@Robin,
Thank you, very brilliant advice.
@chopsuwe
/////
If you use two sensor strips, side by side and separated by a known distance, you can calculate the angle of the board. E.g. The first left hand sensor covers half a second before the first right hand sensor. If the sensor strips are 30 cm apart and the board is travelling at 1cm per second that gives a right angle triangle with sides of 0.5cm and 30cm. Now you can use trigonometry calculate the angle of the board on the conveyor, and use the speed make an allowance for the angle in your width calculation. ////
Sorry, I am confused.
If the sensors are 30 cm apart and the conveyor speed is 1cm per second, I don’t see how the first left hand sensor covers 0.5 s before the first right hand one.
Cant see the right angle triangle neither.
Thanks
If you can make a sketch.
abdelhmimas:
Sorry, I am confused.
If the sensors are 30 cm apart and the conveyor speed is 1cm per second,
If the angle of the boards really can confuse the distinction between boardA and boardB then you could use the scheme I suggested in Reply #39 with two rows of sensors.
Sample all the detectors when the first X sensor is covered. Temporarily determine the board size using the number of covered sensors on the same side as the triggered X sensor. Then check the coverage of the other sensors to get an idea of the angle and perhaps make an adjustment to the board size. I think I would do some tests to see how the angle affects the determination of board size and build that into a table in the program. That way there is no need for complex maths.
It seems to me that 30cm separation would be far too little if you are dealing with 8ft boards. 1 metre or 3ft seems more sensible to me.
Another thought is to organise the flow of boards so they are never skewed so much that simple width detection fails - I think that would be my #1 approach.
...R
@robin
Thank you.
I am not worrying about the boards arriving at an angle.
Mechanically controlled and I have a skewed boards alarm up the line.
Then it looks like Robin's answer #39 will do the job. Nice and simple too.
Do you have any information about how it's done in other sawmills?
@wildbill
If you are talking about the boards skewed,
Every board is tracked down the line from the optimizer to the sorter using an encoder and different sensors.
Right after the saw, there is the skew photoeye, when it detects the board, the PLC compare the estimated encoder count to see the board with the actual one, if they are way different, it means board is skewed and the line is stopped.
abdelhmimas:
@wildbill
If you are talking about the boards skewed,
Every board is tracked down the line from the optimizer to the sorter using an encoder and different sensors.
Right after the saw, there is the skew photoeye, when it detects the board, the PLC compare the estimated encoder count to see the board with the actual one, if they are way different, it means board is skewed and the line is stopped.
With an encoder and a PLC already installed why bother with this Arduino angle? Just add a PE and the encoder count and some ladder logic to do the width check.
dougp, I agree, a PLC with a high speed count card to read the encoder for monitoring conveyor
speed and a photo-eye input for sensing part present would be straight forwards in ladder using
timers. the leading edge of the input starts a resettable millisecond timer, the trailing edge stops it,
a 2" board at 60m per minute would equal 20ms, 3" would be 27ms, and so on(or something like that).
but I would go with abdelhmimas's Arduino or add-on PLC approach so production does not get
interrupted to prove the concept.
I would first try a simple long armed roller type limit switch for sensing part present, phot optics is the wrong
technology for the woodworking industries.
abdelhmimas:
@chopsuwe
/////
If you use two sensor strips, side by side and separated by a known distance, you can calculate the angle of the board. E.g. The first left hand sensor covers half a second before the first right hand sensor. If the sensor strips are 30 cm apart and the board is travelling at 1cm per second that gives a right angle triangle with sides of 0.5cm and 30cm. Now you can use trigonometry calculate the angle of the board on the conveyor, and use the speed make an allowance for the angle in your width calculation. ////
Sorry, I am confused.
If the sensors are 30 cm apart and the conveyor speed is 1cm per second, I don’t see how the first left hand sensor covers 0.5 s before the first right hand one.
Cant see the right angle triangle neither.
ThanksIf you can make a sketch.
You said you were concerned with the boards being at an angle. So use two sensor strips to determine which one the board hits first.
In the diagram below, the sensors in each strip are placed 4 dots apart. The board is just about to cover sensor 1 on strip B. And it is still one dot away from covering sensor 2 on strip A. That's a total 5 dots behind which equates to 1.25cm.
---------conveyor side--------------------------
sensor strip A 1....2.//.3....4
//
//
//
//
//
sensor strip B 1//..2....3....4
---------conveyor side--------------------------
If the two strips are 30cm apart your triangle looks like this
---------conveyor side--------------------------
sensor strip A 1|.1.5cm.//
| //
| //
30cm | //
| //
|*//
sensor strip B 1|//
---------conveyor side--------------------------
The angle you are looking to find is the corner with the *
@chopsuwe
I see now, when you were talking about right and left hands you meant conveyor opposite sides. I though they were in the same side of the conveyor.
I don’t think that I ever said was concerned by the boards skewed.
This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.