Hey all,
I am involved with the Robocup Junior program and I just have a few questions for uall.
First, we are looking to replace or improve our compass system so that we have more accurate readings on our heading. At the current time we have 2 problems with the current compass (CMPS03). First is the magnetic interference that occurs at points around the field. Second is the fact that it works off an i2c interface and we use an MIT Handyboard which does not have i2c. This means we have had a convertor from i2c to analog voltages built for us, but it isnt very good (gives 70 points of differentiation roughly) which is far from what we desire.
We have seen the vector2x compass which looks better suited for our needs but due to limited time and money we would rather come up with a better solution using the CMPS03. Anyone have any ideas on how to better use this compass with the handyboard?
Second question. If we chose to scrap the compass to avoid the interference issue, is there any other sort of sensor that would be able to let us set a point (that we would call north...the goal that we are aiming for) and be able to tell where it is facing in relation to that point and rotate back toward it? We have looked at gyro's and some other systems but have not found anything that fully suits wat we need yet. It isnt really possible to use encoders on the wheel as we have an omni directional drive system which makes it hard to measure the direction that we have been driving in, so dead reckoning isnt really an option as far as I can see. Any ideas on this would be good.
If anyone has a good method of distance measuring I would also be interested. Again, we are using SRF08 ultrasonic modules but due to the i2c interface we have them hooked up to the convertor as well. We have found that ultrasonic readings are not very constant (tend to waver a bit) and are also not as accurate as we would hope. If anyone has a method that does not emit infra-red light (against the rules) I would be very interested to hear it. We have looked at a colour sensor to find where we are in relation to our goalie which would be coloured brightly, and also in relation to the grey opposing goal, but the colour sensors are too expensive for our use.
Has anyone had experience making a mumetal shield? We were looking at this solution to prevent some magentic interference getting to our compass from the floor.
Thanks for your help
Jeff