is there any type of sensor that . . .

[vannic]

hi all.  i'm working on a project at the moment that needs to detect what position a stick (gear stick in this instance) is, along a horizontal plane.  I've been thinking of ideas that are slightly complex and needing brackets, gears/levers and such to switch switches or turn pots or rotaries, but it all starts getting fairly complex that way.  I'm just wondering if anyone knows a way to detect an objects position along a horizontal plane (basically i'm trying to detect what gear i'm in . . . 1, 3, or 5 in the standard H pattern).  unfortunatly the gearstick doesn't have a lot of lateral movement so there isn't a whole lot of room in there either.  any ideas appreciated.

** edit **
just thinking about it a bit more, i guess the theory of a slide pot is waht i'm looking for.  a different resistance based on it's position on the slide.  however i'd rather not use a slide pot as i assume the slide mechanism will fail quite quickly as the gear stick gets heavy use. 

[hotwaterwizard]

a magnet and 5 reed switches could do it low tech.

[hotwaterwizard]

Here is a Diagram for the Reed Switch Idea

The second is a Digital Matrix with Diodes to display a number for which gear you are in.

[vannic]

cheers.  i had gone down the reed switch path early on and decided that my gear stick didn't have enough lateral movement which could likely cause interferance between reed switches in different positions and ending up in false readings.  have you done this project successfully?

[Hero999]

Good idea, it shouldn't be hard to add a lower case 'r'  for reverse.

I think the idea is to mount the device to the underside of the gear stick.

What about using an IR LED and an array of photo-transistors?

Mount the LED on flexible cable and put the sensors underneath it. Ambient light shouldn't be a problem on the underside of the car so there shouldn't be any need for filtering or pulsing.

[vannic]

the electronics of the display aren't a problem.  digital matrix with diodes, PIC, it's relatively simple to do.  the problem is having a foolproof method of detecting the position of the stick.  mine, like i say has very little lateral movement and things like reed switches, photo transistors, hall effect sensors, etc will likely suffer from some mis-readings at some stage.  especially as the stick also has a certain amount of slop to it and moves around significantly while driving.
the best solution i've found so far is some arrangement to detect top, left, bottom and right positions of the stick and using a PIC to "decode" the switch combinations (ie, top and left = 1st, bottom and left = 2nd, top only = 3rd, etc).  my current plan of attack is to do this with a simple 4 microswitch set up or something similar.

[Herman the German]

Hi vannic,

I'd go one further step and check the leverage going into the gear box. There should be a distinctive (relatively) long way of the push/pull rod. Reed switches are a bit large to sense different gears. Using a Hall sensor (TLE4905) you can place them narrow enough to avoid false readings. Their magnetic sensitive area is just a few square millimeters (2sq mm) large.

Mounting small magnets on the rod you won't have to bother with cables going to the bottom of the shift lever.

HtG