• Hardware
  1. Electrical schemes
  2. Printed Circuit Boards
  3. Body design
• Software
• Errors

Errors

The main errors we encountered during the project  mostly came from the electronics part:

• We made a first electrical scheme for the greyscalesensors to eliminate the offset and amplify the signal, but it didn't work, although all resistors and other connections were right. By measuring the voltage on strategical points, we noticed that it were the opamps that didn't produce a negative voltage. This was something we had overlooked when we ordered these opamps. We fixed it by making a new scheme that didn't require a negative voltage.
  
  
• We made a seperate PCB for the IR-ball detection sensors and distance sensors, with opamp schemes for elimination of the offset and amplification of the signals. However, the amplification and offset regulation did't produce any output. This was strange because we used about the same schemes for the greyscale-sensors, which did work. We took a lot of time trying to find the cause of this problem, but eventually we connected  the sensor signals directly to the microcontroller inputs. This was acceptable because the fototransistors produced a signal between 0,1 and 4,3 V and the distancesensors a signal between 0,1 and 3,2 V.
  
  
• During the last day before the deadline there was a problem with one of the H-bridge chips. It could only produce current in one direction and no longer in both, so one of the wheels of the robot couldn't turn forwards anymore. This had to be an internal malfunction of the chip because it was able to spin the motor  in one direction.

Inventor drawing:

An important remark for the future, especially for the students who will use the 3D-printer is to think about the price!! For 1 cm2 it cost 1 euro, that’s a lot. Our first design, cost 760 euro! So we had to change a lot of our design so eventually we got a design of 385 euro.