It was not easy coming up with z box design. I was quite concern on the feasibility of the whole project when Leda Resources decline to construct it even after I've given them technical drawings. The local DIY community wasn't too encouraging in me designing from scratch. A few helpful fellas forwards me links to existing design using some other scanspeak speakers and recommend that I just follow those. It was a good thing as I searched for more references for ideas. One design that I've taken a liking is shown here. It is in bright red. It has a very nice elegance, especially the slant/tilt similar to the SF Cremona.
I would love to achieve the same, but I foresee MDF has to be cut to a high accuracy - ie +- 1mm. Furthermore, the slant is a major design headache. The Ideal slanted design would means cutting the MDF at a precise angle to make the top plate totally flat!!
OK. Enough bs, let's get on with the design proper; The following are the major considerations.
1. Internal volume needs to be at least 52L.
2. MDF at least 18mm thick, 25mm for the front baffle.
3. Internally braced to increase rigidity.
4. Tweeter height around 80cm from floor.
5.
TMM configuration.
6. Curves to allow back-to-back veneering.
7.
Embedded mounting nuts - improve air tightness.
These cut panels are given to the lumber stockist for cutting. I decided on 18mm thick MDF throughout (originally wanted 25mm but was afraid to be unavailable). The front baffle will be 36mm thick constructed by joining two 18mm pieces together. The bottom plate also will be constructed in the same manner for structural rigidity.
The internal bracing is purely my own weird concoction, it would improve the cabinet rigidity nevertheless:
The MDF were ready for collection within a week. I have had no idea how much wood it would turn out to be. When I arrived, the carpenter went to the back of the store and push out a trolley stacked with the MDF. It was really heavy and I've to move everything all by myself - 1st into the cab and later up to my apartment! I would estimate the whole lot to weigh at least 40-50kg.
The MDF nicely piled into one corner of my 'workshop'#1.
Finishing: I will be using Maple veneers. It will be stained deep mahagony or something in that shade. The final coat I am yet to decide on gloss/satin varnish or 'hor e ki' by doing piano laquer gloss.
The Work bench: the main tools are the Dremel 395 rotary tool and the AEG power drill.
A close-up of the Dremel rotary tool with cheap 3rd-party diamond cutter set. An original Dremel bit goes from $7-20 per piece, while this whole box cost less than $10, made in China of course.
The biggest challenge at the contruction stage is the routing. The Dremel has an optional routing attachment which unfortunately mean more $$ into the tools. To control cost, I spent a few days designing and making my own attachment. The damage - under $10 versus ~$100 for the real thing.
This is how it is done - 100% manual hand-guided. You basically rout out around 2-3mm per pass. Extending the router bit until you finally reach the intended depth. Considerable care and concentration is required when routing the edges less you overun the speaker cut-out. The noise is extremely loud and irritating. Ear plugs is highly recommended. Fortunately, there's external constuction on-going at this time. So I did not get any complaints from the neighbours!
Routing for the umpteenth pass: I lost count how many times it took to finally reach the desire depth. As you can see from the amount of saw dust, this is not child's play. Now I can understand why most carpenters are unwilling to do the routing bit. According to the Leda Resource sales guy, it will take 2 man days to make these... abit over exaggerated for an experienced carpenter, it took me about 3 days effort.
Constructing the bottom plate. It consist of 2 pcs of MDF with 4 angle braces for the back and side plates. I use dead weight to press the parts together as the clamps that I can find/buy are too short/small.
Middle bracing: This is the 1st phase of the middle bracing. As can be seen at the top of this page, the final product consist of 3 horizontal members and 2 vertical ones.
Embedded nut: The front baffle design is similar to the bottom plate; two MDF pcs glued together. Instead of a through hole drilled, I am embedding the hex nut in-between the two MDF. To ensure it do not turn with the bolt tightening, I've to 'dig' a hex-shaped hole.
As it turn out, I overestimated the bolt length for the midbass unit. I've to shave about 4-5mm - thankfully, the Dremel is also an excellent cutter- witht the cutting disk and a few hours later...
The FINAL crossover: new and more compact than the mock-up version. Actually, it is 'too small' as I underestimated the required real estate.
Partially completed pieces tied together using normal strings for dry assembly test. Due to the construction accuracy it is quite sturdy. I mounted the drivers to have a visualization of the end result.
Not bad eh? I could not resist hooking it up to hear how it sound since I've already have one xover completed. It is much better than I'd expect from a tied up box! The bass is rather loose with all the air gaps, I now have great expectations for the end result. Lesser box dependent characteristics like the high and mid-high freqencies are much more focused... truly remarkable.
This is one of the most critical part of the assembly... due to the non-reversible nature of glueing, one wrong step and you'd be back at square one. A most horrifying thoughts especially after all the hard work put into it. I started by assembling the base component into modules before fixing them together. The bottom plate forms the base that each module is added. Based on my design, the assembly sequences are as follows:
In the process of assembling the 1st speaker, I realised that glueing the pieces together is not as straight forward as it seems.
This is a ported design. I got these from Leda Resources as well... looks like a one stop speaker parts shop. The port have stiffeners at the curved end. For mounting, I removed these using the Dremel and drilled 8 holes... a little overkill for a none structural part. In its modified form I need to round off the port opening on the back panel.
The following are the shots at different interval in before the final assembly.
more pics will follow.. see the 1st Unit. For now, I am beat. Will take a rest and casually build the 2nd unit - I estimate the timing to be just in time for the arrival of the new pair of midbass (mid May 2005).
The box resonance was clearly audible from the 1st sealed box listening test. I tried many different stuffing from pillow fibers, children playmatt foam to PC foams. The sound were affected for the worse while the box resonance remained. Running my knuckles over the box gives reveals that the speaker walls are quite resonant at places not support by the internal bracing. Knocking on the front baffle gives a dull 'thud'. This means the two layer MDF of the front baffle critically damped the resonant energy.
It would seems a good idea then to 'beef-up' the side, back and top walls of the enclosure. Instead of buying new wood, I recycle the scraps leftover during the making of the internal bracing. After lots of sawing and sanding off the rough edges, I've two indentical sets of stiffener 'bars' to fit the different compartments, they are pasted on using the remaining Elmer's probond glue.
Notice the slanting wood pieces on the side wall (pix on left).
On top of this, I decided to add mechanical support to the structure in the form of wood screw. For each support section, I used 2-3 screws depending on the length.#2
Due to the mistake in using Selley's liquid nails, there are considerable air leaks. I filled up the gaps from the outside using the PVC white glue as well as filler putty. For the inside, I used silicon sealant. The are some literature on vapour emission from the sillicon sealant curing that can corode the drivers. Thus, as precaution I left the enclosure to cure in open air for 10 days before assemblying back the drivers.
see Page 2 for the FINISHING.
#1 I cleared out the guest room purely for this project. My study will temporary serve as the guest room.
#2 This is also a precaution as all the glue instruction advise using mechanical support for high load situations.
note3