avoid table edge

Ben Damon

I was working on avoid table edge program in "robotics with the boe-bot" on page 259.· My problem was the boe-bot kept spinning and spinning.· I figured out that the left ir led was some how sending a stronger signal.· And the right did not see anything.· Should I do the Ir balance thing in the back of the book?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ben Damon



And may God Bless America!

Ben Damon

ok, part of the problem was solved. I had 8500 as the send out, instead of 38500. But the left is still stonger than the right. With the leds pointed down at the same angle, the right flashes on and off, while the left has a solid signal

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ben Damon



And may God Bless America!

Jon Williams

You have minor variations in components -- this is normal. What you need to do is "tune" your hardware to the environment (adjust things mechanically until bot sides respond the same under the same conditions). Doing edge sensors for long distance is tricky because the sensors can be affected by other IR sources.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jon Williams
Applications Engineer, Parallax
Dallas, TX· USA

Andy Lindsay (Parallax)

For table edge detection, you should be getting a solid 0 when your IR detector and LED are focused on the surface in front of the Boe-Bot. You should also get a solid 1 when the IR LED/detector is focused on the electrical tape course outline.

If one of the sensor outputs is kind of sketchy, try the following:

Regarding the tuning that Jon mentioned, make sure that the IR LED and the detector are both focused on the same point in front of the Boe-Bot. There will be two different focal points, one for the left IR LED/detector, and another for the right.

Check and see if the resistor in series with the IRLED has the wrong color code. It's easy to mistake brown-black-orange for brown-black-red.

Check to make sure none of the wires are shorting against each other.

Make sure that the left and right IRLEDs are both properly seated in their shields.

Try adding PAUSE 10 just before the second FREQOUT command.

Some table surfaces work better than others.· Try a sheet of white photocopy paper and see if that gives the right IR LED/detector a strong enough signal.

If none of the above works, try going back to the Your Turn section of Chapter 7, Activity #3.· Although it's not common, you may need a different series resistor for your right and left IR LEDs.· For example, maybe they will both have the same detection power if the resistor in series with the·right IR LED is 1 k (brown-black-red) while the·left needs 2 k (red-black-red).

If the intermittent results occur while the detector is focused on the electrical tape boundary but not on white paper, you may have IR interference from a nearby fluorescent light. Try IrInterferenceSniffer.bs2 from Chapter 7, Activity 2.

Post Edited (Andy Lindsay (Parallax)) : 3/22/2005 8:07:16 PM GMT

Ben Damon

I have changed up the resistors now. I changed the right to 1k resistors. And then I found out the left was too strong. and changed it to next step up 4.7k resistor.

This seems to work, the left is touchy and will turn at any little thing

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ben Damon

And may God Bless America!

Jonb

I did alot of "tuning" for my current project using IRLEDs.

My new method of harnessing their power is to coat them with liquid paper on the back and part way up the sides. Then cover with shrink tubing and shrink using a match. I found that while using very low power the beam could be modified by modifying the IRLED. Also keep in mind that there are various IRLEDs on the market with various emissions angles depending on your needs. Getting the LED to match perfectly and proper placement of the IR receivers is the key though. I also noticed that making a small shield(as the finished project will be enclosed) on the back of·the receivers helped with weird reflections, where an object on the right would be picked up by the left sensor aswell. I'm not using edge detection though, its a matter of fiddling with it till, as mentioned before, the focal point is where you want it.



▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Post Edited (Jonb) : 4/5/2005 4:54:51 AM GMT

Jonb

Another thing I forgot to mention, the problem with IRLEDs is that we can't see them. If you have a webcam, or digital camera with video you can view the beam in real time.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

glouvier1982

I'm trying to fine tune my bot to detect a 1/2 inch drop-off. However, when I get close to the range I want, I start getting a lot of phantom 1's and 0's (depending on where the bot is placed). Any ideas???

Franklin

I'm trying to fine tune my bot to detect a 1/2 inch drop-off. However, when I get close to the range I want, I start getting a lot of phantom 1's and 0's (depending on where the bot is placed). Any ideas???

Yes, start your own thread rather that latching on to one that is 5 years old.:idea:
Also post in robotics if the question is about robots.