I'm tinkering with motor control (the feedback part at the moment)

I've found sites that drill a hole in one of the gears and put optos on either side.

I would like to avoid modifing (IE: cutting) the Tamiya Gearbox as much as possible. A Decal and reflective sensor is one option, another is gluing a small magnet to one of the gears. The main motor revs at 10,000RPM at 3V. The final reduction is 207:1

Ideas?

__________________
Bill
Home of the
Firefly PIC Tutor
Inchworm ICD2

http://www.blueroomelectronics.com

 

use a small proximity detector to count the teeth on the final gear ?

 

What sort of resolution do you need?

Lately I've been monkeying around with some 2D GMR sensors - you strap a magnet to a wheel and (if you can align things well enough), you get something that resembles a sin-cos output.

 

I had the same gearbox and looked at using a hall-effect sensor (allegro, don't recall the part number). I found that using that flexible magnetic material (used for printed fridge magnets) works pretty well. It's thin but you have to find the individual magnets to avoid multiple readings though you could compensate in sw.

you can insert a small piece of the thin magnet material into the first vertical gear in the train - the one immediately following the gear on the motor shaft. It should be pretty clear where it fits. Then make a holder for the HE device so it hangs down next to that gear. I'd make it adjustable so you can move it closer while keeping enough clearance. the problem I had was that the gears have too much side to side play so it takes very careful adjusting. Maybe a thin washer would help keep them from wandering.

One other idea I had was to make a photo-interruptor that could fit between the gears. drill a hole in the gear. there may well be a commercial unit that could fit but it's pretty tight.

 

Why nec drill a hole in one of the gears? U could just count the number of teeths pasing in front of a optical swich (u can make one from a LED and a receiver). An uC can handle most of the functions (counting, math, PWM etc.).

__________________
If is too complicated.....most of the times is the wrong way.

 

those teeth are TINY plus the gears slop around about 50 mils.

 

Y....but it's still feasable

__________________
If is too complicated.....most of the times is the wrong way.

 

Here's a link I found where the fellow drills it. I think a round hole would work as well. I'll have to calculate the gear speed, but it should be no problem for the PIC.

The second gear (easy to get to) 2nd fastest large gear.

http://www.philbot.com/projects/TMG_encoder/

__________________
Bill
Home of the
Firefly PIC Tutor
Inchworm ICD2

http://www.blueroomelectronics.com

 

that looks pretty good but I wouldn't make such a huge slot in the gear. it doesn't look very strong and you only need a one small hole. I don't understand why he made it so big.

 

Yes, it's a really impressive slot, but it must weaken the gear a huge amount. Also it only gives one pulse per revolution, drilling a series of holes could give higher resolution - and I'm sure four evenly spaced holes would be a lot stronger than that slot!.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk