Tom Gardner wrote:
One huge problem I've had is feeding flat wire into machines. For years,
I've used a feeder design that used spring-loaded pinch rollers on 1-way
clutches cycled back and forth with an air cylinder. The variances in the
wire's hardness, oiliness, surface finish, etc. would cause havoc. And,
just the mass of the rollers would cause over-feed. We coped.

The new design uses no moving parts other than the cylinder and linear
bearing car. The wire is controlled by a pinch roller that is just a
standard 1/2" round carbide insert. The roller rides in a angled pocket
with a small spring to coax it into the lower end of the pocket and pinch
the wire against a hard block. If you pull on the wire in one direction,
the roller jams the wire tight against the block. Pull in the other
direction and the roller releases the wire and it moves easily. This
replaces the 1-way clutches. The trick is the angle! Too little and the
roller stays jammed tight, too much and the roller lets the wire slip.

The first prototype used 8 degrees in the pocket, it stayed jammed
sometimes. The next try was 20 degrees with the intention of finding the
far end where the wire constantly slipped. But, to our delight, 20 degrees
just happened to be perfect. We went one whole day without a single wire
problem and logged a15% increase in average productivity. I know that one
day is too small of a test but it is a great start. (one step closer to
retirement!)

We are currently building a new machine that will incorporate a bunch of new
ideas designed to eliminate or reduce the various problems. I think we can
improve an average 30% down-time to less than 10%. I have another wire
feeder design that is just coming up on the machining schedule. It's based
on an entirely different idea using gears pinching the wire. We also made
the machine reset automatically, added warning lights for the operator to
know if the machine is ready, reset, blocks clamped, etc. We also added a
VFD and tachometer to fine-tune the speed, beefed-up weak subsystems and
made a bunch of ergonomic changes. (another step closer to retirement!)



Another good application of a great old idea. If you can find a new
application that someone hasn't though of yet for it you might be able
to make a million. The spindle lock on my Starrett micrometer uses the
same principle and IIRC a guy made a lot of money by using it in a
fencing fastener such as for cattle fencing.