Forum Discussion

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

Dangerous bug in DE2-70 demonstrations

If I'm not mistaken, there is a very serious bug on several of the DE2-70 demonstrations that could create bus contention in one signal. It seems to affect all NIOS based designs except the Control Panel.

The signal is AUD_ADCLRCK, which has a permanently enabled output enable. It should have instead, a permanently disabled output enable.

This is a bidir pin connected to the audio codec. The audio codec uses this signal as output or input depending on the configuration mode, master or slave. The audio code is configured as master mode, it will actively drive the signal.

There are also additional minor issues on the i2c controller. The i2c signals should be open drain, they are not. Also the frequency is slightly higher than the maximum allowed by the specs.

6 Replies

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

    Thanks vjAlter.

    Have you contacted teraSIC? Ideally they should provide an update, but failing that, could you make the fixed files available somehow?
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    18 years ago

    I am trying to contact Terasic, but without success so far.

    The fix is simple. Edit the top module of the design (i.e.: DE2_70_NET.v). Find the following line: 

    assign AUD_ADCLRCK = AUD_DACLRCK;

    Replace with a permanent tri-state assignment: 

    assign AUD_ADCLRCK = 1'bz;

    This should be done on all the designs that use the audio codec in master mode. Seems all NIOS samples do, and seems all have the bug except the Control Panel.

    Note that in same cases there are two identical lines with the assignment causing the bug, one is commented and the other is not. This was possibly, the origin of the bug.

    Check the fitter report to verify the bug was fixed. The AUD_ADCLRCK pin should have a permanently disabled output enable.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    18 years ago

    Hi,I am tring to use TRDB_DC2 on the DE2-70 dev board.But it seems not work.I wonder if the i2c have some problems as you say above.

    Thanks!!
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    18 years ago

    --- Quote Start ---

    Hi,I am tring to use TRDB_DC2 on the DE2-70 dev board.But it seems not work.I wonder if the i2c have some problems as you say above.

    Thanks!!

    --- Quote End ---

    Unlikely. The bug I found is specific to the DE2-70 top module, it is not a bug in the I2C core or on the hardware.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    18 years ago

    --- Quote Start ---

    Unlikely. The bug I found is specific to the DE2-70 top module, it is not a bug in the I2C core or on the hardware.

    --- Quote End ---

    I appreciate your reply!

    I wonder if the I2C protocal in the DE2 board and DE2_70 is alike.

    I mean if they use the same I2C protocal version.

    Would you mind check the problem for me ?Thanks a lot!!

    I really want to know the reason.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    18 years ago

    --- Quote Start ---

    I wonder if the I2C protocal in the DE2 board and DE2_70 is alike. I mean if they use the same I2C protocal version.

    --- Quote End ---

    The protocol version in this case, for external I2C devices, doesn't depend on the board at all. It depends only on the design.