Saturday, January 7, 2012

Banduristan Goes CNC!

All the secrets are revealed.  If you've been following this blog for a while, you may have noticed that it usually takes me about a year to find enough time to make one instrument and, that I haven't produced an instrument in over a year.  The reason is that in what little free time I have, I've been researching, assembling and learning how to use a CNC machine.

The reason I decided to go this rout is to remedy the issues I alluded to in prior posts.  Although I'm a fan of hand building techniques, I'm also a pragmatist when it comes to instruments.  My goal is to make the best instruments I can, and to be able to make enough instruments to leave a mark in the bandura world. My day job keeps me very busy which is why I can only make about one instrument a year.  Using CNC, I think I should be able to increase that output without compromising anything about the instrument.

Additionally, CNC opens up a world of accuracy and precision that's simply impractical using strictly hand techniques. This should allow me to finally design a re-tune mechanism for my instruments that actually works without the buzzing, loss of tone and other shortcomings of current re-tune mechanisms.

The purist might argue that this is too close to factory techniques but frankly, I don't care.  If using the CNC allows me to improve my instruments, I'm going to use it where appropriate.  

This is not going to turn into a CNC blog - there are plenty of other resources out there to learn about CNC e.g. CNCZone and dozens of other blogs and webpages on how people have built their own CNC machines and I'm not going to re-do their work.  Instead, I'll stick to bandura building and how I'm using the CNC machine to improve my work.

Briefly, the process to use a CNC machine is to first design the part in some sort of Computer Aided Design (CAD) software. I use SolidWorks .  Next, one uses Computer Aided Manufacturing (CAM) software to design the "tool paths" or the path which the machine will follow to carve out your part from whatever stock you chose. I use VisualMill for Solidworks.  

The output of the CAM software is something called "G-Code" which is read by the machine controller to actually machine the part.

So, over the last 9 or so months, I've managed to assemble my CNC machine from a rough kit, re-learn SolidWorks (I had drawn my original plans back in 1998 with solidworks) and begin producing a few part for the instrument I'm currently working on.  

As I've mentioned before, carving the neck and scroll is something that gives me fits.  The scrolls you've seen in previous posts are probably the last hand carved scrolls I'll make.  So, I decided to dive in head first and start with the scroll - the most difficult part to design in CAD!  It took me a month or so of a few hours here and a few hours there and some help from people with more experience (thanks Brad!) to get it designed and program the tool paths.  With the holiday slowdown over the holidays, I finally had enough free time to machine my first prototype neck and scroll.

Because of the complexity, I had to machine the part from all 4 sides.  This meant figuring out a way to index the stock from each side and a way to hold it down.  I chose vacuum as my clamping method and extended a technique I found HERE to index the piece on all four sides.  Here's pretty much how it went (remember, click on any image to enlarge it):

 Here's the stock in the left/right indexing position on the vacuum fixture before any machining.  It was critical that the stock be machined to very precise dimensions in order for everything to work out right.

Here's the neck after machining the right side of the scroll.  I haven't machined further down the neck so that the flat area on the right side is still present to index the piece for the next operation

Here you can see the left side after machining. You can see that it's still a bit rough on the back but that will be taken care of later.

Now I move on to the front of the neck.  Those long pockets are simply to make the neck lighter.  This prototype is made from poplar, a very light wood, but for the actual instrument, I'll be using hard maple which is quite heavy.
Finally, the neck is finished from the back side.

Overall, I'm very happy with how the neck came out.  As you can see from the photos, there are still some bugs in the program to work out to eliminate chipout and other problems but, this also shows that even though a machine is actually making the piece, there's still a lot of art involved in the process.  Whether carving a neck by hand or with a CNC machine, the craftsman must still know how to use the tool.