Hello all,
I am looking for a linear sensor for my project that would measure the position of a 3mm rod (end position) that moves fore and aft pretty fast (max range of 30mm, typical range of 10mm and frequency of 10Hz).
Idea is to measure the position, use it to move a servo to move another rod.
Where should I start? rotary or sliding? anything better? (newbie here).
Thanks,
Nr
I would start by searching the web for "linear position sensor". Sensor lifetime may be an issue.
10Hz is not fast. That should be easy to do with basically any sensor technology.
How much drag can you tolerate? Can you push a linear pot with the 3mm rod?
If drag must be near zero then an optical distance sensor - like the ones made by Sharp and available at Polou, Sparkfun, Adafruit and everywhere else - is actually a pretty good option. Put a reflector target on the moving arm and mount the sensor at a convenient position nearby.
A magnetic or hall-effect sensor could work, but I don't have any immediately to hand to recommend for that range of motion.
For high precision, a vernier scale with a serial output can be purchased surprisingly cheaply and give you accuracies better than 1/10000th of an inch.
Thanks for the replies.
Going through the different ideas:
*/ pots; I am a bit concerned at the cycle life; quite a few datasheet mention a total cycle < 100000 and around 1000 cycles per hour. 10 Hz give 36000 cycles per hour.
*/ optical sensors; I hadn't mentioned it, but I am a bit tight in space; the smaller Sharp starts from 4cm; but they do seem to need a fairly big 'target'.
*/ hall effect sounds interesting; one example here Using a magnetic hall effect sensor to see distance on Vimeo for distance. Just wondering about accuracy.
Nr
nr765:
*/ hall effect sounds interesting; one example here Using a magnetic hall effect sensor to see distance on Vimeo for distance. Just wondering about accuracy.
You should not assume that the output from the hall effect sensor is "linear" as stated in that video. However if you can move the rod to different stationary positions, you could calibrate and then use interpolation or curve-fitting within your Arduino code to get a more accurate measurement of the rod's position.
While tiny neodymium magnets are readily available, I would be a little uneasy fitting one to the end of a 3mm diameter rod.
I suggest you consider placing an insulated metal tube around the rod and measuring capacitance one way or another. I would expect that to be fairly linear.
what about softpot? life cycle seems to be fine, cheap and accurate?
Re: "linear" hall effect sensors may have an output that is a linear function of magnetic field strength, but the field strength is not a linear function of distance.
How about an LVDT? The only wear is on linear bearing surfaces. And they can be highly accurate and highly linear. The resolution can be arbitrarily high.
You can even roll your own.