A week ago I purchased a Rigol 1054Z oscilloscope. Since then I've had fun measuring all sorts of stuff: bounce in switches, the PWM outputs of my Pi Pico, voltage spikes in a 555 astable timer, etc.
I decided to visualize how my Raspberry Pi Pico controls a 28BYJ-48 stepper motor. I already had code laying around, and I thought it worked fine. This was my first opportunity to use all four probes at once.
It was immediately evident that something was wrong. My "working" C code should have been bringing one input high at a time.
The happily spinning motor had misled me into believing my code was correct. The sadness is clearer when turning off two channels:
Squinting at my code, the culprit was obvious. I'd accidently shadowed the
int cur = 0; inside my while loop. This is also where I
discovered that the CMake example I'd used didn't include
-Wall by default.
// configure GPIO pins for output
uint pmask = (1<<IN1) | (1<<IN2) | (1<<IN3) | (1<<IN4);
gpio_init_mask(pmask);
gpio_set_dir_out_masked(pmask);
// initial energized state
int cur = 0;
gpio_put(pins[cur], GPIO_ON);
while ( true ) {
sleep_ms(3);
int pre = cur;
int cur = (cur + 1) % 4;
uint toggle_mask = (1<<pins[pre]) | (1<<pins[cur]);
gpio_xor_mask(toggle_mask);
}
I removed the errant int specifier. Lo, with the code fixed, the
scope looks like you'd expect:
This was my first real defect debugged with a scope.