Robobuild

Ok, so the Robobuild session this Wednesday was not quite as well attended as i'd have hopped, but nonetheless useful. It was the first chance i had to play with the full kit and prove to myself that it actually works. After some stupid errors induced only by my lack of homework (editing the wrong python file) i was in control of the motors. The servo was next and that too worked. I learned that the PWM board continually sends out the position signal to the servo, regardless of how long ago the setpos() command was issued. Following this we plugged in the IO board and attempted to flash all the digital outputs on and off. This would work for an unpredictable amount of time before it would cease all activity.

With Rob's help and the MSP Spy By Wire the fault was found. To safeguard against the IO Board 'hanging' if an error occurred on the I2C bus, a timer interrupt in the MSP was enabled at the start of data reception and disabled at the end. If an error happened the timer would interrupt and kill the I2C communication, freeing the bus. However the timer was not being correctly disabled and this was causing the problems.

By the time we fixed this bug it was half four so we retired to Zepler to hammer out the budget. We didn't get very far. In the main meeting as well as forcing everyone to introduce themselves whilst wearing Aron's green hat, we covered all the improvements to the hardware. When broken down it doesn't seem half as much. I'm excited about the screen module, especially if we use an OLED screen. The idea of allowin students to add breakpoints in their code and display appropriate outputs on the display would be really helpful.

Hopefully I'll have the budget - for the electronics- hammered out before the next meeting.