Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
17 years ago

can't scan jtag chain in customer built board

I have a custom built board with EPM570T100C3N, this is the only programmable device in the board. The other chip is a CY68013A. The USB part of the board works well, but the CPLD cannot programmed by Quartus software.

Each time I want to program it, the Quartus tell me "unable to scan device chain, can't scan JTAG chain." I have checked the power, and the connection, found no error.

I had tried with a ByteBlaster cable, and a USB-Blaster clone.

Can anyone give me some clue to solve the problem ?

8 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    --- Quote Start ---

    I have checked the power, and the connection, found no error.

    --- Quote End ---

    But the JTAG hardware is apparently wrong anyway.

    Check with an oscilloscope, that each input signal (TCK, TMS, TDI) is arriving at the respective pin. If so, the TDO output must also show a response in initial chain test.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    --- Quote Start ---

    But the JTAG hardware is apparently wrong anyway.

    Check with an oscilloscope, that each input signal (TCK, TMS, TDI) is arriving at the respective pin. If so, the TDO output must also show a response in initial chain test.

    --- Quote End ---

    I don't have a oscilloscope, the only thing I have is a multimeter.

    I have verified the byteblaster and usbblaster cable in another board that have MAX7000 CPLD on it. Both of the cable are fine. Here is the power and jtag part of the schematic. Is there something wrong ?
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    It's pin-by-pin identical to a EPM570T100 production circuit, that I designed. Did you also verify the correct pin numbering (and mounting position) of the TQFP100 package?

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    --- Quote Start ---

    It's pin-by-pin identical to a EPM570T100 production circuit, that I designed. Did you also verify the correct pin numbering (and mounting position) of the TQFP100 package?

    --- Quote End ---

    I found one of the problem, a 10K Ohm resistor（R6 in schem) was misreplaced by a 1Kohm resisitor. After use the right value resistor, Quartus successfully recognize the CPLD. But in most of the time, the programming will failed during the "programming device" period after blank checking.

    The length of tracks (TCK/TMS/TDI/TDO) are approx. 35mm. Are those tracks too long ?
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    I don't see a problem. TCK pulldown of 1K should also work, it's a recommended value for some FPGA. USBBlaster is able to drive 1k load as well. 35 mm trace length is rather short. There's most likely another problem like bad solder joints or unstable supply voltage.

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    After check the "enable real-time ISP", I can programming the device. It is really strange. I don't know why the programming will fail continuously if "enable real-time ISP" were not checked.

    :)
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    An interesting phenomenon indeed. Real-time ISP basically means, that the existing MAX II design is not halted during programming. Thus in your hardware design, something seem to affect MAX II programming, if the design is halted, sounds odd. It may be a power supply related problem or due to some special hardware connected to the MAX II. Not a common problem, I think.

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    17 years ago

    --- Quote Start ---

    An interesting phenomenon indeed. Real-time ISP basically means, that the existing MAX II design is not halted during programming. Thus in your hardware design, something seem to affect MAX II programming, if the design is halted, sounds odd. It may be a power supply related problem or due to some special hardware connected to the MAX II. Not a common problem, I think.

    --- Quote End ---

    Thank you FvM. I think it is not a common problem too. I build another board, the same to the first one. The same problem occured. Quartus cannot programming the device. But "Autodect" can list the jtag chain contains only one EPM570 device.