Top panel cnc jig

20110721-102528.jpg

Welp, this is the last jig I needed to make for beatseqr production. It holds the top panel portion of the case so the cnc router can do it’s job. When I was just making one beatseqr at a time I had much much less efficient jigs to hold things on the cnc table, so this is a huge improvement. I’m getting repeatable results that I’m happy with when I cut multiple cases in a session. There’s less guess work, less waste, and less cursing… So overall a good situation.

Also on the final infrastructural push, I designed and built a fume extractor for soldering. Which makes me happier when I’m hunched over a circuit board soldering for hours. Here are some photos :

20110721-103307.jpg

20110721-103318.jpg

Horizontal cnc jig results

Alright, this jig I made to hold the dremel flex shaft horizontally worked *beautifully*. I hand-programmed the gcode, and I obviously have some more fine tuning to do, but overall, I’m on the right track!

Here’s the nice clean hole the jig produced. Adjustments required, but almost there.

20110619-110332.jpg

Here’s the jig that’s holding the case down.
20110619-110351.jpg

Video of the jig in action:

Horizontal cnc mod

Alright so one of the last hurdles to jump before I’m ready for production is being ready to cut holes in the sides of the enclosures for the USB cable. I’ve been doing this by hand with a dremel with *extremely* variable results. :-/ So, when my pal Jim was up for a visit earlier this year, he helped brainstorm on some ideas and this was the concept we ended up with. Using the dxf files for the cnc machine, I was able to locate the hole pattern and design a new piece that will bolt on to the existing z axis gantry. This new piece, with the help of some shape lock, securely holds a flex shaft attachment for a dremel… Which in theory should let me jig a beatseqr enclosure to the table and cut nice clean square holes without and problems or sloppiness. In theory. But man, I’m getting really close to having the infrastructure in place to do what I want. Getting pretty exciting up in here!!

20110614-113013.jpg

Solving problems: LRF support

20110531-054204.jpg

When I have a bit more time, I’ll post a video, but I just couldn’t locate mechanically fastening LRF’s to the specs I need. So I drew up a concept on some paper, redrew it in illustrator, exported it to dxf, imported it to sheetcam, and cut a casting mold. I used the smooth-on shore hardness 40 silicone since I’ve decided it’s too tough to work with for my buttons… Turned out pretty good! Maybe a little too tall. But it will be easy enough to slice off this mold and try again. I’m loving the time from concept to iteration. Very fast!! And I’m not committed to hundreds of failed designs. Just a couple few. I’ll even use these for experimental projects.

Version 4.51 silicone button array development

Research and development continues with cutting (and recutting, and recutting) the new silicone button casting mold. I’ll be verifying the dimensions tonight, but things are moving along nicely. I may be able to get one of the new v4.51 boards close to fully assembled. Which I am very
much looking forward to. The last major task is to engineer a jig to hold a dremel flex shaft that will attach to my cnc machine. This will let me cut a hole for the USB jack with precision and repeatability on the case. After that, I’ll be able to kick into a higher gear for producing units and kits. Whew!

20110524-024103.jpg
Fail. 🙁

 

20110524-024125.jpg
a sea of machineable wax chips. This mold would end in disaster.

 

20110524-024205.jpg
Present and future casualties of my learning process

 

20110524-024220.jpg
Try number 3. This one went well after 5 days of chipping away at it.

 

20110524-024240.jpg
Smooth-on dragonskin-30, a two part a/b platinum cure silicone, in the happiest place on earth: a beatseqr button mold

 

20110524-024254.jpg
It. Looks. GOOD!

 

beatseqr 4.5 is alive

The good news is that after I received my new boards, I was able to put one together and everything pretty much worked reasonably well. Well enough to use as demo units. 🙂 So I have at least one more demo unit that I’ll be completing within the next month or so.

beatseqr version 4.5, assembled PCB

The not-entirely-bad news is that I did find some aspects of the board that I evaluated to be sub-optimal, so I’ve made a handful of changes and have submitted the design for manufacturing. I’m hoping to see those boards by the middle of next week.

This was my first ever project using Eagle cad, and I’m feeling pretty good about it. The learning curve was ok for me. I don’t have any professional experience doing this stuff, so I guess Eagle is easy enough. It’s a little tough at the beginning, but once you get into the groove of getting your parts created in a library, you’re good to go.

I pretty much took it as a given that I’d have some kind of critical bug or wiring error that I’d have to eat the costs of. The errors I found weren’t wiring problems, they were problems with drill sizes. All of the wiring was 100% accurate and functional. I ponied up for the autorouter, and I’m glad I did. I was able to move some things around on the layout, adjust some of the my fail-over pins and pads, re-run the autorouter, and was back to being in the fabrication queue with my board house within a couple of hours. Some people just don’t like eagle’s autorouter, but for me it’s awesome. I have something like 500 holes on my board and the autorouter did the job in about 2 minutes. Easily worth the cost. so I’m feeling like I can go ahead and work on the manufacturing optimizations I’ve been thinking about for pre-assembled units without having to worry about give up all of the hard work I put into the last functional layout.

So, some highlights on this round of boards:

  • the board is now properly branded with the beatseqr logotype. Woo.
  • I took a risk on putting holes on the board for the support legs on my faders, and they work great.
  • I took a risk on putting offset pads on the board in order to use surface-mount board headers to interface with the arduino mega.. and they work pretty well!
  • silk screening on two sides worked as I figured it would. I thought it through and reversed all the text on the b-place layer, and that was a good guess. I missed a bunch of stuff on the t-place layer that should have gone on the b-place layer, so that’s corrected in version 4.51. Lesson learned: change the color of the b-place layer… that will make it easier to see what’s on that layer.
  • I put some holes on the board to line up with the mounting holes on the arduino mega, and they lined up perfectly
  • Totally surprised, pleasantly, that my boards came back routed to my irregular shape, exactly as specified in my design. that’s a major post-production step that I just don’t have to worry about. I’m very happy about that.
  • mounting holes for the project case lined up perfectly. This bullet and the one above are epic, because when I was using my board house’s own PCB layout program, I couldn’t specify the irregular board outline I need, and it wouldn’t let me put my mounting holes as close to the edge as I needed them. Both problems solved by going to eagle cad. Plus I can get silkscreening on the bottom layer at no additional cost, so heck yeah.
  • all of the drill sizes that I ported over from my other design worked very well.

So … after a a few months of cooling the jets and learning eagle, I’m back to full speed ahead again.

DJ Mojo drum and bass demo

Yeeeeah, check this video out! This is my friend DJ Mojo in L.A. He was a major guy in the underground electronic music scene in the 90’s, and since then has been documenting the history of our slice of the musical history spectrum, thinking very deeply about new music interfaces, producing remixes, and mix sessions. Here Mojo is demoing Beatseqr on a Windows machine using the Sonar DAW. Awesome!!

Find Mojo on Facebook, YouTube, and Soundcloud