The Subwoofer DIY Page - Projects
The Enigma Subwoofer
07 March 2020

After a few months of operation in my brother's bar as a "loaner" subwoofer, the INF10 driver in the Enigma was damaged beyond repair (the signs suggest that a power "thump" stuck the cone all the in, and the coil eventually burnt out).  So, it was time to find a replacement driver. A search that proved to be quite difficult, as not many drivers had the same or even similar combination of parameters.

I finally came across two drivers that seemed to fit the bill - the Dayton DA270 and the Dayton DSA270.  The specs of both of these drivers are pretty similar, but their rated Xmax figures (6.1 mm and 5.5 mm respectively) were a bit lower than the INF10's rated Xmax of 8mm.  I was concerned that peak output would be noticeably affected, but that turned out not to be the case.  I actually bought both drivers but chose to install the DA270, keeping the DSA270 as a spare.

As the rebuilt subwoofer was destined for permanent use in the bar after it was repaired, I retuned it by shortening the vents a little more. I also added a DIY two stage "protection" circuit that used two PTC devices and a 3 ohm resistor. I designed the circuit so that, if the first PTC device was tripped, the second PTC device and the 3 ohm resistor would be put in series with the driver, and if the second one was tripped, the power fed to the driver would be significantly reduced.  The new subwoofer has survived over six months of almost continuous service at the bar, so it looks like the combination of the new driver and the protection circuit are working as expected.


Another thing I did was put together a better simulation of the subwoofer's performance in Hornresp, using some features that were fairly recently added to the program.  Instead of simulating it using Hornresp's "BP4"feature, I simulated it using the "OD" feature and segmented sections representing the vented section of the enclosure.  This allowed me to not only model the impact of the vent on the volume of the vented section, but to also model **where** it was having the most impact.  It also allowed me to model the impact of partly stuffing the vented section. 

The frequency response and impedance curve predicted by the new model also turned out to be very similar to the actual response of the rebuilt subwoofer - close enough to convince me that this is probably the best way to model offset-driver 4th order bandpass designs in Hornresp at this time.

I also took the opportunity to perform some linearity and distortion testing on the rebuilt subwoofer, and I'm quite satisfied with the results.  THD breached 10% at the lower end of the passband with the equivalent of about 70 Watts of input, almost exactly where Hornresp predicted it would happen, and higher up in the passband the THD dropped significantly.  This was going to be a very clean-sounding subwoofer, all the way up to its limits.

The image below is the Enigma V2's measured frequency response with fed pink noise through an active 24dB/octave LR filter centered at 200 Hz, at a level 10dB below the level at which it reaches 10% THD in its passband. Out of band noise is basically non-existent at this level.


Overall, I'm very happy with how this rebuild of a rebuild turned out, and intend to use this design approach for offset driver 4th order bandpass alignments going forward as, not only does it accurately predict the frequency response of the alignment in the passband, it also allows me to predict what the out of band noise is likely to look like, allowing me the opportunity to "tweak" the simulation to reduce the impact of this out of band noise.