poss. OT, make lever work a switch multiple times

One way might be to hack into a surplus keyboard and substitute transistor switches for the appropriate key contacts and then drive the transistors from a microcontroller that "listens" to the movement of the lever. You may be able to do it using a series of one-shots instead, where the one-shot fires when the lever reaches a position but won't continue to fire if it stays in that position. You'll need to handle contact bounce, of course.

There's probably a clever way to do this using discrete logic but a state machine in a microcontroller would seem to be the simplest.

Alternately, dispense with the keyboard and build it all around a micro that has a USB peripheral and have it emulate a USB keyboard. That gives you the option of using any inputs, within reason, and then generating an arbitrary sequence of virtual keypresses.

Possibly the very (very!) early mice worked with mechanical switches but, up until the advent of the current generation of optical mice, the x-y motion was pretty much universally done using optical sensors in quadrature, via the slotted wheels.

--
Rich Webb     Norfolk, VA

I have used an old keyboard, to solder the contacts of a reed-relay across the print traces for a number of keys. Then you only have to supply 5volt to the reed coils, to remotely press a key.

You need some slop in the mechanical system - two small buttons as close as posible to each other (say less than 1/2"). The button actuator has to slide (easily) within guides - to position on top of one or the other push buttons. The triangular (sawtooth) shape attached to the lever first slides the actuator over to one end stop then pushes it down on the switch.

\/\/\/\/\/\/\/\/\/\/\/\/\ Lever bottom I ||- I-------------|| Actuator in guides I |---| |---| | | | | PB1 PB2

--
Geo

USB HID device:

You will need to modify the code to handle the one key press issue.

Let us know how you get on.

hamilton

nd

ss

Most (for instance PC classic game controls) levers are connected to potentiometers, i.e. variable resistors. Your problem is that the intended action depends not only on the lever position, but on the history of that position.

Conceptually, you need an accumulator (like, an up/down counter) and to update that count when any power/brake adjustment is made. Whenever the lever position is significantly above the accumulator reading, send a 'power' signal and increment the accumulator, and when it's significantly below, send a 'brake' signal and decrement the accumulator.

If 'twere a PC doing the control, this would be a simple matter of software, with the lever connected to one of the game port analog control pins.

i

nd

ss

eed

r

nal

ull

Not sure I follow - but use a 2nd switch (that is otherwise unused, but is still in the direction of travel). By comparing the activation sequence, you can tell which way the lever is traveling.

You're on the right track. You need a beam-break optical system (LED paired with a phototransister), and an encoder wheel. You'll need a pair of the optical circuits and a pair of D flip-flops. You can make the encoder from whatever is easiest to use, but you need to construct something like this

If you look at the bottom under "Incrimental Decoder", it explains how a quadrature signal can be decoded to yield both a step signal and a direction signal. Do a google search for "optical encoder direction".

On Jan 11, 6:55=A0am, "Gazz" wrote: ...

nd

ss

...

ull

That's brilliant!, here's a variatiom (not to scale)

\ /\ /\ /\ /\ / \ / \ /__\ / \ / \ / \/ \// \\/ \/ \/ o \__/ o \\ // up down

use a ball (or roller) instead of the slider and angle the awitches. o is a stop bar so the ball doesn't escape.

--
?? 100% natural