BF592 interrupt error trigger problem


Now I found a rather strange problem. At present, the input rate of my external sport audio data is 1M, and then I set the count in the code to trigger after a certain time, but I don’t know why it is always a little earlier. For example, I set the count to 29 16 bits. Normally, it takes 464us to generate an interrupt, but I found that a DMA interrupt has been generated at around 300us. If the interruption makes him stop, then I will confirm the data, if otherwise he stops, the data will still be transmitted.

Could you pls help me to check this probrem?

thanks !

  • 0
    •  Analog Employees 
    on Aug 28, 2020 1:02 PM 3 months ago in reply to Tom jackson


    Sorry for the delay in response.

    From your mail, I could understand that you are measuring the time for SPORT interrupt on ADSP-BF592 and you are not getting the expected results. Please correct me if my understanding is wrong.

    In order to assist you better on this, please get back on the below.

    1. How are you measuring the time?
    2. Can you please explain your count logic in detail?
    3. From which point to which point the time is measured? Say the start and end points.
    4. I guess the BF592 received audio samples from external source. Is it possible for you try using a simple SPORT transmit/receive code?
    5. Does this timing issue have any impact on your application? If so, how does it affect your application?
    6. BTW, are you testing this on an evaluation board or custom board? And, whether you are using VDSP or CCES?

    Anand Selvaraj.