Motion tracking: sensor options
Archiver
Posts: 46,084
Hi, First-time poster – just getting started in Stamp-based projects (I've
been more of a software guy in the past), have learned a lot from the group
already – thanks for a great resource!
For my next project, I want to try to measure direction and speed of opaque
plastic balls (about 4cm/1.6” in diameter) rolling along a flat surface.
[noparse][[/noparse]I.e., similar to a pair of photogates, but with detection of direction of
motion as well].
Looking to get the highest accuracy / precision / reliability possible while
keeping cost of components manageable for a hobbyist (i.e., <$100,
preferably less than $50). Of course, willing to consider more expensive
solutions if they really work well... [noparse];)[/noparse]
Key details:
Ball parameters (e.g., size, composition) cannot be changed.
Ball can be assumed to have started at a known location (the center of the
start of the track)
Ball can be assumed to be rolling in a straight line (and only one ball at a
time)
Max ball speed will be something like 6 m/s (approx 18 ft /s).
The width of the track needs to be at least 1 foot, but could be up to 2-3
feet. The length of the track will likely be around 6 feet.
Sensor location(s) can be known and fixed.
Track needs to be flat, but if there are options for sensors mounted in the
track that do not affect its surface, this is a possibility.
Sensor types that look promising:
--- Ultrasonic sensor (e.g., Devantech SRF04): mounted on side of track,
project sound waves across track, so it would report a fixed distance (the
wall on the other side of the track) until a ball went through its detection
'cone'. Will the small size of the balls be a problem? Is it possible /
useful to use multiple sensors in the same detection region to improve
performance?
--- Infrared ranger (e.g. Sharp GPD120) : mounted on side of track, beam
would be projected across track like a photogate. I know that the voltage to
distance conversion is nonlinear, but I'm wondering what kind of accuracy /
precision people have been getting with these for moving objects (estimates
I've read seem to indicate 1cm accuracy in reasonable conditions for static
objects). Another problem with infrared: 1 reading per 40ms = 25 readings /
s. Beam width is approx 2cm. At max speed of 6 m/s, the ball would pass
through the beam region in about 3ms, does this mean that the ball might
pass through the beam region without triggering it? Could mount several IR
rangers, perhaps alternating sides of the track in a staggered fashion,
would this help?
Interested in any suggestions about alternate sensor types to consider
(e.g., measure path by using a dense array of very small contact switches at
the end of the track??), or your feedback on the applicability of the
ultrasonic or infrared rangers to my application.
Feel free to email lengthy replies to me directly, and I will post a summary
of responses to the group.
Thx, Paul
_________________________________________________________________
Check out the coupons and bargains on MSN Offers!
http://shopping.msn.com/softcontent/softcontent.aspx?scmId=1418
been more of a software guy in the past), have learned a lot from the group
already – thanks for a great resource!
For my next project, I want to try to measure direction and speed of opaque
plastic balls (about 4cm/1.6” in diameter) rolling along a flat surface.
[noparse][[/noparse]I.e., similar to a pair of photogates, but with detection of direction of
motion as well].
Looking to get the highest accuracy / precision / reliability possible while
keeping cost of components manageable for a hobbyist (i.e., <$100,
preferably less than $50). Of course, willing to consider more expensive
solutions if they really work well... [noparse];)[/noparse]
Key details:
Ball parameters (e.g., size, composition) cannot be changed.
Ball can be assumed to have started at a known location (the center of the
start of the track)
Ball can be assumed to be rolling in a straight line (and only one ball at a
time)
Max ball speed will be something like 6 m/s (approx 18 ft /s).
The width of the track needs to be at least 1 foot, but could be up to 2-3
feet. The length of the track will likely be around 6 feet.
Sensor location(s) can be known and fixed.
Track needs to be flat, but if there are options for sensors mounted in the
track that do not affect its surface, this is a possibility.
Sensor types that look promising:
--- Ultrasonic sensor (e.g., Devantech SRF04): mounted on side of track,
project sound waves across track, so it would report a fixed distance (the
wall on the other side of the track) until a ball went through its detection
'cone'. Will the small size of the balls be a problem? Is it possible /
useful to use multiple sensors in the same detection region to improve
performance?
--- Infrared ranger (e.g. Sharp GPD120) : mounted on side of track, beam
would be projected across track like a photogate. I know that the voltage to
distance conversion is nonlinear, but I'm wondering what kind of accuracy /
precision people have been getting with these for moving objects (estimates
I've read seem to indicate 1cm accuracy in reasonable conditions for static
objects). Another problem with infrared: 1 reading per 40ms = 25 readings /
s. Beam width is approx 2cm. At max speed of 6 m/s, the ball would pass
through the beam region in about 3ms, does this mean that the ball might
pass through the beam region without triggering it? Could mount several IR
rangers, perhaps alternating sides of the track in a staggered fashion,
would this help?
Interested in any suggestions about alternate sensor types to consider
(e.g., measure path by using a dense array of very small contact switches at
the end of the track??), or your feedback on the applicability of the
ultrasonic or infrared rangers to my application.
Feel free to email lengthy replies to me directly, and I will post a summary
of responses to the group.
Thx, Paul
_________________________________________________________________
Check out the coupons and bargains on MSN Offers!
http://shopping.msn.com/softcontent/softcontent.aspx?scmId=1418