This project was a collaboration between Kevin K. and myself. He did all of the visual design, construction and finishing of the box. I supplied him with what I wanted for driver layouts, volume, and did the crossover design. All pictures of Kevin’s build progress, as well as a few videos of CNC cutting the baffles can be found at the following Drop Box link: https://www.dropbox.com/sh/gyobv92es...iFREKcR7a?dl=0
The Silverbacks name comes from the color scheme, the fact that this is a fairly large (not quite monkey coffin size) 3-way bookshelf, and it’s got a big beastly woofer for the bottom. This led Kevin to liken it to a Silverback gorilla.
The project originated with the idea to try and squeeze a 3-way in under the Iron Driver $225 drivers and crossover parts budget. I figured most people would be going with a 2-way and wanted to do something a bit different. We did a lot of searching for woofers and finally settled on the MCM 8” subwoofer (http://www.mcmelectronics.com/produc...-2421-/55-2421) with its fairly small box requirements and healthy xmax. We actually tested out the Dayton DVC 8” woofer as well, since it would be able to handle a higher crossover point, but it required a much large enclosure and hit xmax with a lot less input power.
The biggest challenge with this was that it really required a pretty low crossover point. I was originally thinking of using the TC9 mid, which would have made budget more flexible, but there were two reasons I went a different route. The first was that I’ve used it a number of times already and wanted to try something different. The second, and more limiting factor, was that it really likes a crossover point of somewhere around 400 Hz. The MCM subwoofer, in my opinion, needed to be crossed closer to 200 Hz, so we went with a mid that had a little more xmax and usable extension on the low end. We wound up with the SB Acoustics SB12PAC25-4 4” woofer (https://www.madisoundspeakerstore.co...m-cone-woofer/), which I hadn’t seen used yet, but published graphs looked pretty good and SB tends to make quality products.
The MCM drivers measured dead on spec for T/S parameters, so we ended up going with a 16 liters after displacement and port for the woofer chamber and the box was tuned to 30 Hz with a 10.5” long by 1.5” diameter port that was flared on one end. I intentionally tuned the box a little low. This helped reduce excursion on the woofer in the normal music range and gave a bit of a shallower roll-off on the bottom end. With this tuning and the crossover, the f3 came out to 41 Hz and f6 was 32 Hz. The mid chamber came in at 4L. I didn’t actually measure the T/S parameters of the mid, only the impedance in box and the frequency response, which matched up to SB’s published info very well once you account for baffle step and diffraction. I went with the mid above mainly to keep the tweeter height closer to ear level once on the 22” stands.
The box was built out of ¾” MDF (except for the baffle) and cut on a CNC machine, which is how Kevin got that super cool looking curvature to the front baffle. The cabinet was finished in Italian silver ash veneer, spayed with a few layers of Sealcoat before being sanded smooth, and then finished in a semi-gloss top coat. I had Kevin line the inside of the bass chamber with peel-and-stick vinyl floor tile for some added mass damping and then he added eggcrate foam over the top of that. The mid chamber was stuffed with denim insulation.
As for the crossover, I ended up settling on roughly second order Butterworth at 240 Hz for the woofer LP, 400 Hz second order Bessel on the mid HP, 2200 Hz third order Butterworth on the mid LP, and sixth order Butterworth at 2800 Hz on the tweeter HP . I started out close to this and then played with the voicing a bit until I got something I was pretty happy with. I tried crossing the tweeter lower at around 2 kHz, but it sounded a little grainy. It also increased the parts count, and might have pushed me out of the budget.
Un-spliced measurement (after revision) taken at 30" with 1/48th octave smoothing:
The impedance dips pretty low, around 2.4 Ohms at the lowest, but the phase angle is pretty benign. I ran these for a couple hours in testing and my amp stayed very cool to the touch. I've had other speakers that have a higher overall impedance curve, but have gotten my amp a decent amount warmer. If your amp is 4 Ohm capable, it should be no issue. I'll be trying these on my Marantz HT receiver as well to see how they do with that.
I ended up accidentally using a 3 Ohm resistor on the tweeter circuit instead of the 4 Ohm I meant to use when I assembled the speakers to be shipped to Roman. I was in the front row at the event and found these to be just a bit to bright in the upper midrange and treble. I will be adjusting the crossover padding on the mid and tweeter slightly to account for this and will post the crossover schematic once that is done.
CROSSOVER DIAGRAM: to come shortly
All crossover parts were purchased from ERSE to ensure we could come in under budget. I ended up using all NPE caps on the woofer and mid and the cheapest inductors they had everywhere else, but was able to squeeze in some cheap 250V poly caps on the tweeter circuit. Feel free to upgrade everywhere with one exception. The woofer circuit used a high DCR 2.5 mH 20 awg air core inductor from ERSE. This was not only done for cost savings, but the additional DCR helped to control the peaking on the woofer with the low crossover point I used. I accounted for this in all my box sims as well as the crossover sims. If you change out to a lower DCR inductor, it will cause up to 4 dB of peaking centered at 80 Hz, so please be aware of this.
Overall, this was a tough design to work with but it seems to have paid off. I can’t tell you how excited I was to see our speaker sitting there when they pulled the curtain at the end of the event. What I think really set this design apart was the ability to play very cleanly at loud levels. In the head-to-head, the big drum track with heavy bass guitar really showcased this. It was super clean and the bass stayed really tight. I also got a bunch of comments about how nice the vocals sounded on these.
Some extra info:
You can build this on a standard baffle with a ¾” roundover. The testing was done in a standard cabinet with a ¾” roundover and final measurements in the facetted baffle matched fairly close. Make sure you keep driver spacing and general layout.
I’ve also modeled with the mid and tweeter position reversed, and thanks to the low crossover point on the woofer, you should be able to swap positions if you want the tweeter on top (as long as spacing between the mid and tweeter are the same) without a change in response. It will just flip the vertical null. I have only modeled this and not measured, so there will probably be some minor diffraction differences due to the tweeter being farther away from the woofer cone.