uart problem of 21569

Hi Bin,

From your mail I understand that you are using a serial terminal to send/receive data on UART. Can you please explain the issue in detail. Meanwhile, I'm attaching a simple code for your reference. This can be directly run on ADSP-SC589 or ADSP-SC584 Ezkit. Can you take this code as a reference and check if this could help?

char_echo.zip

Regards,
Anand Selvaraj.

I mean serial terminal periodically(100Hz) send 6 bytes to dsp,dsp send back 6 bytes to computer while it received 6 bytes.

the test result is:the number of bytes which dsp echo back is less than serial terminal send to dsp.hope you see what I mean.

Can anyone help me?

Hi Bin,

I'm looking into this and will get back to you ASAP.

Regards,
Anand Selvaraj.

Hi Anand,

Have you made any process?

Regards,

Bin.

Hi Bin,

Apologies for delayed response.

I had taken a look at your code. Looks like you are using STOP mode for DMA Write and Autobuffer mode for Read. Is there any reason for doing so? Regarding your issue, you mentioned 6 bytes are sent to DSP and then DSP sends back 6bytes. How many bytes are exactly received from DSP?

BTW, did you get a chance to look into the the UART echo back code available in the Board Support package.
<Installation path>\Analog Devices\ADSP-2156x_EZ-KIT-Rel1.0.1\ADSP-2156x_EZ-KIT\Examples\Power_On_Self_Test\common\source\uart_test.c

The BSP can be downloaded from the below link.
download.analog.com/.../ADI_ADSP-2156x_EZ-KIT-Rel1.0.1.exe

Please use this example code and let me know how you get through this.

Regards,
Anand Selvaraj.

Hi,

Does the CoreWrite/CoreRead funtion work in blocking mode or not?I want uart works in unblock mode,cause our software's real-time requierment is vrey high.that's why I using DMA mode for uart,hope you see.

Regards,

Bin.

Hi Bin,

Yes, the CoreWrite/CoreRead APIs are in blocking mode i.e. this API blocks until the complete data is received. It is recommended to use this mode for reading small number of bytes. On the other hand, DMA read/write API can be used for large chunks of data.

Regards,
Anand Selvaraj.