crazyuncleglenn.hackleman.net 

    These are a pair of cherry veneered scratch-built grill-less speakers I completed over the course of three months.


    I started this project to compete with the $1400 pair of klipsch speakers that tweeter was quoting my father-in-law for a home theatre system installation. I ended up doing the entire install for him including some more budget in-wall surrounds and center channel. Overall, tweeter was asking $13-15k, and we got away with less than $5k total, most of that being the 1080p LCD! Acoustically they are comparable to the units tweeter wanted to sell and cost only $300 in parts. They also look a hell of a lot better. Of course with labor they are well over $3k, but gifts are gifts. He specified size and he wanted a rich, warm finish to complement the other wood furniture in their house. Custom speakers, here we go!

    The boxes for the speakers start life as 2'x 4' sheets of 3/4" medium density fiberboard. This stuff is easy to work with, dimensionally stable, and has good damping characteristics. Since this speakers are to be veneered, the prefinished surface on the MDF really saves a lot of time on something no one will see anyways. I cut up the sheets on my cheapo ($99) Ryobi table saw. For glue I tried polyurethane glue for the first time and found it worked well. It expands as it dries so ensures a perfect seal. Unfortunately it costs more and will stain your skin brown if you don't wear gloves. One of its other characteristics is that it is waterproof, unlike traditional yellow woodworkers glue. However, that doesn't mean anything in this application! The boxes are clamped and reinforced with 2" drywall screws that are predrilled and countersunk.

DSC00892 DSC00893 DSC00899 DSC00896

Once the glue dried I measured and marked the driver cutouts. These were for the minor mounting diameter, as I planned to route the recesses for the drivers so they mount flush with the front baffle. This is required for aesthetics as speaker grills are not part of the design. This improves dispersion and reduces distortions since the surface is perfectly flat, there is no grill frame to get in the way of the sound. Since I did not feel like spending $45 for a router bit that was the right dimensions, I used one I already had and fabricated a new brass bearing for the router bit on my lathe. This worked great! As yu can see it is much smaller diameter than the original bearing, this increased the bite of the bit by a lot.

DSC00900 DSC00901 DSC00906 DSC00907

After routing was complete I tested to see if the drivers actually fit, I went ahead and cut the holes for the ports. The ports themselves are PVC pipe with and elbow, and glued into place with polyurethane glue. The interesting thing about desinging ported enclosures using thiele-small parameters is that as you make the ports larger, you have to increase there length to maintain the same frequency cutoff. This puts an upper limit on how large your ports can be, as they change the box volumne as you make them longer! While this begs the question, why not make them small? Well, you want to make them as large as feasible to lower your vent mach, which with small ports makes your speakers whistle as they play - not good. So like everyhting in engineering design, there are tradeoffs. As you play with the equations it makes you appreciate the fact that great sounding speakers cannot be small! Sorry bose, but your wave radio sucks :)

DSC00916 DSC00917 DSC00918 DSC00921

After the ports were dry I used began the laborious finishing process. This consists of sanding out any inaccuracies in the MDF cuts and filling the screws in with drywall spackle. I use spackle because it dries quickly and sands easily. I have never found wood putty to be useful - typically it just crumbles when you sand it, and is fairly expensive. I ran 3 fill/sand coats on each box to level everything out and get a good glueable surface. Next I sprayed the driver recesses flat black since the edge will be visible and we want it to be dark and not stand out. When I was coming up with tools for veneering I noticed that both lowes and home depot had no j-rollers or "squeegees". So I bent up some mild steel and used some scrap oak dowel to mke my own j-roller. Don't laugh, its actually very strong and can apply plenty of pressure. Later when I found a source for a store-bought tool I never did buy it since it felt flimsy by comparison.

DSC00923 DSC00925 DSC00924 DSC00926

To apply the veneer I decided to use contact cement, another first for me. I used the full strengh, full flammability, full brain-destroying kind - I had read that the less-flammable variants were prone to peeling or lifting - definitely not cool! This stuff is pretty interesting, you basically paint both surfaces with it, let it dry to the touch, and then put them together. As soon, I mean, the very absolute instant that they touch, you will never, ever get them apart again. Which is both good and bad. Good becuase it will stick forever. Bad, becuase you only have one chance. So there are various methods of making this bond happen straight and bubble free. I first tried the wax paper method - this didn't work well for me because the contact cement is just slightly tacky, and when you try to remove the wax paper it distorts the veneer position. So instead I used the babob skewer method, where you use a few bamboo skewers to keep them seperated, and then draw them out one by one, and attach the veneer a little at a time. I used my j roller to prevent bubbles and a large oak dowel like a rolling pin to apply tremendous pressure. There is this little cheesy plastic tool that holds razor blades for trimming the veneer after you glue it. It actually worked well! After the trimming was done I began the punch-out process for the driver cutouts.

DSC00930 DSC00931 DSC00927 DSC00933

The drivers are located because they are soft spots through the veneer - I used a box-cutter to score the major portions and there is a method I found that works with the thumb to crack the veneer on the edge of the driver process without splintering. Then I can carve it out with the box cutter very accurately. I punched out the rear port holes at this time as well. They are beginning to actually look like speakers! The cherry is a little blonde at this point and with just finish it wouldn't be very dark. Its possible to use the sun to darken cherry but this is usually very time consuming (days upon days), so I opted for staining them. I originally chose a walnut finish which turned out to be mottled, light, and way too "cool". So I decided to use it as a base stain (what choice did I have) and find a warmer stain to topcoat with. I went ahead and stained and urethaned the thing twice after the first stain dried, and started looking for fixes.

DSC00932 DSC00936 DSC00937 DSC00945

To test for different possibilites I went down to HD and bought a couple of cans of stain/poly mix called "polyshades". This product had very polarized reviews online, either people like liked it or absolutely hated it. I went ahead and ignored all the reviews and bought it anyways since I needed to test a bunch of different combos. I made up some swatches with the "mistake" on them and started experimenting. Ended up finding combo I liked and applied it to the speakers. The first coat went on well and improved the appeareance. I put on a coat of normal poly after that one dried and then put on a second coat of the polyshades. By then it was looking great and I decided to hand-rub on the final three poly coats in satin clear. All together the speakers have more than 9 coats of stain/poly. It gives the finish a lot of visual depth.

DSC00947 DSC00955 DSC00958 DSC00950

In the last shot you can see where I installed the crossovers. The inside of the box was lined with acoustic dampening foam, also a first for me. I used spray adhesive to attach it the the interior walls. On some previous pics you can see the 50 Ohm L-pad on the back near the bottom port which is wired to the tweeter. This allows one to adjust the teeter level to match the room (tweeters are typically much more efficient than woofers, so its often necessary to be able to attentuate it) whilst still keeping a constant impedance as seen by the crossover. In acoustically "live" rooms the tweeters get turned down (lots of hard surfaces) and in dead rooms (lots of absorbing materials like couches and carpets) they get turned up. Since our house has a fair amount of carpeting they were up fairly high when I tested them but when I put them in my father-in-laws house I had to turn them down. There aren't any shots of it but after wiring up the enclosure I filled it with teased dacron polyester fibers to increase apparent box volume. Again, a first: I always used cheap fiberglass insulation in the past. There isn't any advantage to the dacron aside from the fact that it doesn't itch the hell out of your skin when working with it :)

DSC00953 DSC00959 DSC00964 DSC00962

As you can see the speakers are a WTW (Woofer/Tweeter/Woofer) design using 6.5" unshielded Dayton drivers and a silk dome tweeter. The woofers are 4 Ohm wired in series to give an effective impedance of 8 Ohms. I decided on 8 Ohm since not many run-of-the-mill receivers are capable of driving less than 6 ohms. I should know, I ran into the same problem on my Tower Speakers - originally they were 4 Ohm and then I realized that I coudln't find a reciever that could drive them, and I would have had to spin my own to not break the bank! I had to order new drivers to change the impedance. Although Building 4-Ohm-capable amps is easy, you just double up the final transistor stage and/or use a lower power supply voltage - but it would have doubled project time, since an amp is easily a huge project unto itself.

DSC00969 DSC00971 DSC00977 DSC00976

My father-in-law is very happy with the speakers, and my only regret was that I didn't get to keep them myself. They really turned out great.


copyright 2008 glenn hackleman
home blog email