Flex-8 - Dayton DA215-8
The Dayton DA215-8 shares the same frame as the DS and SD series woofers but uses an aluminum cone and slightly different style surround the combination of which provide a very smooth frequency response, probably tied with the Silver Flute variant for best overall in that regard (+-1dB 100Hz - 10Khz).
It does have the lowest sensitivity of the bunch coming in a hair under the DS215-8 at roughly 84dB 1W/1M. Bass extension is good with an f3 of about 50Hz (4pi) with an f6 near 42Hz. These would provide very good bass performance in room for most music without a sub. I used these with the port full length with tuned them to ~35Hz, this provides a more gradual taper of the bass towards tuning which I feel sounds better and makes the bass tighter and more natural then a higher tuning giving a flatter bass response but sharper rolloff does.
Distortion performance is mostly decent though the strong breakup resonance from the aluminum cone does cause a spike in the harmonics that land on it mainly the 3rd order. This is not a big issue in listing but I also include the ability to implement an optional distortion trap filter which will reduce the peak in those harmonics, more detail on this filter below. The breakup itself is well controlled by the lowpass so it does not affect the fundamental response in a negative way at all.
The provides good linearity up to about 100dB above which it starts to run out of linear travel as a quick increase in distortion and compression can be seen above the 100dB sweep level. Use of a amp in the 100-200w range should give you most of what this speaker is capable of, obviously you can use lower powered amps if high output levels are not required.
The high frequency performance/sound quality is largely the same in all the designs so this section will likely be a cut/paste for each. The LaVoce DF10.101LS with the Celestion H1SC-8050 is a very nice combo, the smooth easy to work with frequency response results a very clean clear uncolored high frequencies with remarkably low distortion and compression. The high frequency driver has to work very little in this design so its composure is maintained well beyond the point where the 8" woofers are begging for mercy. The only negatives are the slightly tighter coverage of the horn then I would have liked and the early drop in high frequency output past 16kHz. For the cost I don't mind as they are not glaring issues that you notice immediately and the narrow dispersion could actually be a benefit depending on the desired use of the speakers if wanting a more focused dispersion pattern to keep excessive energy off the walls.
If you are looking for an 8 ohm design this one is likely the most well rounded of the bunch with a very smooth frequency response resulting in excellent midrange and bass definition.
Crossover Schematic:
Port Tube:
Port tube was used at full length for this woofer, but could also be cut a couple inches shorter to give a slightly flatter the bass response if desired at the cost of extension.
On this crossover the C1, cap on the high frequencies should be a polypropylene style cap as well as C4 and C6 for the optional distortion trap but all the others in the crossover are designed to use non-polarized electrolytic. The ESR in the NPE capacitors adds damping in the low pass circuits which proves beneficial to the shaping of the frequency response without the need for additional resistors on the C2/C3 parallel legs. C2 and C3 can be changed to poly but it will result in some additional output in the midrange near the crossover.
Resistors are standard 10w wire-wound.
L1 inductor should be a 20 gauge air core, L2 and L4 should be 18 gauge I-core, L3 & L5 should be a 18 gauge air core.
Crossover BOM with links to suitable parts.
C1 - 3.3uF Poly
C2 - 15uF NPE
C3 - 22uF NPE
C4 - 1.0uF Poly (Alternate)
C5 - 22uF NPE
C6* - 4.3uF Poly
L2 - 2.0mH 18awg I-Core
L3 - 0.80mH 18awg Air Core (Alternate)
L4 - 3.0mH 18awg I-Core
L5* - 0.15mH 18awg Air Core
R1 - 10 Ohm 10w
R2 - 0.82 Ohm 10w
R3 - 12 Ohm 10w
*C6/L5 required only if adding the optional distortion trap filter.
Loudspeaker Drivers / Horn:
Note: This page contains affiliate links which if used allows me to earn a small commission if those products are purchased at no additional cost to you. All of the drivers and parts for this design were purchased, nothing was provided by the affiliated retailers. Any commission earned just helps offset the cost of the build and allows me to continue to design and publish more free DIY speakers like this one.
Optional Distortion Trap Filter
Implementation of this filter was inspired by an article posted by Purifi Audio, which describes using a series notch filter which provides increased impedance in the path of the harmonic distortion components generated by modulation of the inductance of the voice coil at the frequency of the cone breakup peak. The link to the article they posted appears to be a dead now but a working link can still be found using the Wayback Machine (Link here).
The Dayton DA215-8 with its aluminum cone gives a very smooth frequency response but the metallic cone does present a strong resonance or breakup peak near 6kHz. The fundamental of this peak is well suppressed by the standard lowpass in the crossover but you still see spikes in the harmonic distortion components that land on and are amplified by the resonance. Since the frequency of the peak is much higher then the crossover frequency an optional series notch can be implemented with minimal impact on the existing lowpass filter and results in a reduction in severity of the harmonic distortion spikes that land on the resonance peak.
The filter for this design itself is a simple LC filter placed in series with the driver on the output of the crossover. A 0.15mH 18awg air core and 4.3uF capacitor are placed in parallel and provide the correct trap or tuning frequency, this is shown in the crossover schematic. I experimented with adding a resistor for an RLC filter but for this design the resistor added no benefit to the function of the filter.
The following measurements are taken semi-nearfield to the woofer with the normal lowpass filter in the crossover applied, first without then with the optional distortion trap filter. While it does not eliminate the spikes in distortion you should notice the reduction in 3rd, 5th and 7th order harmonics. Additional harmonic distortion measurements of the full speaker can be seen with and without the optional distortion filter in place in the full set of measurements further down on this page.
Full measurements for the Flex-8 Dayton DA215-8 Variant
The following measurements were performed on my 10' tall outdoor turntable, measurements taken at ~5.66v / 2M on the tweeter axis which provides the same approximate SPL as 2.83v / 1M (Nominal 1W/1M).
The measurements were gated at 14ms and blended to diffraction adjusted nearfield woofer response below ~300hz.
No smoothing applied to the frequency response measurements.
On Axis Response
CTA-2034 Style Spin
Estimated In-Room Response
Flex-8 (DA215-8) Harmonic Distortion at 85, 95, 100, 104 and 105dB/1m with and without optional Distortion Trap Filter.
104dB sweep included as there was some amp clipping at the 105dB sweep level.
Note: minor differences in distortion may been seen in lower frequencies between the two sets of distortion measurements (with and without distortion filter) due to slight change in position of the setup as the two sets of measurements were not taken back to back.
Compression at 85/95/100/103/104/105dB normalized against 75dB:
Flex-8 (DA215-8) Impedance: