> Linkwitz Lab LX521 project
Linkwitz Lab LX521 Dipole Loudspeaker Project

The LX521 loudspeaker, designed by the late Siegfried Linkwitz, is a dipole speaker with an acoustic radiation pattern that remains fairly constant with changing frequency. The exceptional Linkwitz Orion was partway there, but this design takes it even further. As a result, the off-axis reflections of sound coming from the walls should closely resemble the on-axis sound. This provides a better spatial rendition of the "acoustic space" in which the music was recorded. In other words, more realism - a fine goal for good loudspeakers.

This speaker will have a hybrid active/passive crossover. Pluto and Orion, also designed by Linkwitz, are completely active loudspeakers. The LX521 differs by being a 4-way speaker (tweeters, upper-midrange, lower midrange, and a pair of 10" woofers) and uses a simple passive crossover between the upper-mid and lower-mid. Active speakers require additional amplifiers for each active driver, and this can get costly. The LX521 requires at least 6 channels of amplification if the two woofers are driven in parallel as one unit. This requires a robust 2-ohm capable amplifier.


I ordered plans and circuit boards for the ASP (analog signal processor) from Linkwitz Labs shortly after the LX521 speaker design was announced in autumn 2012. Even though I was partway through my construction of a set of Orion loudspeakers, I took the leap. Because these replace the Orions, that hurt a little. I am still pondering whether to finish the Orions, or to abandon that project at ~80% complete.

 


Construction Log, Page One - Baffle template, bracket, and woofer box fabrication.

[Skip to page two...]  
[Skip to page three...]  
[Skip to photos of the finished speaker...]  
 
Plans Arrive [November 20, 2012]

<--click any thumbnail to enlarge

 I've ordered the plans and ASP circuit boards from Linkwitz Labs. They arrive in about a week via Priority Mail.

[NOTE: Linkwitz passed away in September 2018. Plans are not available directly anymore, but Madisound and http://www.magiclx521.com have taken over distribution of them. Also, newer versions of the LX521 use DSP for the active crossover instead of ASP analog circuit boards.]

 
Planning Baffle Fabrication

The photo shows the baffle layout being sketched on 3/4" MDF. I plan to use this as a master template for routing subsequent copies. This piece won't be used on the loudspeaker itself.

The baffle is an unusual shape. The width of the baffle changes with each driver to make it small acoustically. Because it is an open baffle loudspeaker, the fabrication of the baffle should be easier than with a box-type loudspeaker. I'm still exploring what materials to use (black-painted MDF or plywood, hardwood, etc.). I've also ordered some unusual size Forstner bits to cut the tweeter holes and the upper midrange speaker hole.

 

The Baffle Bracket

In the plans, the baffle is held to the structure with a bracket that I'm making from a combination of 1-1/8" MDF and 18mm Baltic birch plywood. I used my table saw sled to rough cut the pieces, and also used it to cut the angles on the brace uprights. You can see the hold-down stick in the first photo to keep my hands away from the spinning blade. It has a rubber facing on the side that contacts the work piece.

I've already taken some of the parts to the router table to put on a small 1/8" radius on the edges of the MDF "foot" and the plywood sides. That should help prevent skinning my knuckles on sharp edges when I'm working with the passive crossover to be located between the uprights. [Edit: Later on in this project, I re-routed a much larger chamfer instead of the small 1/8" roundover in order to match the appearance of the rest of the bridge.]

In a slight deviation from plans, I made the brace "foot" into a trapezoidal shape to echo the angles designed into the baffle. This brace assembly will be painted black like SL's, probably using  Lamp Black water-based General Finishes Milk Paint. It will be sprayed or rolled-on after a good shellac coat is done to seal the wood/MDF.

 

Work continues on the Pattern for the Baffles [November 30, 2012]

I ordered some Forstner bits from Woodcraft to cut the upper mid and tweeter holes. I could have done it with the router, but I dislike the process - especially when the weather is marginal. If I can't use the router table, I route MDF outdoors because of the dust. With poor weather outdoors, the Forstners will make life easier.

One Forstner is 1-7/8" in diameter (tweeter) and the other is 3-1/8" in diameter. The 3-1/8" diameter is slightly too small for the driver (by about the thickness of a business card), but I used a sanding drum on my drill press to enlarge the hole in the pattern to the needed dimension. This enlargement took about 10 minutes with pauses for measurements along the way.

A 3+ inch Forstner is about the upper limit that my drill press can swing without straining. I took light cuts, and repeatedly pulled the cutter from the hole to clear chips and to cool the tool.

I was very pleased with how clean the Forstner bits worked. Not much dust went airborne - unlike cutting the holes with a router.

After I cut the holes, I trimmed the overall length of the baffle on my table saw. The table saw's sled made this very easy.

To cut the outside shape of this irregular baffle, I took it to my Sears Craftsman band saw. I've improved this hand-me-down band saw over the years so that it cuts straight and smooth. However I quickly discovered that the narrow throat of my band saw is too small for the baffle when it was angled for some cuts. I did as much as I could with the baffle facing up, then carefully marked the problem cut on the back and cut the rest that way. If turning it upside down hadn't worked on the band saw, I was prepared to use the jig saw for any remaining problem cuts. After all, this has to be close to the line, and not exactly on it. The excess will be trimmed to dimension in a subsequent step.

I am able to cut fairly close to the mark with the band saw, but I'm leaving a little extra material to be trimmed off either with sanding (if there isn't much material left to remove), or on the router table using a pattern bit after I tack down some scrap pieces of MDF to use as a temporary guide for the cuts. I eventually went with the router table approach because it is so straightforward and fast.

This pattern must be well-made. The resulting baffle copies will echo any flaws it contains. Outside edges must be smooth and straight. The router bit will ruthlessly expose any sloppiness or irregularities.

 

 

Continuing Work on the Brackets [December 2, 2012]

With the baffle pattern in its current state, I was able to mark out the mounting screw hole pattern on the baffle. I drilled 1/8" pilot holes in the baffle, and used transfer punches to mark the front of the bracket.

I plan to use 1/4"x20 machine screws to hold the baffle to the bracket so that I can easily interchange baffles. I may make painted MDf baffles to start with, but later switch to hardwood as time and materials become available. I like being able to change my mind.

I drilled 1/4" diameter through holes on the bracket front, and used a 3/8" diameter counterbore with a 1/4" diameter pilot to have near perfect concentricity. The changeable-pilot counterbore tool was purchased at McMaster-Carr a while back, and it works splendidly. It has interchangeable pilots to align with just about any size pre-drilled hole.

After creating the 1/2" deep x 3/8" diameter counterbores, I inserted 1/4"x20 threaded brass inserts into the rear side of the baffle bracket. That leaves about a 1/8" inch "land" of material on the front to eliminate the chance of pullout.

The threaded inserts are installed using a T-wrench that I purchased from Woodcraft years ago. I have to use a washer with it to prevent it from wedging into the back side of the threaded inserts, but that's easy. While installing the inserts, I keep a small square handy to judge how straight the insert is going in.

I may use screws that are longer than the thickness of the baffle bracket and baffle combined, so I bored 5/16" diameter holes into the edge of the bracket sides. This will permit a fairly long screw to pass completely through the threaded brass insert and still fit without bottoming.

I added a couple of small holes on the vertical centerline of the bracket in case I wanted to mount a hardwood baffle without any mounting screws visible from the front of the speaker. They are just "what if" holes just in case. It is easier to drill and countersink them now than after the bracket is assembled.

Once everything was fitted nicely, I put a thin layer of glue on the mating surfaces, and tightened the 1-1/4" deck screws that I used to clamp it together.

Things are nice and square, with just a tiny bit of misalignment. It's nothing that a couple swipes of a hand plane or a few minutes with sandpaper can't fix quickly. It's funny how something can dry-fit perfectly, but there's always a slight bit of misalignment when glue is applied!

 

Baffle Material Thickness Issues

On the Orion/Pluto/LX521 owner's forum I posted a couple of pictures of two kinds of plywood being measured with calipers. I'm not concerned about increases in strength that come with thickness

My concern is having enough cavity depth to hold the tweeter. Thin plywood, no matter how well it is made, will not allow the tweeter to seat. My sample of Baltic birch plywood is only 0.67 or 0.68 inch thick. The hardwood plate attached to the front and to the back of the baffle is 0.13" thick. Added together, you have a cavity for the tweeter body that is 0.80 inch deep.

However the tweeter is 0.85 inch deep. It won't fit. The feature on the back of the tweeter interferes with the rear tweeter plate (Part C) when it is mounted. (Note that it's not shown in the illustration to the left.)

One could use a Forstner bit and drill about halfway through the thin hardboard piece on the rear to provide a space for the tweeter protrusion. You'd have to do the front plate too because there are two tweeters. While this is possible, it slows down construction.

If baffle material that is a minimum of 0.72 inch is used (typical for some USA-specification "3/4 inch" plywood), then the tweeter hole/cavity inside will be just deep enough to contain the tweeter. I'd still be worried about buzzes and rattles if the tweeter fits with zero clearance. If I were using 0.72 plywood, I'd put a piece of very thin LD polyethylene foam in the cavity first to absorb any contact.

 

Perfecting the Pattern Edges

Because the baffle pattern was first cut on a band saw, the edges were not as smooth and straight as you get from a table saw. However a table saw can't cut the angles needed for the baffle. The edges were left a little proud of the pencil layout lines because I knew that I had to trim to the final shape somehow.

To trim the edges right to the pencil mark and to make them very smooth, I cut some MDF scraps into various wedge shapes for use. I used double-stick tape to adhere them to the baffle pattern exactly at the pencil lines.

I took the stack of parts to the router table to do a light trim pass to match the straight edges of the scraps. There wasn't much left to remove because the band saw cut was very close to the layout lines. I estimate that I had to remove only 1/32" of material in the trim operation.

I cut the left edge and the right edge is two separate passes. I simply reused the scraps from the left side on the right side for the second pass. I merely flipped them over and used fresh double-stick tape.

Easy work. Faster than sandpaper.

I used spot putty on only one edge where the band saw went very slightly into the line. When the putty dried, I sanded it flush. Then I gave the MDF pattern a coat of shellac to "case harden" it for its eventual pattern duty.

In the last photo, you will spot the 1/8" hole drilled between the two tweeter bores. It will hold one of the 1/8" metal dowel pins used to hold the pattern and the work piece in registration. I also drilled 1/8" holes to mark the centers of the mounting screw holes, and a second dowel pin will be inserted through one of them for registration. The two 1/8" steel dowel pins will hold the pattern securely to the stock for flush trimming around the edges.

The holes are placed strategically in places that are either hidden later (i.e., under the tweeter sub-baffle), or are used/enlarged later (one of the mounting screw holes). I plan to use 1/4"x20 threaded fasteners through the baffles into the brackets to assemble them.

 

Back to Brackets

I drilled through holes in the base of the bracket assembly for #6 x 2" deck screws to hold the base and the upper assembly together. The base is 1-1/4" MDF from stair tread purchased from Lowe's.

I needed to drill perpendicular pilot holes to receive the screws in the legs of the upper bracket assembly, but the shape of the top part made that difficult. It wouldn't stand square by itself. I solved the problem by using one bracket to support the other upside down bracket. Then I drilled the 3/32" pilot holes square to the material.

In the router table, I routed a 45 degree chamfer around the base of the bracket assembly for appearance. Because most of the base is behind and fairly far from most of the drivers, I presumed that there wouldn't be any significant sonic change.

I also made the base for the bracket a slight trapezoid shape to echo the angular baffle appearance. Angular is in!

 

Woofer Box Joinery

I am deviating from the plans slightly in an attempt to improve upon the joinery of wooden parts for the woofer box. The items shown in red (Parts A, D, & E in the top picture, and Part C in the bottom picture) have been slightly changed to meet these goals. This will allow me to use biscuits to join the baffle to the box in some areas, and I will move a butt joint to the top surface for visual reasons. I also plan to miter the angled woofer baffle where it meets the front of the box.

I completely redrew the plans because I'll use Baltic birch plywood which measures 0.68" thick. This is somewhat thinner than the nominal 3/4" material used in SL's plans. To get a tight fit everywhere, I had to determine new sizes of the parts for the cabinet. Redrawing to scale allowed me to do this.

For non-owners of plans, I'm sorry I can't present more detail or show an overall picture here. If I did so, I might be inadvertently releasing intellectual property. For Linkwitz plan owners, these pictures combined with your existing plans on page 11 should be enough to understand my approach.

 
Baffle Material [December 15, 2012]

While I was shopping at the local farmer's market, I saw a vendor selling wooden cutting boards there. It is someone with whom I have worked on some projects at the university, and his woodworking looked solid. I asked about getting some cherry boards glued up into a panel to serve as baffle material. His similar sized kitchen cutting boards were $45 each, and he said that it would be the same for an all-cherry panel. He also stated that he could plane the thickness close to what SL has specified in the plans.

I sent him an email about getting the work done, and supplied this sketch (sorry about the lack of supplied dimensions - intellectual property, you know). Cherry should look nice, and I have lots of cherry boards in the garage for other parts of the build (bridge) if I choose to go that direction.

 

Tweeter sub-baffle fabrication [December 17, 2012]

I attempted to make do with a hole saw to cut the radius at the bottom of the tweeter sub-baffle. However hole saws are terribly inaccurate. This hole saw, almost new, probably has a whopping 1/16" runout. In addition, the center hole wants to wander, the saw chatters when it doesn't like the feed rate. It was a miserable experience. The cut edges were chewed-up and not clean. I decided against doing any more work on this project with a hole saw. I discarded the parts affected by the hole saw, and began with a new approach.

A router, a 1/4" spiral downcut bit, and a hole cutting jig (a Jasper jig, in my case) was a better way to go. However, the tweeter sub-baffle is small and hard to hold. I decided to make a jig to hold the work. With a little thought, the jig performed a couple of other duties too, as you'll see.

The jig is made from a scrap piece of 3/4" MDF, and pieces of Masonite the same thickness as the tweeter sub-baffle. The work piece has to sit flush on top of the jig so that the router base passes over it smoothly. To allow a little clearance between work and the jig sides, I used a piece of heavy paper as a shim. This provided just a little clearance so that I could remove the work after it had been routed. It also accommodated the slight variations in the width of each sub-baffle blank.

I also carefully marked and drilled three 1/8" holes in the jig that locate the center of the two tweeter locations (the low tweeter in front, the high tweeter in back of the baffle and the center of the arc to be routed on the bottom. The two tweeter center holes are used to guide a transfer punch for marking the hole center, and the punch also helped remove the work from the jig. I used double stick tape to keep the work in position, and it is then difficult to remove from the jig. Using a transfer punch, a light tap through the jig onto the back of the sub-baffle  loosened it, and simultaneously marked the center of the tweeter hole. Using the punched mark for alignment, I bored the tweeter holes afterward using one of the Forstner bits shown earlier in this build log.

The 1/8" hole located at the center of the arc did two things - it let me flip the jig over and drill the sub-baffle blank from the rear for perfect alignment. It then held the pin in location for the router jig.

The jig didn't take long to put together, and it did a satisfyingly accurate job with all aspects of fabricating the tweeter sub-baffle. The punch marks left by the transfer punch were extremely accurate. In the second-to-the-last photo in this group, you can see how the ruler line (the 6" graduation) is exactly at the punch mark when the left and right edges were on their respective marks.

I used my table saw to cut the angled sides of the front tweeter sub-baffles, and trim the "legs" of the rear tweeter sub-baffles (per the photos on SL's site, not exactly to plan). All that's left is to mark, drill, and countersink the mounting holes for black #4x3/8" sheet-metal screws. And paint it, of course.

I'll keep this jig in case I want to change the  material for the tweeter sub-baffle. While it's currently ordinary Masonite, some thin cherry or other attractive hardwood might look nice. For now the Masonite, painted, will work fine.

I have to keep momentum going on this project. There are several other commitments looming that will occupy weekends and all spare time.

 

The Woofer Box [December 19 & 20, 2012]

I spent most of today cutting down some 3/4" Baltic birch plywood for the woofer box. I began outside on a very chilly day cutting some large sheets to more manageable sizes. I used sawhorses and a circular saw on a track. After doing that, I took the small pieces to the table saw for another pass at rough cutting. My goal was to get clean edges and square corners. After that, it's easy to cut the pieces to the final size fairly accurately.

I use a crosscut sled for both crosscutting and also for minor ripping duties. It works well both ways.

When crosscutting, I try to use blue painter's tape on the cut line to reduce chipping of the plywood. I'm using a 40-tooth general purpose blade, and it needs a little help when crosscutting plywood. I dislike splintered edges.

When rough cutting outside on the sawhorses using a circular saw, I knew that my crosscut sled would handle only 24-1/2" pieces between the front and back. You can see from the first photo that I just made it. The width of the piece just fits into the sled.

While I have a Rockler track and a flip stop mounted on my sled, it can't quite reach much beyond 22" without exiting the track. I resorted to the old-fashioned way of c-clamping a block of wood to the fence. It worked fine as you can see from the third photo. I have a detachable table saw stop for crosscutting longer boards, but I didn't' have to use it today.

With each speaker build that I attempt, I try to achieve even more precision than the time before from my tools. I wanted to rough-cut cut a board to 24", and you can see that I did OK in the photo with the tape measure.

Even better is the parallelism that I get from edge to edge. Take a look at the measurements made with a caliper at the top and bottom width of one of the boards. It's parallel to 1/1000". That's satisfying! [Yeah, I'm bragging a bit! It took a while to get to this point.] This was a rough-cut to square up all sides with clean edges. Now if I can keep this precision for all the final cuts!

I now have all pieces F cut to final size, and the top panels are the appropriate length. There's much more work left on these.

On my Orions, I spent a great deal of time hand planing the plywood boards smooth and flat before cutting them to final size. The plywood had a nice, nearly glass-smooth finish after planing. I'm not doing that here, and I somehow feel like I'm omitting a step that will ease painting later. We'll see.

Update December 20, 2012
All table saw work is done. I mitered a couple pieces to either improve appearance, or to improve the joint between parts B and E. I don't have a crosscut sled for doing miters, but an Incra miter gauge that I had careful adjusted for squareness did an admirable job. This gauge is used only on the right side of the blade, and sees little use. After mitering one end, the pieces were cut to the final dimensions on the other end.

All the cuts look good. If I can't get a square box out of this, I need a different hobby.

 

Biscuit Cutting [December 21, 2012]

It was cold and it snowed today, so I didn't go outside to route driver holes. Instead, I took a little time to cut slots for biscuits, and to check the fit of the newly cut parts.

I like to use a biscuit cutter because of how it keys the pieces of a project together. Dry fits are much easier to do because the tight fit of biscuits in the slots holds them together, and gluing later on is much easier.

I have thought about the sequence of assembly that will keep things as square as possible. I'll put the front "skirt" on the top panel first, gluing them together while both parts are biscuited together with the side panels. That will register the parts so that they will fit the side panels exactly later on. I'll remove the side panels when the glue dries and continue with other parts.

I use a biscuit cutting jig to ease work cutting the biscuit slots. Over time it has evolved to solve problems that arise - i.e., warped boards. I have a pivoting hold-down arm that presses the work flat against the jig surface if the work is warped slightly. I have been able to apply so much pressure that the MDF base bowed, mis-registering the slot, so I glued a piece of 1/2" Baltic birch plywood under the work surface to stiffen it.

To cut face slots in vertically oriented boards, I have a push stick for safety. It keeps my hands far from the spinning cutter.

One of the remaining problems I have with the biscuit cutter is that its carriage has a little up/down slop in it. I'm still trying to figure out "windage" to give me perfectly flush joints each time. In one of the pictures, I did it. The metal square is flush along the joint. In other joints, I applied corrections the wrong way, and there are slight mismatches between mating parts. Because it was a bone-headed error, I wrote down the corrections on a reminder sticker ("Press down on handle when cutting face slots")  and added that on the jig. I don't use it enough to remember the steps months later, so I have to write things down.

However even the larger mismatches are only 0.020" from flush, so it isn't a big deal. A couple swipes with the hand plane will level things.

Curious to see if my modifications worked, I dry fitted even the baffle parts even though they don't have biscuit slots cut. If I do cut slots in them and the side panels, it will be a tricky deal, and I will have to remove the cutter from the jig for some cuts. I have a plan though.

I was very satisfied with how the parts fit together at this point. The miter cuts on two pieces worked perfectly per the drawing I made earlier. I deviated slightly from SL's plans here, but I believe that it will prove to be a better woodworking joint in the end, and will improve appearance a smidge.

 

Routing Holes [December 23, 2012]

The weather warmed enough today to go outside for the messy task of routing driver holes. Sawdust flies everywhere. Outdoors, I merely use a leaf blower after I'm done to clean up. Indoors, I'd still be vacuuming crevices for hidden sawdust instead of writing this.

Of course I'm a prisoner to weather. Lately we've had rain, strong winds, then snow. And cold temperatures that make work outdoors unpleasant.

I took advantage of today's weather. It began as a chilly 31 F morning, but by the time I was done, it warmed to 42 F. Snow was melting.

I used the drill press to drill 1/8" pilot holes for my Jasper Circle jig. The jig rotates around the pin in the wood, cutting neat circles and arcs. My first two passes were made with a Bosch 1/4" downcut spiral bit. I've found that they leave a cleaner edge to the cut without splintering. Then I switch to a Bosch 1/4" upcut spiral bit to finish the work, cutting through to the other side. The upcut bit cleans out waste better in deep grooves.

I also noticed something malfunctioning with my router. The depth-stop fine adjustment was turning rapidly all by itself when the motor was turned on. I'd estimate that it was completing a revolution in 10 seconds. That's a problem with stepped routing like I was doing today when each revolution changes the depth by 1/32". To bypass the problem, I turned the depth stop all the way to its limit so it could no longer turn by itself. I used the coarse adjuster to get me close. Close is good enough for this work, but I'll have to remember to look into fixing that later. [Update - there's a small 3/16" i.d. x 5/16" o.d. o-ring in the adjuster to provide friction and to seal the threads from debris. It must be worn, and it's inexpensive to replace.]

Even though I used the downcut bit first to cut clean edges initially, I still had a couple plywood panels chip at the edges after I switched to the upcut bit. I tried to sand a small bevel on the remaining panels with a sanding sponge, and it seemed to help, but not eliminate, the problem. Thankfully the chipped edges will be hidden by driver flanges or will be rabbeted to a larger diameter in the case of my top panels.

I still need to rough cut some 1/4" hardboard, then route the circular covers for the tops of the woofer boxes. I cut openings in the top of my boxes to allow for access when mounting the heavy woofers. The openings will allow easy use of tools. It will be covered up with the circular covers after drivers are installed. If you are thinking of manhole covers found in streets, you wouldn't be far off. Same concept.

Now, If anyone has a really clever use for the 3/4" thick disks left over from speaker building, just let me know. I'm starting to collect way too many.

 

 

Marking and Drilling Woofer Mount Holes [December 25, 2012]

I spent a little time in the workshop today marking and drilling screw holes for the woofers. The shop was very cold today. I'll have to start wearing a coat in there when I work.

To mark the screw locations, I place a woofer into the routed opening and center it using some marks made with a square and some little punches made with the tip of my calipers. The original pencil centerline is still visible for use. I measure the center to center distance of the actual holes in the woofers, divide by two, and set the calipers to that distance. I reference from the centerline and make a left and right hand scratch using the calipers on either side of the centerline. Then I use the combination square to draw short lines through the caliper scratches. The steps are repeated on the left side, which is 910 degrees opposite the original marks.

These lines are visible through the woofer mounting holes when it is positioned correctly. The woofer sits in an opening that is about 1/16" (for me, at least) larger than the driver itself, so these lines aid in centering the woofer in that opening. It isn't perfect, but it gets me close.

Once the woofer is positioned as desired, I use transfer punches through the woofer mounting holes to make a small center mark.

Afterwards, drilling the holes is a piece of cake. I drilled the holes large enough (7/32") to easily accept #10-24 screws. I'll use locknuts and washers on the back side.

 

Treating Edges [December 26, 2012]

I managed to cut the recess for my "manhole cover" in the top of the woofer box. It will allow access to the top woofer when it comes time to mount the driver and solder wires. Note that this is my modification, and is not part of SL's plans. I will fabricate round pieces of 1/4" hardboard to fit into the recesses. They will be installed with weather-strip gasket material to make a tight, rattle-free joint.

I cut the recess on my router table using a 1/2" rabbeting bit. I knew that this plywood was susceptible to splintering from my earlier routing work, so I applied some shellac around the area to be cut. My thinking was that the shellac would sink into the top wood fibers and serve as a glue, binding them together. It was an experiment to reduce the tearout.

It didn't work. Even with a very light first cutting pass, I had fairly significant splintering of the wood along the cut edge. Thankfully the woofer box will be painted, and  Famowood-brand wood filler is an option to hide the chipping. With a couple of applications so far, it looks good.

Hindsight thought - the Bosch 1/4" down-cut spiral bit worked best with this plywood to prevent tearout. I should have made a light scoring pass with it before cutting the through hole in the part. [Hindsight alert #2.  What the heck - maybe cut everything with the spiral down-cut bit! I wonder what sort of new problems that would bring about?]

I'd like to find a way to put a 1/8" radius along the cut edge, but no router bit I own will fit into the shallow recess to make the cut. Even the small-pilot Dremel bit that I ground down hits bottom in the recess before the cutting edges make contact. I'll probably just use some sandpaper to ease the edges and be satisfied. In the end, this part is mostly hidden beneath the bridge anyway.

I revisited the tweeter sub-baffles with a small block plane to chamfer the edges along the sides and tops. This went very fast, and took only 10 swipes down each edge. It's looking more finished now. Primer and paint will come later - probably in spring when the weather improves.

 

A Day of Drilling [December 27, 2012]

I needed a way to clamp the angled baffle to the bottom of the woofer box for when I glue them together. There's nothing simpler than using screws to clamp a joint tightly while glue dries.

I didn't want the screws to show from the listening position, so I planned to insert them from the bottom of the box into the miter. To estimate position and depth, I stood the pieces on edge, and just placed different screws into position - by eye. I settled on #8x1" self-drilling screws from McFeeley's as a good length if I counterbored the screw hole about 0.150" below flush. They were positioned about 0.7" from the box's front edge.

Because of the need to counterbore and to countersink the land for the screw head, I used a countersinking tool that does it all simultaneously in one drilling operation. My kit is cheap, and the cutting edges are fairly dull, but the #8 bit did an OK job considering the tool quality. Once decisions were made considering screw size, placement, and depth, the drilling went very fast.

I also placed three mounting holes in the piece at the top inside of the box. The screw holes here won't show either. They will assure a good clamping pressure when the time comes.

I needed to fabricate the "manhole covers" that cover the access holes in the top of the woofer boxes. I will make these 7.5" round parts from 1/4" hardboard. After rough cutting some square blanks, I drew the cut outline. Before I cut away the center, I also drew the bolt circle for the mounting screws. I plan on using 6 screws in each cover to hold it in position.

From mathematics, when you adjust a compass for drawing the desired bolt circle, the compass radius is exactly correct for marking the spacing between the 6 holes on the bolt circle. I drew arcs that intersected the bolt circle to mark locations for the screw holes. I plan on using small #4 x 1/2" flathead screws to mount the covers to the top of the boxes.

I also contacted Madisound by phone today to add items to my order before it ships. I added 8mm female quick disconnects for spade terminals (part number QC8MM) to fit the large woofer terminals.

The quick disconnects that I added to the order don't cost much money, but by including them in the driver order, shipping costs are essentially free.

 

More Woofer Box Work [December 28, 2012]

I didn't have as much time today as I'd like, but I got a couple of things completed for the project. First, I routed the "manhole covers" from 1/4" hardboard using a downcut spiral bit - outdoors at 31 degrees F. Brrr! I made the OD 7-7/16 inch to allow a 1/32" clearance all around (1/16" total difference in diameters). The rabbeted recess for the covers was within a couple thousandths of an inch of the 7.500" target, and I knew that I had to allow a little extra room for paint.

Back inside from routing, they looked good in place. There's a small 1/8" pivot hole drilled through the center of the piece for the router jig, but I will probably not even attempt to fill it with Bondo or other filler. It will be too small to make any difference at the woofer crossover point of 120Hz, and it's hidden under the bridge.

I also cut biscuit slots to join the two angled baffle pieces inside the box. I might choose to not screw them together  at the joint, but to just clamp  in position for gluing.

After all that was done, I did another dry fit of the parts with biscuits inserted to hold it all together. The fit was very good. I'm still pondering methods of placing biscuits accurately enough along the sides of the baffles to avoid running screws through the side walls of the enclosure. Screws are simple, and will be mostly hidden by the bridge of the top units, but there's something about the challenge that appeals to me. We'll see if cold logic (use screws!) wins out, or if I attempt new techniques to get good registration with biscuits.

At any rate, the fit was good and I'm pleased with progress so far. I am looking also for a way to fasten the top angled baffle to the miter on the spacer to hold it in place while glue dries. Nothing appears easy, but drilling pocket screws at a 45 degree angle is one possibility. The drill press table tilts, and I'd have to clamp securely to do this. It seems like a lot of work though.

One last photo in this group was made looking through the "manhole cover" port on the top panel. Because my shop is illuminated by different compact fluorescent bulbs mounted in round clip-on fixtures, the color temperature for photography varies among them. The differently angled pieces were illuminated by the different color temperature bulbs, and I found the effect fascinating. Consider it speaker art.

Today, I ordered some items from Parts Express for the project. I plan to inset round Speakons into the tops of the woofer boxes, at the rear. That will make it easy to connect and disconnect the mid/tweeter part from the woofer box for moving. I expect that I will need to adjust position many times once they are in my listening room to find the best sound. I've never had dipoles before, and I'm sure that they will react differently to my room's acoustics.

While I could surface-mount the 2" diameter Speakons, they will look better recessed into the top surface of the woofer boxes. I'll use a  2-1/8" diameter Forstner bit to do that. Forstner bits are available in a Woodcraft store that's about 45 minutes away by car, but the we have about 5" of snow on the ground and it is still coming down.

No driving today for non-essential items.

 
Not Much to Report [December 29, 2012]

I sanded the parts for the woofer boxes today. It's boring work, but some sanding needs to be done before assembly. After assembly, access becomes limited for interior parts.

I did manage to try some Mirka Abranet sanding pads on this project. I've heard good things about them, so I bought a small assortment to try. I began with 120 grit, and followed that with 180 grit. They worked well.

Dust, however, settled on everything in the shop. Even with the dust collection bag built into the sander, fine dust managed to get everywhere. While it's possible to attach a vacuum to the sander, I couldn't run my Shop Vac for a couple of hours without damage, or blowing the circuit breaker on the power strip. I suppose that I should refrain from power sanding indoors, but the weather outside is miserable.

I have some cleaning to do.

 

Holes for Speakons and Pocket Holes [December 31, 2012]

The Speakon quick-connect parts were ordered from Parts Express on Friday. They arrived via FedEx this morning around 10 am. That's fast service.

I drove to Woodcraft in Parkersburg yesterday to pick up two Forstner bits -  2" and 2-1/8" diameters. I though that maybe I'd try the smaller 2" even though the Speakon is 2" also. Having both the recess and the part 2" in diameter provides no clearance in the recess for paint, etc. However stated specifications are sometimes off a little, so I thought that I'd try it anyway.

It worked. I'll have a little paint buildup in the recess, but I can always file the outside of the Speakon flange a bit to make it smaller. The tight clearance looks very tidy, and I was afraid that going to the larger 2-1/8" recess would look sloppy by comparison.

I began by trying the 2" WoodRiver bit (their house brand, usually Asian in origin) on a piece of scrap plywood. Even though the WoodRiver-brand Forstner bit was inexpensive ($12.59) compared to a higher-quality Famag ($46.19), it did a great job. Both the 2" and the 2-1/8" WoodRiver bits were in stock, too, which wasn't the case with the Famag brand. I did use a fine-grit diamond file to touch-up the cutting edge of the 2" bit before I used it. It had a burr on the leading edge that was easy enough to remove.

Once the test with scrap wood worked out OK, I went on to the real baffles. I set up a fence of sorts on the drill press so that the baffle wouldn't move around much. It also aided finding the center when it came time to cut the 1" through-hole with another Forstner.

Once the holes were cut, I rounded the underside through-hole edge on the router table. The 1/8" chamfer bit was already in place in the router table, and it took about 10 seconds per panel to do. Now I won't cut myself on sharp, splintery edges when it comes time to wire the Speakon.

This work should really pay off when it comes time to move the speakers to find the best sound. The Speakon quick-connect system is a proven design.

A few more preparatory steps (drilling wire holes, etc.) for the woofer box panels, and I should be able to glue-up soon.

I also cut some pocket holes for the upper baffle. I need a way to clamp it for gluing, and pocket holes with screws made sense here. I used my Kreg pocket hole jig to cut the holes on an angle. The 1" long screws will exit the baffle at its end and enter Part E. I'll drive the screws after glue is applied and other box parts are positioned correctly to ensure a good fit and squareness.

For now, I just placed the parts on end to get a sense of the arrangement. If I can, I'll use the Kreg plugs to fill the pocket holes. Even though they are well under the top panel of the box, I don't want to see them if it's possible - even when I crouch down behind the speaker.

I didn't have enough time to begin gluing, but I did start planning the wiring run, and where to put through-holes in the baffle. I am planning to place the larger 8-pole Speakons left-of-center on one speaker, and right-of-center on the other. It will shorten the speaker cable run a few inches, and I believe the wire routing makes more sense visually. A big wire in the center splitting off left and right didn't seem appealing.

I also plan to use Wood Artistry's Speakon brackets instead of fabricating them from wood. I'm borrowing them from my Orion build because it will be easy, and because of appearance.

 

Gluing Begins [January 1, 2013 - New Year's Day]

I had all day to work on the speaker today. No interruptions, nothing.

I began to glue together some sub-assemblies. In the first photo, I glued the top panel to the short front panel. I did not glue them to the side panel. Because I used biscuits, I could use the side panel as a jig for gluing. To prevent the two pieces being glued from sticking to the side panel, I put some wax paper between the top/front pieces and the side panel. After about an hour, I pulled the now-glued top and front panels from the side panel for further work.

I needed to run the clamping screws through Part E into the top panel accurately, but was having a little difficulty with the piece moving while driving screws. It has a miter which makes clamping more difficult. To solve that problem,  I used a corner clamping block made for another project to apply firm pressure to the part, and drove the screws home.

Now it was time to glue the upper baffle to Part E at the miter. I positioned the parts using one side panel as a jig, and used the pocket screw holes that I drilled yesterday for clamping screws. I allowed the glue to dry several hours for this sub-assembly because of the small, unsupported glue area. I wanted to be sure of a solid joint.

After a safe period of time passed, I pulled everything apart again and plugged the pocket screw holes with Kreg plugs. They were Maple (close enough) and were glued in place. After the glue dries, I'll use a hand plane to trim the plugs flush to the surface. Right now, it looks fairly rough.

One thing that I noticed with all the fitting is that the side panel material is slightly bowed. I'd have to clamp with a considerable pressure in the center to ensure good contact with the baffles. If I were to use biscuits, that would mean clamping with cauls. That's too much work, so I made the decision to use screws through the side panels. I spaced the holes for the #6 deck screws 5" from the front edge and 5" from the back edge where they intersected the baffles. To do that, I positioned everything in its final assembled position, then traced the baffle edges onto the side panels with pencil.

It's not ideal to have screw holes in the sides of the boxes, but they are flat head and countersunk. . I located the screws so that they'd be covered the bridge of the speaker, and not visible in normal use. I've milled countersinks for them about 0.020" below flush which will allow me to use Bondo putty to cover them over, if I wish to anyway.

 

More Progress [January 2, 2013]

Yesterday I filled the pocket holes with wooden plugs. The glue dried by morning, so I planed then sanded the protruding parts of the plugs flush with the surface of the baffle. When painted, you won't be able to see the plugs at all. Also, this is up very high underneath the woofer box, so you'd have to crouch to see them anyway. It was just something to do while glue dried elsewhere.

I had forgotten to mark the screw locations for the "manhole cover" on the top panel, but with only #4 screws, pilot holes were not necessary. Good thing, the top panel had the short front panel piece already glued to it making difficult the use of the drill press. With the little #4 screws, all I had to do was to use some heavy cardboard (measured 0.023") as a spacer to center the hardboard cover in the recess, then drive the screws down. Easy.

I included a picture of my very small, crowded shop with both of my "work tables" occupied with work. The "work tables" are a piece of MDF on my table saw, and the top surface of my router table. Oh, and the little 12" square top of my cheap band saw. The room is located under my garage, and has low ceilings too. Access is limited, and it's crowded because of its use as a storage area too.

When the glue dries, I pull the subassemblies from the side panels and move on to other pieces that need work.

[Later today - more gluing]

I began gluing up the sub assemblies into the part that fits between the side panels. I forced parts square to the side panel that is used as a jig for the assembly and gluing. Note that I left these pocket screws open - I inadvertently glued the small Part E to the top panel before I realized that wouild make it very difficult to clean up the plugs. It's not an issue structurally. Even the appearance isn't affected unless someone crawls around on their knees looking at the top part of the woofer box, and even then it's neat. It's one of those things that bothers me more because I did a step out of plan.

To clamp the two baffles together without using screws, I used a 6" clamp through the woofer hole, and two others pulling on bolts that I had inserted into the driver mounting holes. It made a handy grip point for applying good clamping pressure for the glue. There are biscuits in the joint too, and those were liberally glued before insertion. Note the use of wax paper to keep the squeezed-out glue from adhering the assembly from the side panel. I will remove the glued together parts tomorrow for a little more work (drilling pilot holes into the baffle edges) before the final step - attaching the side panels.

Working step-wise like this allows me to check the quality of fit as I go. If everything were glued at once, I wouldn't have time to check squareness, adjust clamps, whack pieces into proper alignment, etc. before the glue set up. I don't have enough arms and hands for that sort of approach.

Note that some of the clamps you see in some of the pictures aren't being used for clamping glued joints. I'm using some of them to flatten a slight bowing in some of the side panels, and in others, I clamp assemblies square to each other. By clamping them square to another part and flattening any slight warping, I simulate the position in which the pieces will be eventually assembled.

 

Gluing Again [January 3, 2013]

Gluing takes time when faced with limited work space and clamps. I've begun gluing the side panels to the center assembly. I noted with some irony how square things are before the glue goes on. I know that the gluing process, being somewhat rushed, always tweaks squareness. I wish there were a way to assemble the panels and then somehow wick the glue into the joints. Plastics, yes, wood, no.

I'm glad that I put 4 screws through the side panels to tie it to the baffles. It would be very difficult to apply pressure on the mid-panel without using some sort of cauls. I still resorted to using one makeshift caul because of the bow of the side panel. It was barely making contact between screws in one location.  I placed a piece of scrap plywood on top of the panel with a few folds of paper towels in the center to apply local pressure. It's not ideal, but it did take out some of the bow of the panel and made a little better contact with the baffle edges. It would have been OK without it, but it helped, I'm sure.

I finished gluing the sub-assemblies together on the other box. This time I was a little better prepared for the baffle clamping by using screws, washers and nuts at the clamping point. The last time, it was a rush to jam the screws through the holes while glue was setting. It worked, but this looks tidier and it places less stress on the wood around the baffle holes.

I've employed just about every clamp that I own today, as well as both flat surfaces on which to work. Naturally, after gluing, I can't do anything more while it dries.

I'll probably not update this log for a few days until gluing is done. After all, how many pictures of clamping/gluing do you want to see? (If you think this is bad, just wait until spring when it comes time to paint. Prime, sand, spot putty, prime, sand, etc., etc., etc....

Flash update.... Madisound sent an email stating that the back-ordered drivers have shipped. I'm officially behind now.

 

Woofer Box Gluing DONE! [January 6, 2013]

The top picture in this group shows the last box about to have its final side panel glued on. That was the last gluing task for these boxes. I finished gluing the woofer boxes together yesterday, and removed the clamps this morning.

Time to step back and take a look.

They are looking good. Not perfect, but very close. As I expected, gluing tasks aren't exact, and a little sanding here and there to flush some joints is still needed.

I am very pleased with the 45-degree miter at the lower front edge. That came together very, very well, and only the slightest touch up was needed at the front edge to make it perfect. I sanded a very small radius there to blend the angled piece with the box. There's no gap, no ridge, just smooth goodness now.

Squareness of the boxes is also very good.

I still have to sand away squeezed-out glue from the joint locations. When it warms up in spring, I will probably use some sort of grain filler or heavy, sandable primer to hide most of the appearance of wood grain from the painted surface. The plywood edges are particularly visible without a little work with fillers, primer, then paint. I know that from experience with the Orion woofer box.

My available free time left to work on these is rapidly disappearing. I have the rest of this week away from work, then I will be very busy for a while. I will have small bits of time in the evenings, and on the weekends (those that I don't work), so maybe its time to start on the electronics. Electronic assembly work is more suitable for small blocks of time.

 

[continued on page two]

   

 [top]


[main audio page]


All content  on this page is copyright © 2013 William R Schneider. All Rights Reserved.