So unfortunately, I’m still waiting on my batteries to arrive meaning I have to make do with some extra Li-Pos I happen to have lying around (…as you do).
I gave a demonstration to one of my lecturers – soon to be my project supervisor over the summer, and he showed an interest in a minimal implementation with the two front facing proximity sensors so I might be providing him with a parts list and schematic at some point.
This is about the most gutting part of this project so far! I was wondering why my SwarmSense module wasn’t responding to modulated light and as it turns out, I’d used the wrong opamp package in eagle with the power and ground on the completely wrong pins argh! That deems all of my SwarmSense boards as useless, along with wasting some of my precious phototransistors (around £1.30 worth). This is really annoying as it means I can’t actually test the main reason I initially thought up these robots!
On the plus side, I’ve now prototypes a handheld controller using the portable NRF24L01 library and my new (oooh) STM32F0 discovery board. I’m using a PSP joystick as the input and it seems to do the job at the moment though it’s a little dusty and scratchy. I think my servos however aren’t particularly well calibrated as they sometimes drift leading to me needing to change the thresholds within software as I can’t reach the trimpots now.
For future revisions, I’m going to have two front mounted proximity sensors so the robot can see on which side the obstacle is and choose which direction to turn. Along with this, I’ll probably include reflective optical sensors for the wheels so the robots can have an idea of how much distance they’ve covered.
For the SwarmSense module, I’m probably going to change from an inverting amplifier with specially selected capacitors and resistors – used to give a bandpass response, to a proper MFB bandpass filter due to the better frequency response.
Different responses of RC configurations