads\Bulletin10A.pdf>, for those who want prescriptive design information;, for those who what to delve into the equations and understand the theory.

But I think it’s even more important to understand that the non-ported enclosures applied to model speakers are typically NOT tuned to the speaker elements they are designed for.  Certainly most modelers who build our own non-ported enclosures do not acoustically optimize their enclosure volumes to match specific model speakers.  In general, more enclosed volume is better in a non-ported enclosure.  So I really wonder why RTV-ing a speaker face-down on the bottom of a steam engine tender, over holes available or drilled in the tender floor, isn’t the best possible approach—since that uses all of the available volume in the tender?  Of course, in such an installation it is critical that all openings in the tender shell not covered by the speaker element be sealed!  (RTV works fine for that, too.)

So what is the value of the rather tiny speaker enclosures offered by the manufacturers?  Very simply, these tiny enclosures help because they cover one side of the speaker element!  If you realize that a speaker element is basically a plate that moves back and forth in the air to generate pressure waves that we hear, it’s very easy to realize that leaving the back of a speaker element open allows some “short-circuiting” of the pressure waves from the back side of the element to the front side—which greatly reduces the amplitude of sound that we would hear once we’re a foot or more away from the speaker element.  It’s that simple—the small enclosure covers the back-side of the element, preventing short-circuiting of the sound generate by the front side of the element.  And it really doesn’t matter, for this purpose, whether the speaker element’s front  or rear face is enclosed—we get the intended benefit either way.

But the smaller the volume of the enclosure, the higher the “low frequency cut-off” of the enclosed speaker.  That’s another way of saying that the enclosed speaker becomes less capable of reproducing low frequencies.  Why?  Well, the enclosed air acts like a spring—when it is compressed, the air pushes back on the speaker element.  This raises the “natural frequency” of the enclosed speaker.  The bigger the enclosure, the less the effect of the air spring, and the better the low frequency response of the enclosed speaker.

So the tender shell can serve the same purpose as an enclosure, and provide larger volume behind the speaker element than an “off the shelf” enclosure.  So why use a separate enclosure around a tender-mounted element??  Of course, diesel folks face a much tougher challenge, and it’s almost always much easier to use an off-the-shelf speaker enclosure in a diesel.  But for goodness sake, whether it’s a tender shell or an off-the-shelf item—use a sealed enclosure.

A third topic worth considering is the wavelengths of sound these speakers generate, in the context of human audition.  When the sound wavelength gets much longer than twice the distance between our ears, we lose the ability to localize sound accurately—i.e., we lose the ability to determine what direction it’s coming from.  My ears are about 7” apart, so for wavelengths longer than 14”, the direction the sound is coming from becomes increasingly hard for me to determine.  A wavelength of 14” corresponds to a frequency of ~965 Hz (Hertz; cycles per second).  I (and all other two-eared people) have essentially no ability to accurately localize sounds at 500 Hz or lower frequencies.  But above about 2000 Hz, I (and most others) can localize sounds very well—within a few degrees of the direction from which the sound is actually coming.

Why is this useful knowledge?  Well, suppose your speakers only produced sounds in the 200 Hz to 1000 Hz frequency range.  In that case, if you were more than a foot or two from your HO steam locomotive, you really wouldn’t be able to tell whether the sound was coming from its tender or its boiler.  But if your speaker produced sounds at frequencies between 2000 Hz and 10,000 Hz, you might be annoyed by the fact that it’s really easy to tell that the sounds are coming from the tender—which in a real locomotive (other than a triplex, Garratt, or SRR tractor articulated) would NOT be emitting any exhaust chuffing, or whistle, or bell, or air pump cycling.

Most of our tiny speaker elements are useless in generating much sound energy at or below 200 Hz—and I suspect they are “falling off” pretty darn fast below 600 Hz.  And the bigger (1” diameter and larger) speaker elements probably strain to produce sounds much above 2000 Hz—so sound effects like bells and whistles can tend to sound a bit damped when we only use “large” speaker elements.  This is why most home speaker systems have “woofer”, “mid-range”, and “tweeter” speaker elements in each enclosure.  These different speaker elements divide up the responsibility of reproducing the audible frequency spectrum, taking on the low, middle, and high frequencies respectively.

And it’s why we have dual-speaker sound decoders, like the QSI Titan series.  You can mount a “large” speaker in a steam loco tender to reproduce the lower frequency (longer wavelength) sounds that we can’t “localize” anyway, and a “small” speaker in the boiler that reproduces the higher frequency (shorter wavelength) sounds that we humans can localize quite well.  Boiler-mounting a “small” speaker puts the higher frequency sounds closer to the stack, bell, whistle, and air pumps than tender-mounting, while tender-mounting the “large” speaker still allows the sound system to reproduce the lower frequency sounds (which we can’t localize well anyway) that make the experience much more realistic.  By all means, put the “small” speaker in an enclosure, and aim it at the boiler opening around the gearbox, or up the stack, if the stack hole is open to the boiler.  Oh, my, does this extra effort pay off!!!

But please consider using the entire tender as your sealed enclosure (except for openings in the tender floor) for the “large” speaker, as described above.  If you don’t like the results, then you can always go ahead and use one of the “off the shelf” enclosures, and aim the output up through your coal bunker.  I’m doubting you’ll feel the need.

Finally, make sure that nothing in or on your tender or loco is loose and likely to “buzz” when you crank up the volume.  Any tender detail item that gets to buzzing will just ruin the whole sound experience.

-Eric Bott

From: NW-Modeling-List [mailto:nw-modeling-list-bounces at] On Behalf Of NW Modeling List via NW-Modeling-List
Sent: Tuesday, October 27, 2015 6:08 PM
To: nw-modeling-list at
Cc: NW Modeling List
Subject: Re: sound decoder ?

You are absolutely right! Oops! I guess old age is sneaking up on me! And the port design opens up a whole different set of variables! LOL But these little speakers just do not move enough air for that! I have found that the "Air Suspension" Install work wonders for HO sound decoders. Anything that ups the low end and stops the wrattles goes a long way!
Thanks for correcting me!
Ray Russell Sr.

In a message dated 10/27/2015 3:07:35 P.M. Eastern Daylight Time, nw-modeling-list at<mailto:nw-modeling-list at> writes:
Actually a sealed box is called Air Suspension, Bass reflex implies a ported design.
Mark Lindsey
Stuck in the 1930’s
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the NW-Modeling-List mailing list