Small Speaker Enclosure Design vs. What I see on the Market

[billfitzmaurice]

The response curve is only part of the equation. You need to look at the maximum SPL chart to know what it can actually produce. Your version of WinISD can't do that, or a lot of other things. Get WinISD 0.7.

[DHembree]

The resonance and compliance of the driver will determine what box size will be required. There is a very wide range in 3" drivers. For example, you can get to 40 Hz with the Tang Band W3-1876S in a reasonably small enclosure, but you sacrifice max output. The NS3 on the other hand, as shown by the green trace in your plot above will only extend to about 50 Hz in a similar size enclosure. The Audioengine A2+ you have linked to only extends down to ~65 Hz. The other big factor is that the A2+ are active, self-powered speakers. They may very well have boost in their electronics to allow a smaller than optimal enclosure, but still have "extended" frequency response.

Thanks for this info. I am encouraged to know the A2+ is only good to 65Hz. Maybe I need to get a pair of Tang Band W3's for comparison against the NS3.

Is the response curve of my 162 in^3 reasonable, or should is there room for improvement?

[1100xxben]

The resonance and compliance of the driver will determine what box size will be required. There is a very wide range in 3" drivers. For example, you can get to 40 Hz with the Tang Band W3-1876S in a reasonably small enclosure, but you sacrifice max output. The NS3 on the other hand, as shown by the green trace in your plot above will only extend to about 50 Hz in a similar size enclosure. The Audioengine A2+ you have linked to only extends down to ~65 Hz. The other big factor is that the A2+ are active, self-powered speakers. They may very well have boost in their electronics to allow a smaller than optimal enclosure, but still have "extended" frequency response.

[DHembree]

Thanks for all the input!

I will certainly check out the Sprite design, but need to know this design process as I will be needing to adapt and design several various enclosure sizes.

Paul, I was primarily referring to bookshelf style speakers akin to the Audioengine A2+ and the like. Their relationship between box volume and driver size does not appear to comply with what the calculators suggest. I obviously do not have the T/S data for the drivers they use, but cannot see how they can be so radically different than other similarly sized drivers.

I have uploaded a screen capture of the NS3 suggested Vb (blue trace) vs. my proposed 162 in^3 design (green). My inclination is to push the corner as far to the left while maintaining the flattest overall response. The displayed curve is my compromise. Feel free to comment.

Thanks so much!

[philthien]

Commercial speakers, of all sizes, tend not to worry so much about the correct Vb as they do the WAF.

Yep, and the enclosure is actually the most expensive "component" in the system. The smaller the cabinet, the lest the cabinet cost, the less the packaging costs, the less shipping costs, the more wives approve, there are many reasons cabinets are too small for optimum performance.

[Paul Carmody]

When you say "commercially available speakers," do you mean actual passive speakers, or do you mean the Bluetooth/Wifi/Sonos kinda thingys?

[Chris Roemer]

Are you going to measure the T/S parms yourself ("in-hand" drivers, as opposed to mfr's specs)?
MOST 3" drivers don't reach anywhere near as low as an NS3/ND90, hence the smaller box sizes.

[spasticteapot]

I am designing/building my first speaker using T/S data. I have ordered three different 3" speakers for evaluation, but am focusing on the Aura Sound NS3-193-8A.

The NS3-193-8A is near-as-makes-no-difference identical to the Dayton ND90 8 ohm. This is the driver used in Paul Carmody's Sprite soundbar - just chop it in half and lose the amp, and you have passive stereo speakers.

[Sydney]

... optimized figures for Vb? ...

"The whole point of fitting a woofer to a particular box volume is to control the response characteristics of the combination" ( from the Loudspeaker Design Cookbook )
With that in mind think of a points in a performance envelope rather than a strict "optimum." The designer gets to set response characteristics.

[DHembree]

So I'm not the only who battles WAF?? Good to know!

Thanks for the info. I'll look for WINISD.

[fpitas]

Commercial speakers, of all sizes, tend not to worry so much about the correct Vb as they do the WAF.

A sad truth.

[billfitzmaurice]

Commercial speakers, of all sizes, tend not to worry so much about the correct Vb as they do the WAF.

[dcibel]

Solution is to actually model the cabinet response and determine if the speaker meets your design criteria in a cabinet size you are looking for.

Download WINISD alpha and go from there, ditch the "online calculator".

[DHembree]

I am designing/building my first speaker using T/S data. I have ordered three different 3" speakers for evaluation, but am focusing on the Aura Sound NS3-193-8A. I have found some on-line calculators that produce Vb based on the speaker's T/S parameters. All of them suggest (ported) enclosure volumes that seem quite a bit larger than many of the commercially available speakers that use the same size driver. Could it be that these commercial speaker designers are using less than optimized figures for Vb? The enclosure I want to design has to fit into a space that would not allow for the Vb numbers the calculators are suggesting. Any help is appreciated!