Update 31Oct25 ~ Nest GR8 Room Audio Upgrade... I moved the new Peregrine TABAQs to the Nest GR8 Room. Here they are, still unfinished, in the room in a good location against the oblique corners flanking the monitor, and raised a foot. I measured in pure direct stereo, both with and without the BDS correction circuit, and the improved location really obviated the need for correction IMO. I rebalanced/re-EQed the system, using Denon's internal DSP, for 5.1 surround for cinema (Dolby D/PLII) and music (DTX Neo 6). Such a great upgrade from the NHT mains. Definitely time to spin up a double bass reflex for the center channel now.



Tang Band W4-1337SDF drivers... I have some keen "leftover" Tang Band full range 4" titanium W4-1337SDF drivers. They were strong potentials, and tested, in the FrugalHorn FH3s during a bake off. I liked them and kept them but they lost in that cabinet to the Mark Audio Pluvias. I also have a spare sheet of 12mm Baltic birch plywood. So, I decided to build some new speakers with the leftovers. I wanted to employ some new acoustical physics ideas and started out on paper & computer. I compared a quarter-wave transmission line design to a double chamber bass reflex design in mathematical models. Here's the driver spec...

DRC TABAQ... A mass loaded (the air in the speaker and port) QWTL design by Bjorn Johannessen for an alternative Tang Band 3" driver was modeled in King's MathCad software; it's called the TABAQ (Tang Band Quarter-wave) and has been proven by DIYers and commercial builders/sellers. I 'built' something similar (in hornresp) to fit my 4" driver well and accommodate its Thiele-Small parameters. The Qts of the driver is on the edge of effective for this kind of speaker, at .34, but it is super efficient at 89dB. I futzed plenty with King's math too but needed to find an alternative to MathCad first (which only works on Windoze)... that never panned out well enough on macOS or Linux.

A MLQWTL cabinet design accentuates some bass (below the driver's resonant frequency) by delivering an in-phase lower frequency (a quarter-wave) from the rear throw of a full range driver, through a tuned port. The entire 'line' of the speaker from closed end to port exit is a vital dimension to get a quarter wave 'exactly' and the port dimensions are key for tuning that low frequency. I've done a design prior including a taper along the line (which helps with resonances), but not a forward firing rectangular port cabinet with a straight line (stuffing will be important). Here is the hornresp modeling of my TABAQ with (horn) section dimensions, 80gr of polyfill into 2/3 of the 'line,' and the acoustic power response (just look at the low end and ignore the bumps; else will ultimately be smoother). I should get ~1/2 octave at least below 70Hz out the port, or 50Hz. The modeling shows higher but I think I'll do better; hopefully I can even get to 45Hz.


DCR (Double Chamber Reflex)... A double chamber reflex (or dual reflex) speaker cabinet employs two separate but connected chambers each ported to the outside for accentuated bass response. Such a design performs similarly to a straight bass reflex cabinet except that distortion can likely be reduced (and perhaps size too) by its being more efficient in using the throw length of the driver over two resonant frequencies. Since the TB W4-1337SDF has a somewhat short Xmax of 2.5mm (I know there is no standard measure but relative to other TBs, it's short), I figured improved efficiency might work well in such a cabinet. Such designs have been implemented by DIYers from designs like the Cordelia from Scott and Bandit from Roman, and Claudio Negro's work. DCR cabinets can be smaller and the overall port length (which for some tunings can be too great to be a straightforward fit) is distributed throughout the cabinet in smaller pieces. Here is the Bandit, a tuned DCR for a Tang Band 4" driver not unlike my leftover.

Modeling again was done using hornresp. A standard vented (bass reflex box) for this driver would be ~ 4L and would get down to 70Hz basically.
A DCR box would be a bit larger and deliver the same but smoother bass. I don't know why port 1 looks longer in the diagram; it is correct and equivalent to the others in the parameter settings. The response of this cabinet may be too weird at 170Hz. I didn't model deadening the smaller chamber, which'd help.
Decision Time... I'm gonna start by building the MLQWTL, a pair of TABAQs. It just appeals to me a little more and will likely have better low end performance. If I have sufficient scrap materials from the 12mm and the other 15mm and 18mm scraps I have, I might also built the DCR and compare them.
Building My 'Peregrine' TABAQ... I headed to Summit (where my tabled power tools reside) and followed my simple cut sheet. The TABAQ is a very straightforward project with straight cuts and butt joints; stuff was sliced up and ready to assemble in <75min. The only modestly tricky part was putting a 101mm diameter hole in a 110mm wide front baffle. Bjorn never eased the back of the front driver hole and with just 12mm plywood, I didn't either. A Peregrine falcon, found frequently above my ridgeline in COS, is small but can fly downward faster than terminal velocity ~ like my light tower TABAQ loaded with the W4-1337, it slays unsuspecting prey.
The build continued through the day but with less sawdust in the PM. I added some no-impedance poly-holders at 2/3 of the line length. I installed the binding posts (found some nice cheap ones for just $10/4pcs). I weighed out 80g of poly for each speaker.
They are done but for the sanding and staining. They sounded best to me when propped up a smidge in front as they are pretty short with the drivers well below ear level when listening seated. Overall they are very impressive with plenty of bass without a sub and great mid to highs. I did get the bump at 45Hz for which I hoped. I could use a baffle step loss correction circuit; the baffle is very small. The SPL sweep shows this and is a little wobbly. I will work that with room placement ultimately. I could use DSP as an alternative to either the BSL or other defects.


Update 24Oct25 ~ Baffle Diffraction Step (BDS) Loss Correction, Circuits, & Testing... I've modeled and built circuits for narrow baffle speakers (like these) before. Lower frequencies sound 6dB lower below 8X the width of the baffle as to wavelength. (1/8 * 13560 in/sec)/5.3in ~ 320Hz for these. That's about right from initial testing (in a sub-optimal room location). The circuit which will reduce frequencies above this by 6dB uses a 1mH inductor in parallel with a 7Ω resistor in the powered signal path. The parts I used and circuit design are shown below along with my completed 'packaging' of the circuit. $20 got me parts for 2 pairs of these BDS loss correction circuits. I don't generally like to just 'create heat' but in the interest of a flatter SPL sweep, I'll do it.
As mentioned far above... I don't feel I need the BDS correction circuit in the speakers' final locations.