x0x, finishing touches
Time to wrap up this project on the blog end too!
I’ve had the x0x complete and in use from around fall 2011 (first live at Dekadence Dataparty 2011), but finishing this text has gotten pushed aside.. Prioritizing masses of this and that over writing, as usual. This delay however has also a little in common with the “extra” I wanted to do for this project.. I give you ‘Jam the x0x’:
Not by far the first take I did of this song, but a one-go live recording without additional editing anyway! Besides x0x I’m using XBase09 for drums, MC-202 routed through a MX-200 effects unit and x0x through Filterbank 2.
On to the build..
Picking things up from the previous work stage, I next moved on to painting the case. For the main color, I felt like using something champagne-ish. I knew I’d be doing the decals plain black (B/W laser printer on sheets of Lazertran), so going for a bright/light main color is good for text readability.
The color I ended up with was nr. 55650 from the Motip metallic car repair spray range. To maintain compatibility, primer and varnish were picked from the same product range. Undoubtedly any other acrylic spray would’ve been just as fine. When applied, the 55650 tone actually turned out a bit more darker and gold-ish than what the cap on the can suggested. Nonetheless, I didn’t consider the difference bad enough to justify doing everything all over again (repainting chores = meh).
One detail the paint seemed to emphasize, was the slightly misaligned angled step / bend on the top panel. It just looks way more uneven than prior to paint being applied. But little I can do about it now. Considering this was my second time doing a “real” bend on a sheet bending press, I guess it could’ve turned out worse.
Whilst waiting for the various layers of paint to dry, I cut a couple of side panel test pieces to get some ideas what they could (and should) look like. I also used these to evaluate which varnish might match the 55650 best. As the panels require relatively little amounts of it, I decided to try all the various leftovers I have and pick the most appropriate one out of these instead of buying a full can.
Material-wise there was nothing else than 15mm thick pine on the shelf, so that it shall be for now. Maybe I can replace these with some “classier” wood at a later date.. (as if).
The four screw approach on the side panels looked out of place, so I left out the bottom front hole altogether for the final version. The panels press tightly enough against the aluminium sheet with three screws, so no problem there.
Whereas painting was a no-brainer, getting the decal design sorted out took me better half of june 2011. I tested a couple of different designs, but eventually settled for the good old “something new, something old, something borrowed”. A recipe that works well for most things creative.. Not that I’m all the way “wedding” here, as the part about “something blue” is lacking. Unless of course you want to count in the ESD mat on my desk ;)
Anyway.. for the Lazertran decals I first tried using uncut A4-sized sheets. This was supposed to help me avoid the usual seams you get when using Lazertran cut to smaller bits. But as it turned out, a full-sized sheet actually stretches slightly at least when water transfer method is used. Despite the initial decal placement was spot-on, the resulting sheet was badly displaced when transferred to the panel surface. If the sheet was aligned in the middle, the patterns were over a millimeter off target on both edges! Ugly seams it is then, yay.
Not that this was the only (or even the last) bit of fail. To add a little extra to the design and make the piano keyboard look slightly more “piano”, I decided to mask-tape a rectangle and paint this resulting area white.. And for the first time ever, the paint mask bled through leaving visible edges to the panel.
One other thing was, I didn’t bother sanding away the “orange peel” surface left by the main color. Using masking tape on a surface like this usually causes minor bleed (sort of ripples?) on the edge of the masking tape as it doesn’t fasten seamlessly with the uneven surface. To mask up this bleed I added some thicker outlines to the edges of piano decal. Easy solution!
Fancy on the side
I also wanted to add a groove to the inside of the side panels. This way they would sink in with the panel a bit, making the edge seam look less of a “seam”. I first decided to practice free-hand milling the grooves using a router bit:
Pretty crap results, no surprise there.. With some additional practice, what at turned out a better solution was to first “outline” the grooves using a cut-off wheel and then mill / chisel away what was left in-between.
CNC’ing these grooves would’ve been an option IF the side panels actually were symmetrical. But as they aren’t I simply didn’t see any point in creating milling programs for both individually. These would be very much single-use, so I was way faster off doing this work stage entirely free-hand.
With the varnish, I applied four layers and after thoroughly dry, a very light smoothing polish.. Piano finish, as if I’d bother! The resulting surface is not the smoothest ever, but will do. The unevenness of the surface is visible only when looking up very close and mirroring against light.
During final assembly and testing, my potentiometer sub panel mounting solution turned out inadequate. Whereas the method of hot-gluing board risers to surfaces has worked fine before, adding countersunk screws to the mix proved a failure. The epoxy joints just didn’t hold strong enough because of a smaller layer / contact surface.
End result: When fastening the board spacers to proper tightness, some of the epoxy joints broke off and I had to dismantle everything and re-do broken joints using a thicker layer. I guess this here was the limit of how compact a assembly you can make with epoxy. Just slightly too little contact surface for things to hold together.
Furthermore, this fine tuning of mech bits contributed to sub panel getting pushed away farther from the top surface (as thicker layer of glue was used). As the switch shafts are currently so short, even miniscule dimension changes contribute in rather dramatical ways, and in this case it made the slide switches nearly unusable! The way they are now is, the shaft is still moveable using the tip of your fingernail but none still protrude enough over the top surface to be slid in a “comfy” way. So to say, replacing the switches immediately became a pending fix.. Just need to keep an eye out for some taller shaft spares. Part & source recommendations welcome of course, if you have them.
Summing up this project, boy oh boy it took a while to bag up! There certainly was plenty of practice and unexpected surprises along the way. Nonetheless everything I learned here will definitely help avoiding problems with, say, the Klee project I have under the works at the moment. Guess if I’m using Lazertran ever again ;).
My x0x turned out slightly “temperamental” in that the MCU tends to lock up / crash randomly before the unit has warmed up. This is something that still is waiting to be pin-pointed & fixed. Not that much of a problem, just need to turn on the x0x a while before doing anything with it.
Oh and the completed unit? Since the not-so-recent addition of ‘project pages‘ (it was new when I started writing this text), I’ll just save pictures of the completed unit there. Go check it out here.