Position sensor for a carriage that moves along a rod

[tijoseymathew]

i am working on a project. in which a carriage moves along a rod much like a printer carriage.
what are the best options for postion sensing ?
I thought up these:-
1. use a stepper and a switch at one end. so ate intialization the stepper rotates till this switch is closed and then positioning is easy.
2. use a dc motor with a optical strip which gives the positon.

the first has the problem that i cant find a stepper with the right no. of steps per rev.
and for the second one i dont know how i can sense such small lines.

one another idea is to use resistive wire strip.

any other ideas??

 

[kender]

Re: postion sensor.

What's your desired precision for the position measurement? How long is the rod?

Optical strip is a good idea.
If the carriage uses rack and pinion with teeth to move along the rails, you can add a rotary encoder to the motor. If you have a frictional rack and pinion, I would add a separate non-driven wheel with a rotary encoder for position sensing.
If your rod is short, you could use linear potentiometer.

P.S. I’d like to invite your question to a specialized group dedicated to sensors: **broken link removed**

 

[tijoseymathew]

postion sensor.

about one mm on a 25cm rod.

yeah optical strip is good but i dont think i can find those sensors around here for a reasonable price.

rack and pinion with teeth is not what i want to use , as then i will have to worry a lot about friction.

linear pot seems the best .
but is there any other alternative?

 

[rikie_rizza]

postion sensor.

If its only 25 cm I would choose optical method. Not using strips, but measure the light source attenuation.

Heres the Idea:
1. One end is an IRled light source.
2. The other end is an IR sensor. Photo diode or photocell is a good choice. I prefer this is the moving part.
3. The closer the IR to the sensor the higher the voltage.
4. Convert using adc, calibrate, and you're done! Cheap and easy. use microcontroller's internal adc.

 

[kender]

Re: postion sensor.

If its only 25 cm I would choose optical method. Not using strips, but measure the light source attenuation.

Heres the Idea:
1. One end is an IRled light source.
2. The other end is an IR sensor. Photo diode or photocell is a good choice. I prefer this is the moving part.
3. The closer the IR to the sensor the higher the voltage.
4. Convert using adc, calibrate, and you're done! Cheap and easy. use microcontroller's internal adc.

Ligh output of the LED is dependent on teperature. So, the calibration should include temperature too. To my ignorance, I can't find a single application where the light attenuation would be used to measure distance (except, may be, distance to other galaxies).

 

[Pavel_S]

Re: postion sensor.

Try to use optical encoder like is in mouse. You can count pulses from phototransistors (phase shift between signals gives information about direction of movement). Of course you will need additional optical sensor for base positon of carriage.

1. Carriage goes to base position
2. Carriage goes to desired position counting pulses from optical encoder.

 

[tijoseymathew]

postion sensor.

all the optical sensors require gr8 adjustments whihc are quite beyond my scope. I think that the linear pot. method is best. i am also investigating on the merits/demerits of using a screw type rod to move the carriage using a stepper

 

[Pavel_S]

Re: postion sensor.

With potentiometers there are some "small" problems - wiping out (depend on kind of movement), long term stability, temperature stability. It is not the best solution for positioning or position reading especially for mentioned resolution.
I think long screw with stepper motor is the best solution ...

 

[rikie_rizza]

postion sensor.

Yea, but long screw makes the thing move slower.

 

[nandhu015]

Re: postion sensor.

all the optical sensors require gr8 adjustments whihc are quite beyond my scope. I think that the linear pot. method is best. i am also investigating on the merits/demerits of using a screw type rod to move the carriage using a stepper

I think you can use opto interruptors which need no adjustments

regards
nandhu[/img]

 

[kender]

Re: postion sensor.

all the optical sensors require gr8 adjustments whihc are quite beyond my scope. I think that the linear pot. method is best. i am also investigating on the merits/demerits of using a screw type rod to move the carriage using a stepper

I think you can use opto interruptors which need no adjustments

regards
nandhu[/img]

nandhu,

How would you sense a 1mm object with the interrupters that you show on the picture? They are much larger than 1mm. Do they have lens? If so, how good are the lens. Focus adjustment of some sort is usualy needed for lens.

 

[nandhu015]

Re: postion sensor.

hi kender

they do not have any lense. there is just an led and a received fitted in. (like the one we use to see in a mouse)

I thought of using it for counting pulses. and detecting end positions

Nandhu

 

[kender]

Re: postion sensor.

hi kender

they do not have any lense. there is just an led and a received fitted in. (like the one we use to see in a mouse)

I thought of using it for counting pulses. and detecting end positions

Nandhu

Hi Nandhu,

The receiving surface of the sensor is, probably, 5mm in diametrer. But we need to detect 1mm strips that are 1mm apart. There will still be some variation in the outptu of the sensor, although not the clear pulses, as if the strips were wider then the sensor.

- Kender

 

[nandhu015]

Re: postion sensor.

Hi Nandhu,

The receiving surface of the sensor is, probably, 5mm in diametrer. But we need to detect 1mm strips that are 1mm apart. There will still be some variation in the outptu of the sensor, although not the clear pulses, as if the strips were wider then the sensor.

- Kender

Hi Kender

As it seems bigger the slot in the sensor is just 1mm (I personally checked it ) in MOC7811. I think you may also get smaller aperture with smaller devices.

Regards
Nandhu