• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 97 subscribers
  • Views 979 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

AM3359: GPMC wait input timing in Synchronous Burst Read

Daisuke Maeda
Guru 20465 points

Part Number: AM3359

Why was F22 and F21 swapped in Figure 7-18 on the datasheet which is changed from SPRS717J to SPRS717K?

F21 and F22 are described in Table 7-23. They will not need to be swapped.

Best regards,

Daisuke

  • 0
    TI__Mastermind 30731 points

    Hi Maeda-san,

    Figure 7-18 was updated so F21 and F22 match Figure 7-21, but I have concerns about both figures in the SPRS717K version of the data manual.

    I have found the LitBug where this change was requested and will follow up with the tech writer to understand the intent and correct these figures in the next data manual release. The AM437x DM should have also had this same change applied (incorrectly).

    Figure 7-18. GPMC and NOR Flash—Synchronous Burst Read—4x16-Bit (GpmcFCLKDivider = 0)

    F22 label is incorrectly used for setup time instead of hold time

    F21 label is incorrectly used for hold time instead of setup time

    Figure 7-21. GPMC and Multiplexed NOR Flash—Synchronous Burst Write

    F22 label is correctly(?) used for hold time - hold time of WAIT inactive signal latched on a rise edge clock (before low going pulse) - seems incorrect to me.

    F21 label is correctly(?) used for setup time - setup time of WAIT inactive signal latched on a rise edge clock (after low going pulse) - seems incorrect to me.

    Thank you for bringing this issue to our attention.

    Regards,
    Mark

  • 0 in reply to Mark M
    TI__Mastermind 30731 points
    This E2E thread triggered the change the data manual: e2e.ti.com/.../2186218

    The waveform is trying to show the setup and hold time for WAIT assertion as well as when to deassert WAIT so that data is latched on an expected clock edge. I will prepare an explanation to clear up the confusion and submit it for the next revision.

    Regards,
    Mark
  • 0 in reply to Mark M
    Guru 20465 points

    Hi Mark-san,

    Thank you for your reply. Sorry for my late reply.

    "The setup time F21 defines the time before CLK high transition that is required to indicate that data is valid.  Similarly, the hold time F22 defines the time after CLK high transition that is required to indicate valid data."

    I am still confused because these times seem to be not illustrated correctly in Figure 7-18 and Figure 7-21. I would appreciate to give an explanation to clear up the confusion here before the next revision is released.

    Best regards,

    Daisuke

  • 0 in reply to Daisuke Maeda
    TI__Mastermind 30731 points

    Hi Maeda-san,

    The easiest solution is to redraw the two figures where F21 and F22 are shown to appear like below.

    These drawings simply show the setup time and hold time of the WAIT[x] signal as it is latched by the rising edge of GPMC_CLK.

    We are coming up with a new figure that depicts when the WAIT[x] signal must be valid in order for the latching of data at the expected time (as defined by READACCESSTIME). I will share with you that figure once we have created it.

    Does this help?

    Regards,
    Mark

  • 0 in reply to Mark M
    Guru 20465 points

    Hi Mark-san,

    Thank you for your reply. Sorry for my late reply.

    Mark Mckeown said:

    We are coming up with a new figure that depicts when the WAIT[x] signal must be valid in order for the latching of data at the expected time (as defined by READACCESSTIME). I will share with you that figure once we have created it.

    I would appreciate it if you could give me the update here.

    Best regards,

    Daisuke