Forum Discussion

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

JTAG Configuration problems

Hello,

I work for a company who uses altera chips in one of their products. I was recently shown how to use a JTAG chain to program one of our boards with 7 chips on it. The demonstration went off without a hitch, and the next day, I was expected to program some more boards exactly like the one demonstrated. The problem was when I went to program a new board, the program would not recognize the chips on the board. I plugged in a board previously programmed, and the software recognized all the chips.

I was taught how to program the boards using Max-PlusII v 10.23 with a ByteBlasterII. I also got the same results using Quartus II v 10.0 with a USB Blaster Rev. c.

I used the JTAG Debug tool on the Quartus software, and found that the software could not specifically ID the chips on the chain on a blank board. (they come up UNKOWN_7032DD or something similar) On the preprogrammed board, the chips are identified normally.

Note: this is a product we have been programming successfully for years, the circuit is exactly the same on the programmed board that works, as it is on the unprogrammed ones that don't. I strongly suspect I am missing something very simple in my inexperience with altera programming. Everyone at work with any expertise in this is unavailable this week, I have already spent an entire day trying to get this to work to no avail, so if anyone knows what I am missing, I would be grateful for any help you can offer.

7 Replies

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

    Can you probe the JTAG pins on the devices in the chain?

    For example, are all the devices BGA packages, with vias you can probe on the bottom?

    What OS are you using the parallel port ByteBlaster on? I have a jtag_tester program that can be used to toggle TCK/TMS/TDI that you could use to generate activity on those signals, and then you can probe with a scope to see where things stop working.

    Cheers,

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

    The devices are 44 and 84 pin PLCCs (altera MAX 7000 series) I could find the right pins from the schematic, but the package makes probing difficult at best.

    Both programs are running off of Windows XP.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    14 years ago

    I just tested the attached zipped .exe under Windows XP with a byteblaster II cable (no FPGA attached though). It still works: 

    
$ ./jtag_tester
Test application for JTAG chains
=====================================
Select JTAG interface:
=======================
1. Altera ByteBlaster
2. Cypress UltraISR
Enter option: 1
Windows NT/2000 version: 5.1.2600 (A280105)
ByteBlaster detected ok.
ByteBlaster enabled ok.
Test options:
=============
1. Toggle TCK
2. Toggle TMS
3. Toggle TDI/TDO
4. Check JTAG chain.
5. Count the number of devices.
6. Read out the device IDCODEs.
7. Exit.
Enter option: 6
Reading device IDCODEs:
11111111111111111111111111111111        <no device>
<hit enter to continue>

    You can use this program to toggle pins, read ID codes, etc.

    Perhaps it'll help debug your issue.

    I'd recommend first probing the TCK signals at each device and closely looking at the rising edges. If there is not a clean rising-edge due to transmission line reflections, then it will cause problems.

    Cheers,

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

    The TCK signal seemed to have a clean edge for both the programmed card and unprogrammed cards, although the tester program couldn't verify the chain or get an ID on either type.

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

    --- Quote Start ---

    The TCK signal seemed to have a clean edge for both the programmed card and unprogrammed cards, although the tester program couldn't verify the chain or get an ID on either type.

    --- Quote End ---

    Did you try it on the working board?

    Here's the tester code. It does the bare-minimum with the JTAG interface to generate correct transactions. It should work with your devices.

    Cheers,

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

    The problem has been resolved, what they failed to tell me was that this batch of boards was made with some bad chips. Replacing the affected chips solved the issue right away.

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

    Thanks for posting the resolution.

    Glad to hear the problem is resolved!

    Cheers,

    Dave