Hi
When using the ETPU channel to set the PWM output duty cycle to 0 or 100%, there is about 0.19% duty cycle.I would like to know what factors are involved in this phenomenon and whether it is related to the MPC5634 chip itself.
Hello,
Are you using the standard NXP PWM eTPU function? To update duty cycle are you calling the fs_etpu_pwm_duty() API? What is your timing configuration - system and eTPU clock frequency, TCR1 counter frequency, etc.? The standard PWM function is designed to output all the way from 0% to 100% duty cycle, so you should not be getting small pulses if commanding full 0 or 100%. I did a simple test in simulation with the standard PWM function, initializing one instance on a channel and driving 20%, then 100%, 0% and finally 50%, and it worked fine - see below. But of course there are other potential variables - see questions above.
Hello,
I downloaded a build of the PWM eTPU function from the ASH WARE eTPU Code Builder and then got it up and running in simulation using your configuration - MPC563xM, 80MHz system, TCR1 at 10MHz. All looks fine running 0 to 100% duty cycle. I also downloaded a build from the CodeWarrior eTPU Function Selector and as expected that works the same.
This code:
/* initialize eTPU */
if (my_system_etpu_init())
return 1;
/* start eTPU */
my_system_etpu_start();
/* initialize a PWM output at 20 % duty, 20kHz */
fs_etpu_pwm_init(ETPU_PWM_CHAN, FS_ETPU_PRIORITY_HIGH, 20000, 2000, FS_ETPU_PWM_ACTIVEHIGH, FS_ETPU_TCR1, etpu_a_tcr1_freq);
at_time(1000);
fs_etpu_pwm_duty(ETPU_PWM_CHAN, 10000); /* 100 % , 20 kHz */
at_time(2000);
fs_etpu_pwm_duty(ETPU_PWM_CHAN, 0); /* 0 % , 20 kHz */
at_time(3000);
fs_etpu_pwm_update(ETPU_PWM_CHAN, 10000, 10000, etpu_a_tcr1_freq); /* 100 % , 10 kHz */
at_time(4000);
fs_etpu_pwm_update(ETPU_PWM_CHAN, 10000, 0, etpu_a_tcr1_freq); /* 0 % , 10 kHz */
at_time(5000);
fs_etpu_pwm_update(ETPU_PWM_CHAN, 10000, 5000, etpu_a_tcr1_freq); /* 50 % , 10 kHz */
Produces this output:
I've attached my project for you to look over - you should be able to use the etpu_gct.[c,h] files and other pertinent source to try it in your setup, or at least check out code differences.
Hello,
Are you using the standard NXP PWM eTPU function? To update duty cycle are you calling the fs_etpu_pwm_duty() API? What is your timing configuration - system and eTPU clock frequency, TCR1 counter frequency, etc.? The standard PWM function is designed to output all the way from 0% to 100% duty cycle, so you should not be getting small pulses if commanding full 0 or 100%. I did a simple test in simulation with the standard PWM function, initializing one instance on a channel and driving 20%, then 100%, 0% and finally 50%, and it worked fine - see below. But of course there are other potential variables - see questions above.
