Shallow Mount Coaxial Three Way


A low profile, high performance three way coaxial speaker design for in or on wall mounting.

This was a design that I had floating around the back of my mind for a little while. Originally I was just looking at using the FTX0617 in something, the specs of the coaxial make it a poor choice for a general two way so I thought about what woofers I could pair it with. I started going through the woofers I had on hand until I came across the GBS250, the woofer modeled well in compact volumes and is obviously very shallow which is what inspired me to see how shallow I could make a cabinet using the pair while retaining usable extension. I did some basic enclosure math and modeling but the design sat only an idea until a forum member approached me with requirements that this speaker filled nearly perfectly at which point the real work started on the project.

This design was of special interest because I didn't know of any existing speakers that fill the same niche as this speaker would, the slender profile and yet high performance/capability. The main goals I was looking to achieve were a speaker that is designed specifically for on or in wall mounting (in a standard 2x4 cavity provided that the baffle was oversized and sat on top of the drywall), provide high output/dynamic capability, excellent dispersion characteristics and last but not least great sound quality.

Did it meet expectations?

The FTX0617 coaxial isn't an easy driver to work with but after taming it's high frequency response I find it provides excellent transparent sound quality in the treble without any objectionable traits. The high frequency dispersion characteristics from this driver are also excellent, it's rated at 100 degrees nominal and it certainly provides wider, more consistent coverage in the top octaves then any of the other 6.5" coaxials I have used.

The high frequencies blend near seamlessly with the midrange on the FTX01617 with all the benefits a coaxial provide, namely the point source like dispersion pattern without the drawback of off axis nulls or uneven power response in the vocal range inherent in more conventional multi driver designs. With the the load of the lower frequencies lifted by the woofer the 6.5" midrange on the FTX0617 really opens up resulting in an effortless and clear midrange presentation even when the volume is pushed.

The woofer may have surprised me the most, mainly from the bass extension it gave in the small cabinet being quite unexpected, as I was anticipating an f3 around 65hz when I was designing the cabinet, not the sub 50hz extension that resulted after the crossover implementation. In such a compact enclosure the woofer has no trouble staying in control with lots of power either. Excursion remains at fairly safe ranges up to 400w and the speaker really starts to come alive when you feed it decent power.

This design certainly met all the expectations I had for it when I first set out designing/building them and I would have no trouble recommending them for compact high quality LCR or surrounds if extreme output requirements are not a consideration. They are capable of getting loud enough for reference levels in small/medium rooms given enough power but won't win any SPL shootouts against proper Pro Audio designs due to the medium-low sensitivity and efficiency of the design. That said these clearly outclass more conventional Hi-Fi designs when it comes to their dynamics, clarity and composure at high volumes/power levels.


Sensitivity of this speaker is ~87dB @ 2.83v, however it is a 4 Ohm design so the nominal efficiency is only 84dB/1w. No free lunch when squeezing things into such a compact sealed enclosure yet retaining good bass extension.

Power handling is 200w RMS / 400+ Watts Program. The woofer is the limiting factor for power handling as the high efficiency coaxial is almost always just be loafing along at a few watts.

Impedance hits a minimum of roughly 3 ohms at 100 Hz but any amp capable of 4 ohms will have no trouble driving this speaker.

Crossover Design

I didn't really hold back on the crossover design as I was trying to extract the most I could from the drivers, the only real limitation being what I could physically fit into the enclosure.

Most of the work ended up going into the HF driver as it's response did not lend itself to a simple filter. Essentially it's a 2nd order highpass (though electrically acts more like a 3rd order near the crossover) but with a pair of notch filters and another pair of contour filters mixed with padding. I would have adding something to damp the peak at the top end but that rise starts so high (>18k) you don't really notice it and the peak also flattens out some as you move off axis.

The midrange was much easier to work with and therefore is a much more straightforward circuit, a 1st order highpass and 3rd order lowpass with some padding and paralleled components to get the required values and keep the size down. The highpass does interact with the impedance of the midrange resulting in an electrical slope that's somewhere between 1st and 2nd order (~9dB octave).

The woofer uses a 2nd order lowpass with a coil in series with the parallel leg of the filter. The coil results in a notch and increases the slope of the filter without disturbing the phase much. You loose the lowpass slope as it flattens out above 2 kHz but it stays at least 30 dB down for the remainder of the range and this woofer isn't doing much at those frequencies anyway. A large capacitor is placed in series with the woofer this helps to give a more linear electrical response near the crossover and the interaction with the impedance peak at the system resonance causes a high Q highpass which provides over 4dB of passive gain just below 50 Hz allowing the design to achieve it's surprising bass extension.

The crossover between the midrange and the tweeter is ~1650 Hz and between woofer and midrange is ~250 Hz.

The Slipstream was designed and the crossover tuned specifically for in/on wall use, because of that there is minimal baffles step compensation needed with how shallow the cabinet is. One of the other benefits of the shallow/wide cabinet is the minimization of SBIR. The reflection off the front wall does not cause as large or sharp of a cancellation as a narrower/deeper cabinet would and if mounting in wall that eliminates the issue entirely.

2022 update:

New measurements I took showed I previously measured with the mic positioned a little too closely which was masking some of the SBIR effects on the frequency response in the lower midrange. I ended up changing the L1 coil from 6.0mH to 4.0mH to improve the response in the 200-500hz range, it also minimized the dip that was seen ~150hz and slightly lowered a peak ~80hz. This filled out the sound in the lower midrange a little more.

After that modification I noticed the upper midrange now seemed a little pulled back and I did notice the 2-4khz range looked a little recessed in the measurements. I had to swap around a few components on the high frequency network to raise that range ~1dB without affecting the rest of the frequency range. R5/R6 from 16 to 20 Ohm, C8 from 4.3 to 4.0uF, R9 from 12.5 to 15 Ohm and R10 from 20 to ~56 Ohm (anywhere from 47-68 Ohms should work). That slight bump in the upper midrange/lower treble added a bit of detail and keep the overall sound more balanced and to my liking.

Optional: R10 can be removed entirely if you wish which will increase the entire treble range just under 1dB. I tired it like that for a little while and it gives the speaker a more forward and detailed sound but I also noticed the treble on some of the more bright recordings were a bit too in your face for my liking, which is why I added the 56 Ohm resistor there to lower the treble output ever so slightly.

I've also been listening to them with C5 swapped to 33uF instead of 68uF which results a tad less output in the lower midrange taking some load off the midrange driver.

Crossover Schematic:

Note Reversed Polarity on HF and MF Drivers.

*If mounting in-wall the C5 capacitor should instead be 33uF*

Frequency Response on axis and 20 degrees off axis:

Note: Speaker measured flat on ground facing up to simulate 2pi on wall mounting.

Measured outdoors, 55ms gate, 1/24th octave smoothing.

Off Axis Response (0-90 degrees):

Note: Speaker was measured on my turntable stand here and gated so low frequency response is not accurate.

Enclosure Design

The cabinet is 14" wide x 20" high x 4.5" deep I designed it around a 3/4" baffle and 1/2" for the rest of the cabinet to keep dimensions to maximize internal volume for the given footprint. 3/4" material could be used if the cabinet was made 1/2" wider/taller and 1/4" deeper (14.5 x 20.5 x 4.75").

The mid range chamber is 7.5" x 7" x 3.25" or ~0.1cuft gross, made from three 7.5" x 3.25" pieces glued in a U shape to the top of the cabinet. All the free space in this chamber should be filled with recycled denim, fiberglass or rock-wool leaving just a small cup for the motor/magnet of the coaxial to sit into.

The woofer chamber is ~0.35cuft gross, the crossover is a tight squeeze on the three walls surrounding the woofer and the remaining space filled with denim/fiberglass/rockwool. I was expecting an f3 in the 65-70hz range but the combination of proper damping and the series capacitor in the crossover circuit allowed that to be pushed down closer to 50hz.

Since this speaker is designed for on or in wall mounting changing the depth of the cabinet is not advisable.

Build Photos

I was commissioned to build 5 of these for a forum member so I have a collection of build photos from that process which show the speaker assembly in more detail.

Assembling the cabinets, here I'm getting ready for the baffles to be glued on.

Got Clamps!

Priming the cabinets.

Moving onto the paint.

I threw the drivers in to check how it looks.

Crossovers Galore!

MF crossover board mounted to left side of cabinet.

HF crossover board mounted to bottom of cabinet.

LF crossover board mounted to right side of cabinet.

Crossovers installed.

Installing terminal cups.

Damping in midrange chamber.

Damping in woofer chamber

Installing the drivers.

Grill frames assembled and chamfer cut.

I used a sandwich of 3/4", 1/8" and 1/2" MDF to make them thick enough to cover the woofer with enough room for 8mm of outward excursion.

After painting black and getting read for fabric installation. Using adhesive transfer tape which will allow the grill fabric to be stretched and held in place before stapling.

Closeup of the tape before the backing is removed showing how thin it is. It's basically just a thin double sided adhesive layer.

Stretching and attaching the fabric, have one side attached here.

Fabric attached.

Excess fabric cut away.

Grill assembly finished.