AnsweredAssumed Answered

PWMTRIP interrupt in ADSP21992

Question asked by porter on Mar 19, 2013
Latest reply on Mar 26, 2013 by Mitesh

 

We are moving an existing system to ADSP21992 from ADMC331.

There is a difference in ADMC331 and ADSP21992 processors regarding PWMSYNC interrupt execution after PWM TRIP interrupt occurs.

Please refer to following information from both datasheets.

 

ADMC331.pdf(PWM shutdown)- A falling edge on the PWMTRIP pin provides an instantaneous, asynchronous (independent of the DSP clock) shutdown of the PWM controller. All six PWM outputs are placed in the OFF state (as defined by the PWMPOL pin). In addition, the PWMSYNC pulse is disabled and the associated interrupt is stopped.

.

ADSP_2199x_hardware_Reference.pdf(PWM Shutdown & Interrupt Control Unit):- A falling edge on the PWMTRIP pin provides an instantaneous, asynchronous (independent of the DSP clock) shutdown of the PWM controller. All six PWM outputs are placed in the OFF state (as defined by the PWMPOL pin). However, the PWMSYNC pulse occurs if it was previously enabled and the associated interrupt is, also, not stopped.

 

We are facing a problem in our system functionality due to this behavior in ADSP21992 and we do not want to continue PWMSYNC ISR execution after PWM TRIP interrupt. Is it possible to achieve this in firmware and how?

Moreover we observe that if pwm trip occurs while pwmsync is being executed, the time that pwmsync interrupt takes is increased significantly, to quantify, usually pwmsync interrupt takes 20uS but when pwmtrip occurs it takes upto 50uS sometimes and never less than 30uS.  The pwm trip interrupt has just  instructions.

Can any one tell us why pwmsync interrupt takes so long to execute?

and

Is it possible not to continue PWMSYNC ISR execution after PWM TRIP interrupt?

 

Thanks in advance.

 

porter.

Outcomes