Forum Discussion

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

Cannot detect device with own JTAG cable

Hello,

I have just placed a MAXII on a custom board and am trying to program it with a ByteBlaster cable made following the usual schematics online.

The programmer reports:

Error: JTAG chain problem detected

Error: No device detected

Error: The TDO connection to the download cable might be shorted to GND or is an open circuit

Error: The TCK and TMS connections to the last device might have problem

I've checked the JTAG pins of the CPLD and TMS and TDI are both held high while TCK & TDO are held low. When I run the JTAG Chain Test from the PC:

-TMS does nothing (stays high)

-TDI I can see pulsing strongly

-TCK again strong pulses

-TDO I can see something but it is very weak (less than half a volt)

I've tried two power supplies (3.3v) same result on each. This JTAG cable has never acctually worked I tried it on a known working board and got nothing so I replaced the IC with a 74HC244 (originally a 74LS244) when I realised the LS required 4.5v+ to operate.

To make it just that little bit harder the development PC is a 64bit desktop so the device is bring programmed from an old laptop. (Figure of 8 lead - no earth)

I noted before that if I just kept pressing the Test JTAG Chain button it would sometimes find a 'device' whether anything was connected or not.

1. Is the odd output of TDO anything to worry about or is this some symptom of the cable?

2. Does anyone have any suggestions on how to do debug a JTAG cable?

6 Replies

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

    --- Quote Start ---

    -TMS does nothing (stays high)

    --- Quote End ---

    JTAG can't work without toggling TMS. Either the ByteBlaster is defective or you shorted TMS to VCCIO on your board.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    16 years ago

    Hi FvM thanks for your reply,

    I've checked the board and TMS isn't shorted. I am about to check the cable and wondered what is the worst case scenario of making a cable without the 74HC244 - i.e. just straight through?

    EDIT UPDATE: Right I checked the JTAG Cable and 1. it is sending the TMS signal and 2. It seems to be getting 3.3v out of thin air for the power supply for the 74HC244 which is meant to be off the board (which wasnt connected)

    Ive also found a broken track on the PCB on the TMS line and am patching that right now.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    16 years ago

    Hi

    You must have signal in TMS always than press the Test JTAG Chain button , if not you have a problem with your ByteBlaster. Where are you measuring the signals? You can disconnect the signal TMS from the board maintainning the power supply and measure the TMS just in the cable?

    Can you check the cable in another board?
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    16 years ago

    Hi acruvad.

    I checked the signals both on the board and on the cable. I could see the signal on the cable but on the board, which I now know must have been down to the broken track.

    I am still not sure about the cable though, as it didn't work earlier on a known working board even though I am pretty sure that was down to the wrong buffer IC.

    Ideally I would be able to find some guidelines on the cables beyond the schematics, things like acceptable noise levels (I have some 100mv waveforms appearing when nothing else is driving the JTAG pins), decoupling requirements and maximum length, etc.

    As you can probably tell this is the first time i've acctually needed to really think about JTAG.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    16 years ago

    --- Quote Start ---

    It seems to be getting 3.3v out of thin air for the power supply for the 74HC244 which is meant to be off the board (which wasnt connected)

    --- Quote End ---

    Not connecting VCCIO can be actually another reason, why the self elaborated "Byte Blaster" isn't working correctly. In my opinion, a HC244 isn't particularly suited for 5/3.3V level conversion. You should rather use a buffer, that allows to overdrive the inputs above the supply, e.g. VHC244. I've used a respective ByteBlaster replacement for years, see attachment, but I prefer USB Blaster performance.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    16 years ago

    Hi FvM, Thanks for those schematics.

    I isolated and patched the broken tracks. I also changed the decoupling capacitors in the cable. I tried to find a 74VHC244 but unfortunately I'd need to order one so for the minute I am still using the regular HC244.

    There was also a problem with the header on the board, but once that was fixed Quartus started detecting the device.

    I haven't tried programming yet. At the moment I am trying to get Quartus on my development PC to connect to the JTAG server on the programming PC, but all the JTAG tests check out fine so it appears to be working.

    Thanks for your help!