This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Trouble in using serial port communication bettween EVMDM 6437 DVDP and linux redhat as4

Other Parts Discussed in Thread: CCSTUDIO

Hi,

    I use the serial port to communicate bettween linux redhat as4 and   EVMDM6437 DVDP. The board sent a data eg. 99 to the linux system, but linux system  received is not 99. Except, while the linux system sent a data eg. 99 ,  the  board received garbled .    

      Thank you very much!

  • At first glance, it seems the baud rates are not aligned between the DM6437 EVM and your Linux Redhat machine.

    Just to verify things, have you attempted another terminal emulator program like HyperTerminal or TeraTerm on a Windows machine as a sanity check?

  • Hi

                Thanks for your reply ! 

                At fist ,  the baud rates are aligned (the  DM6437 EVM ' s baud rates is equal to Linux Redhat machine  's  , namely  115200 ) bettwen the  DM6437 EVM and  my Linux Redhat machine  .  During the work ,when I found this problem, I tried changing the other baud rates to make it  work . And the results are the sanme .Except  that ,I didn't use the DM6437 EVM ,using the serial port  to communicate directly bettwen windows system and linux sysytem.The results is ture.So I don't kown what's wrong with them.

                I'm looking forward your reply !  Thanks very much !

  • I'm sorry, but it is difficult to ascertain what you indicated in your experiment with connecting the linux Redhat machine with the Windows system.  Did this work?

    Your message seems to indicate that is didn't.  Unless I am reading it incorrectly.

  • Hi,

           Yes,in the work,I used the serial port to communicate bettween linux system(kernel2.6.9) and DM6437 EVM(CCS3.3) with the DM6437 EVM linked to the windows xp system (CCS3.3) by USB. The programme I used in DM6437 EVM side is its' system's  (C:\CCStudio_v3.3\boards\evmdm6437_v2\tests\uart).

         During the communication, (1) linux sysytem sent some data(e.g.,int 90,97 or 99) to  DM6437 EVM ,and the data that DM6437 EVM received showed in windows' screen is the messy code. (2) DM6437 EVM sent some data(e.g.,int 10,90,97,or 99) to linux system, the data that linux system received is correct value, but the data that windows' screen(CCS) showed is the messy code. (3) DM6437 EVM sent some data of Char or String to linux system, linux system received its ASCII, but windows' srceen (CCS) also displayed messy code. 

         So it's difficult to me to understand.And I am look forward your reply!

         Please do me a favor! Thanks very much! 

     

  • Are you referring to the board level UART test found on Spectrum Digital site (http://c6000.spectrumdigital.com/evmdm6437/reve/ ).  If so, it appears this is a loop back test, which means you had to make some modifications to make it talk to a PC?  Perhaps if you could share this CCS project, we may be able to provide more advice.

  • Hi,

                            Now I want to send my personal pragramme to you ! Can you provide me your email ? And now in the work , EVMDM 6437 DVDP  can't  be linked to wondows system (CCS). Is there anything wrong with chips in the EVMDM 6437 DVDP ?

                           Thanks very much!

                              

  • If you do not mind everyone else seeing your program, you can post it as an attachment.  Otherwise, you can send it in thru www.ti.com/support with a brief description of your issue (or perhaps a link to this post thread)

    Jason Yang said:

     And now in the work , EVMDM 6437 DVDP  can't  be linked to wondows system (CCS). Is there anything wrong with chips in the EVMDM 6437 DVDP ?                          

    Are you suggesting you cannot even connect to DM6437 on EVM using CCS anymore?  If so, I guess a bad chip is always a rare possibility; however, in our experience this is more often than not a software issue (JTAG drivers or CCS Setup).

  • 6724.uart.rarHi,

            OK. Last time I have sent an email to your TI support . And now connecting to DM6437 on EVM using CCS only is OK without serial port connected, but when connecting serial port bettween EVMDM 6437 DVDP and linux system, it is failed to connect.This is the programme in linux system, 7776.uart2ti.rar.This is the programme in the CCS8030.uart.rar

          Using  PCI also has the same problems,it means that it is failed to connect bettween EVMDM 6437 DVDP and linux system.

           Thanks very much!

  • Hi,

           We have not communicated with each other for a long time. And can the programmes work in your test? Is there anything wrong in the hardware or the software? Please do me a favor !

           I wait  for the results with bated breath!  And thanks very much!

  • Mr Juan Gonzales :

              Did you receive and help me test the programme which I sent to you on 05-22-2009?  Is there anything wrong with the hardware or software?  And now I have to stop my work because of  those problems. So I  look forward your reply eagerly ! Please do me a favor !

               Thanks a lot !

  • Jason,

    Unfortunately, I have not been able to do this yet; primarily because I do not have a DM6437 EVM, or the necessary setup to do this.  I will need to work with Bernie (who is also very active on forums) to see how we can move forward with this.  Thank you for your patience and understanding.

  • Mr Juan Gonzales,

              Sorry for inconvenience. And please do me a favor.

              Thanks a lot!

  • I tried setting up your project and I got everything to build and run but I am seeing failure in the DSP test, it looks like it is never receiving anything, the PC side program seems to run indefinitely but gives no indication of error. It kind of concerns me that the DSP side example looks to be just the Spectrum Digital board validation test but with the PC continuously spraying out data (I think) I would expect the receive side of the example to get something. Based on the behavior I suspect something is wrong with the transmitting code from the PC though I have not taken an oscilloscope to the PC output to see what it is actually driving if anything.