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.

CCS/TMS320F28335: Connection Problems with old exDsp and CCSv7.2

Part Number: TMS320F28335

Tool/software: Code Composer Studio

I'm having very confusing problem with two old ezDsp f28335 boards.  This happens on 2 different boards and 2 different development PC's. 

For a long time the board connected and downloaded code to RAM reliably.  Recently it has started getting a whole host of errors:

  • C28xx: File Loader: Verification failed: Values at address 0x08000@Program. This is at the very start of RAM. I don't think the memory is bad, if I move the program to say 0x9000 it does the same thing at the new address
  • Leave it sit for a while, and it might download and run correctly, for a while. It then get a "Target power fail" error, even though the board has not been touched, much less power removed
  • With CCS disconnected from the target, the PC seems to connect/disconnect repeatedly from the usb as evidenced by the up/down "kedunk" sounds produced. Even with no activity on CCS
  • When it fails, it often produces the message that "the selected processor (TMS320F28335) does not match the found processor ()"
  • Sometimes when it does download apparently properly, when I tell it to run it hits and "illegal instruction" trap immediately
  • Often when trying to download and run, when failing it gets "ADC Calibration not complete , check if device is unlocked and recalibrate." But I've never touched the passwords or security module, and it will occasionally run correctly for a while.
  • There are more symptoms, they all seem to be

I am really puzzled about this, but I'm starting to suspect some sort of usb driver/CCS/usb debugger connection problem.  I tried uninstalling the usb driver and letting windows reinstall it, but no change. It seems to suggest jtag connection problems at various points in the process.

I looked for updates to the drivers, but they are built into CCS now, Spectrum Digital's only work up to v3.1

Should this work? Does CCSv7 work with this ezDsp?  Is there any debugging I can do to further diagnose the problem?

Thanks much in advance.

Mike

  • Mike,

    >>I am really puzzled about this, but I'm starting to suspect some sort of usb driver/CCS/usb debugger connection problem.
    I think you are on the right track. In my experience these intermittent issues are caused by HW: either a USB cable or the connectors at either end of the USB connection. Another issue that used to happen was power supply failure on the target device/board - either from problems with the barrel jack connector, the plugpack or even internally to the board.

    By the way, CCSv7 works just fine with this board.

    Hope this helps,
    Rafael
  • Some additional information that has developed lately.

    To recap, this all started when 2 ezDsp F28335 boards (old) that had been running fine were used to try and run a life test. Each worked fine until they were run for a long time, on the order of 30 hours continuously. This happened sequentially, first one board was run and failed, then the second was tried and it failed the same way.  Failure means it could not connect with CCS or when it did, it would fail shortly there after.  Even plugging it in without connecting with CCS, you can here it sporadically disconnecting and connecting at random via the windows sounds.  Leaving it sit over night or a long time (not sure how long) would allow a connection and it would run for a while, on the order of a few to 10 minutes. 

    During one of these times, I managed to download flash code to the device.  Now after it has sat a long time, when I plug it in it will run for a while, then seems to just stop. It will communicate on the serial port to the pc and handle commands appropriately for a while, then it seems to just stop.  Might have thought my code did something bad, but these events also include the debugger disconnecting, and my code should not be able to do that.

    I'm about ready to give up on these boards and move on to a newer version.

    Mike

    Mike

  • Mike,

    >> I'm about ready to give up on these boards and move on to a newer version.
    I suspect this would be your last course of action.

    Before you do that, you could double-check supply voltages - maybe something started to go bad after many hours of operation. Schematics are available at:

    c2000.spectrumdigital.com/.../

    Also, since you are getting the "kedunk" sounds (brilliant description by the way), it indicates the embedded JTAG interface is bad. You may try to use an external JTAG debug probe connected to P1 (this has happened in the past with some boards I have here).

    Hope this helps,
    Rafael