Subject:
|
Re: GBC at BrickFest 2006
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 6 Sep 2006 19:47:50 GMT
|
Original-From:
|
steve <sjbaker1@airmailSTOPSPAMMERS.net>
|
Viewed:
|
4726 times
|
| |
|
|
Brian Davis wrote:
> In lugnet.robotics, Rafe Donahue wrote:
>
> > > detecting the ball itself via reflected light
> > > is a tougher issue...
> > I had success with a ball detector using only one light sensor...
> > Across from the light sensor, two studs away, was a silver brick,
>
> Now that is a nice solution, and not one I'd thought of. Not having the silver
> brick, I could still use a (non-LEGO! Aaah!) solution of a small pice of
> reflective tabe or something. Cool.
There are actually three ways to approach this:
1) Using something reflective on the far side of the ball chute -
have the sensors light bounce back into the sensor - so the
reading decreases when something gets in the way of the
reflected light.
2) Put an actual lego light brick on the far side of the ball chute
to shine light into the sensor directly.
3) Make the far side of the chute very dark so that very little of
the sensors own light is reflected off the chute - but when a
light coloured ball passes by, it reflects more light back into
the sensor than the chute does - so you see the light level
INCREASE as the ball goes by.
Which of these works best? Um...well, it depends.
The Lego soccerballs have black and white patches - so the amount
of light you get reflected back from them is quite variable depending
on which way they happen to be pointing.
In the end, you have three light level numbers to consider:
A = The level when there is no ball there.
B = The level when the ball comes by 'white side up'.
C = The level when the ball comes by 'black side up'.
What you need is the technique that gives you the biggest
difference between (A) and both (B) and (C).
If you choose (2) then block off the sensors own light source. This
(I believe) will give you the best chance since when the ball blocks
the light, there is no light bouncing off of the ball from the sensor
side of the chute - so it doesn't matter whether the ball comes down
white side up or black side up - or even if someone feeds you a
basketball (which are brown).
If you choose (3) then don't just use black bricks that are parallel
to the face of the sensor. Even though they are black, they are
quite shiney. You need to make that wall of the chute face off at
a slight angle to the sensor so that reflected light goes someplace
else and not back into the sensor. You might also experiment with
some black cloth (eg one of the Harry Potter MiniFig cloaks) because
those aren't as shiney as a black brick.
|
|
Message has 1 Reply:
 | | Re: GBC at BrickFest 2006
|
| (...) That's the method I used on my own ball counter. During testing, I fed it with (counted) hundreds of balls, it never missed (or over counted) one. During LW2005 it counted about 15000 balls a day during a week... some wear now but still (...) (18 years ago, 7-Sep-06, to lugnet.robotics)
|
Message is in Reply To:
| | Re: GBC at BrickFest 2006
|
| (...) Now that is a nice solution, and not one I'd thought of. Not having the silver brick, I could still use a (non-LEGO! Aaah!) solution of a small pice of reflective tabe or something. Cool. (...) I know that there's one power transformer that is (...) (18 years ago, 6-Sep-06, to lugnet.robotics)
|
28 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
