Signal gloves revisited (or, 0.77$ of plastic and dirty silicon can make a world of difference)

So, as I said last time, the interface for the signal glove left a lot to be desired.  The button was easy to miss, and difficult to depress for any appreciable time.

Since I was putting together another Mouser order for a significantly more ambitious project, I decided to toss on some Hall sensors (an idea courtesy of my dad, who came up with it within about 0.23 seconds of hearing my button problem).  For those not in-the-know, a Hall sensor detects magnetic fields by using the Hall effect, which takes advantage of the force exerted on moving charges (electrons) by magnetic fields.  By applying a current 'down' a slab of conductor, you are forcing electrons to move; if there is a magnetic field perpendicular to the plane of the slab, it will apply a force to the 'left' or 'right' of the slab.  This force will cause the development of a voltage between the left and right sides of the slab, which can be measured; the magnitude relates to the magnitude of the magnetic field, and the sign of the voltage relates to the direction of the magnetic field (into or out of the face of the slab).

By placing a Hall sensor on the closest knuckle of the index finger (toward the thumb) and a magnet on the thumb (toward the index finger), I can switch on or off the signal without having to exert any force (beyond the force needed to move my thumb).

Hardware

The ideal solution would be 'backward compatible' with the existing switch, and be able to exist alongside it. The original design used a pull-up resistor from the input pin on the microcontroller; the switch shorts the pin to ground.  Fortunately, open-drain is a common configuration for digital outputs; this more-or-less means that the output is a normally-open switch to ground, rather than actively driving any particular output voltage level.  This gives the user more options for use, especially as regards logical level voltage thresholds.

The one I chose was one of the cheapest ones that came up on Mouser (TLV4906K from Infineon, part of an order for a more ambitious couple of projects whose invoice is posted here).  The part accepted the voltage range of interest (down below 2.5V) and didn't soak up altogether too much current (nominal 4mA).  The altered schematic, including the hall sensor, is shown below.

The hall sensor and its 100Ohm SMD resistor were assembled along with a set of wires and epoxied to a bit of substrate, similar to the original switch.  It was then sewn into the glove and routed to the controller module as before.  While I had the glove inside-out, I also sewed a felt panel over the controller module to make donning and doffing easier and to protect the module.  I also put a suture into one of the palm-back LED bases to keep it pointed in the correct direction when my fingers are fully extended during signalling.  The most difficult aspect of the upgrade was getting at the relevant pins on the module connector and soldering to them without damaging any of the other leads or the glove itself.

The new hall module is shown in the two pictures below.  It is mounted on the index finger, facing the tip of the thumb.  Two rare-earth ring magnets are sewn onto the thumb-tip to allow the Hall sensor to trigger the lights when the thumb is moved close to the index finger.

The Hall sensor trigger is much easier to use and more consistently triggered than the pushbutton switch.

Lessons

The main lesson was "magnets are awesome and should be a part of every design ever".  The only down side is that a strong magnet can have deleterious effects on magnetic storage media (does that count as a pun?); however, this shouldn't generally be a problem, especially in this wondrous age of solid-state media.