Multi-channel High Frequency PWM with registers and phase shift capabilities

[HermannSW]

vcgencmd "measure_clock" is really helpful in determining Pi Zero PWM clock. After reboot it is 0, after enabling and different settings it gives the current clock. For 1 not 9.6MHz but 5KHz gets selected, that seems to be the reason why tc_lib was not able to determine the period:

Code: [Select]

pi@raspberrypi01:~ $ vcgencmd measure_clock pwm
frequency(25)=0
pi@raspberrypi01:~ $ gpio pwmc 0
gpio: clock must be between 0 and 4096
pi@raspberrypi01:~ $ gpio pwmc 4096
gpio: clock must be between 0 and 4096
pi@raspberrypi01:~ $ gpio pwmc 1
pi@raspberrypi01:~ $ vcgencmd measure_clock pwm
frequency(25)=5000
pi@raspberrypi01:~ $ gpio pwmc 2
pi@raspberrypi01:~ $ vcgencmd measure_clock pwm
frequency(25)=9600000
pi@raspberrypi01:~ $ gpio pwmc 3
pi@raspberrypi01:~ $ vcgencmd measure_clock pwm
frequency(25)=6400000
pi@raspberrypi01:~ $ gpio pwmc 4
pi@raspberrypi01:~ $ vcgencmd measure_clock pwm
frequency(25)=4800000
pi@raspberrypi01:~ $ gpio pwmc 4095
pi@raspberrypi01:~ $ vcgencmd measure_clock pwm
frequency(25)=5000
pi@raspberrypi01:~ $ gpio pwmc 3590
pi@raspberrypi01:~ $ vcgencmd measure_clock pwm
frequency(25)=5000
pi@raspberrypi01:~ $ gpio pwmc 3589
pi@raspberrypi01:~ $ vcgencmd measure_clock pwm
frequency(25)=6000
pi@raspberrypi01:~ $

Hermann.

[HermannSW]

Got a 100MHz logic analyzer today, and used it to verify the 26.21MHz measured averaged in previous posting. The logic analyzer measured 26.26MHz, quite close to the 26.21MHz measured via tc_lib. Details in this posting:
http://forum.arduino.cc/index.php?topic=417851.msg2888554#msg2888554
(right click for detailed view)


Hermann.