Close Encounters of the Mechanical Kind

I have mentioned before, at least in passing, the historical sea change in how machines are controlled that has occurred, particularly in the last 10-15 years.  More than 30 years ago as a systems analyst at Bell Labs, I recall telling management that "it's a software world now", and their future projects would be late because of software problems, not hardware.  Today, almost everything beyond a certain level of complexity is controlled by one or more microprocessors, some sensors, some actuators, and a big pile of software.  For example, even a low-end contemporary automobile has more than a dozen processors running the various parts of the car.  General Motors says that two-thirds of the budget for the transmission they designed for a hybrid gasoline/electric SUV was for software development for the dedicated processor that operates the transmission.

 I have recently been presented with the opportunity to explore one of the historical alternatives to software control.  I came into possession of the floor-standing Regina music box that sat in my Grandmother Cain's parlor for many years.  It's a handsome piece of antique furniture, shown here in the garage after unpacking from the shipping crate.  This device plays music encoded using projections stamped into thin metal 15.5-inch diameter disks.  As the disk rotates, the projections pluck special gears that in turn pluck the tuned teeth on a steel comb to produce sound.  Somewhat surprisingly, there are a couple of places where you can still have new disks made.  The playing time for a disk of this size is about a minute.  This particular music box includes a disk-changing mechanism that can cycle through a dozen disks, lifting each in turn into position to be played.  Grandma's music box doesn't work enough to play disks -- yet -- but here's a YouTube video of a similar model that shows the basic operation.

The mechanism itself is fascinating (or a nightmare, depending on perspective).  A strong spring motor turns slowly; the turning barrel for the motor has a number of cams cut into it; each cam moves one end of a lever that, through various linkages, is translated into the proper motions at the proper time.  In one sense, I am incredibly lucky.  The box is stopped at the point in the cycle where all of the interesting things involved in changing a disk are happening.  This makes it possible to manually move the followers into the depressions in the cam and watch the corresponding levers move.  If the mechanism had stopped in the middle of playing a disk, working things out would have been much more difficult.

Perhaps the most interesting item is the governor that controls the speed of the spring motor.  The critical piece of it is that gold-colored widget near the top of this photograph.  The video linked above includes a brief shot of the governor in operation.  It spins at high enough speed that air resistance becomes a factor.  That air resistance is transferred back to the spring motor through a set of gears that provides a few-hundred-to-one mechanical advantage.  An ounce of pressure at the governor translates into tens of pounds of resistance that the spring must overcome, slowing it down.  The governor has two spring-loaded wings that unfold when it spins too rapidly, increasing the resistance and slowing the motor.  I'm not sure that I ever knew enough physics to actually write down a set of equations describing the system.  Or to prove that there's a single stable equilibrium for it.

So, what worries me most about the problems involved in restoring this disk player mechanism?  The people who have provided me information thus far all seem to be significantly older than I am, and I'm no longer young.