Bannner IR sensor Measures Ball Exit Velocity

[Dale(294engr]]

The banner IR sensor has a range ~1' using blue/green poof balls

(IR doesn't reflect from balls with a black half, Visible LED type may - TBD)

I conducted a test yesterday (prelim) using a 1+ meter AL pole over the ball exit path with 2 sensors attached directed downward,
1m apart
black output lead pull up = ~2.2mA, (produces neg. going ball pulse)
5.6k to 12v or 2.2k to 5v)

Oscilloscope used for time between their falling (leading) edges.
83.3ms = 12m/s (1/.0833)

NOTE: In lieu of Oscilloscope (or external electronic timer)
RC may be used with software to measure time:

between two sensors using 2 input ports (above)

or one port to measure pulsewidth for Simple method (below)

Example: measured 60ms on 12.5" skyway
direct driven by lg CIM at 12.6V (~2500 rpm)
so 1/.06 = 16.7m/s

The 2 sensor method eliminates
ball dia. and edge dropoff errors at cost of 2nd sensor
by measuring leading -edge start to leading -edge stop times

-----------------------

Simple 1 sensor method: (assumes sensor sees full 7" ball dia):
assumes 7" ball diameter =measured path leading to falling edge = pulsewidth

Also requires blue/green balls (IR reflection is not reliable with BLACK balls)

the Velocity in m/s is 1/(5.624*pulsewidth_sec)

1 / ( (39.37"/m /7"ball)*pulsewidth_sec)

Example: pulsewidth of 15 ms = 11.85m/s

12m/s = 83.3 ms/m /5.634 = 14.8ms for 7" ball path

realistically sensor likely senses a bit short of 7" due to the sphere edge slope. Measure the test ball(s) dia + Perform Edge detection test for an accurate dia. to use for better accuracy -- 6.9 or ?


note
select the white sensor output w/pullup R to +V for positive going pulses

Dale.Hall@earthlink.net
Team 294 Redondo Beach, CA

note:
m/s meters/sec
ms = millisec (1/1000 sec)

[Gary Dillard]

Do you think this would work using the banner retro reflector and sensing the ball as a blocker rather than a reflector? Then it wouldn't be sensitive to the black side of the ball.

[Dale(294engr]]

Gary,
A separate emitter & detector would eliminate the black IR absorption problem
at the cost of a bit more complexity,

i.e. two wired & optically aligned devices per location
(+ may need to shield ambient w/ tube or use modulation)

(1 loc pulse width of 7" ball or
2 loc's for distance traveled reducing error from dia & beam pos on ball)

Some teams may only have balls that are half black.

then I suggest the simpler method, spray paint test ball (black half or all)

Verified - works/repeatable:
simpler: one device position per loc, self aligning / shielding (sensivity adj)
I set up this test in 15 min w/old real time scope (DSO or storage better)
I was impressed with the results of two sensor time repeatability and

promising 1 sensor approach using 7" ball width to measure exit velocity (pro: sensor close to exit, con: error if beam not centered on ball) looks very promising, i.e. "simple as possible but not simpler" A.E.

??? re: retro-refl
Our IR & Vis Banner Sensors contain emitter & detector & range adjustment in one compact package and rely on retro reflected beam.

Broken beam approach requires physically separate Emitter & Detector & likely a hood (or modulated?) and good stable optical alignment

Dale(294engr]

[Gary Dillard]

Thanks. We are playing around with this as well. The IR sensor is listed as "diffuse" and the visual red sensor is listed as "retro-reflective", but the wavelengths aren't much different between the 2 sensors while the range is significantly different, so I wasn't sure if there was some difference in design that would make them behave differently.
I wasn't sure if the detection spectrum was sufficiently different that it would react better to a specific retro-reflector target while ignoring a different diffuse target (the ball) by tuning the sensitivity.

[Dale(294engr]]

Originally Posted by Gary Dillard

Thanks. We are playing around with this as well. The IR sensor is listed as "diffuse" and the visual red sensor is listed as "retro-reflective", but the wavelengths aren't much different between the 2 sensors while the range is significantly different, so I wasn't sure if there was some difference in design that would make them behave differently.
I wasn't sure if the detection spectrum was sufficiently different that it would react better to a specific retro-reflector target while ignoring a different diffuse target (the ball) by tuning the sensitivity.

I omitted this Banner IR sensor option:
Either sensor may be used with a reflector, i.e blocking mode. IR sensor is effective > 1' effective to detect all the ball colors, without separate Emiter / Detector (re: my earlier reply - just need to position reflector and look for loss of signal in code.)