• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 34 subscribers
  • Views 2943 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.

TCA9803: Clock Stretching and overshoot during Slave-Master hand off

Hithesh
Genius 3290 points
Part Number: TCA9803
Other Parts Discussed in Thread: PCA9517, , TCA9517, TCA9406

Are there app notes or article that go into detail about Clock stretching and overshoot during hand-off. 

  • 0
    Genius 3290 points

    The PCA9517 errata mentions a bit, but not in detail

  • 0 in reply to Hithesh
    TI__Guru 62920 points
    Hitesh,

    I don't know of any app notes on this topic, unfortunately. You can find some good info in this post, though:

    e2e.ti.com/.../552045

    The basic idea is that the rise time accelerators could incorrectly trigger during a low level if there is a large enough pulse or overshoot (for example, when the bus transitions from a low level close to 0 V to a slightly higher low level formed by a buffer device's static voltage offset).

    Max
  • 0 in reply to Miguel Robertson
    Genius 3290 points
    In the post you linked, it mentions rise time accelerator detects 30% Vcc rise and triggers on it.
    What is the exact voltage level or voltage range?
  • 0 in reply to Hithesh
    TI__Mastermind 23226 points
    Hello Hithesh,
    I am a bit confused about the post. Are we talking about the TCA9803 or the PCA9517? The TCA9803 will not have this issue because of our constant current architecture with current sensing and no static voltage offset. The PCA9517 has this problem but it is really only a problem if you have a series buffer or if you have a RTA device that is placed on the bus. During the transition from determining which side has control of the zero (low) there is a delay where the bus could travel up high enough to trigger a RTA (0.3xVcc), which would give you a false one (high). The TCA9517 has better performance with respect to this problem due to the reduction in delay, thus the reduction of time the bus rises and could trigger a circuit, like the RTA, that is connected to that bus. The RTA we are talking about are not part of the PCA9517 but rather another device on the bus (either a series buffer that uses RTA or a parallel RTA that is connected directly to the bus). Our devices that have the RTA are set to go off at ~0.3xVcc, they aren't super exact because they don't need to be. So, RTA change the voltage level it trigger with Vcc. A linear relationship.

    Please make sure to explain the question or concern as detailed as possible.
    -Francis Houde
  • 0 in reply to fhoude
    Genius 3290 points

    Hi Francis,

    Sorry about the confusion.

    When I posted this thread, I was looking for any articles or references to Clock Stretching and Overshoot.

    TCA9803 mentioned this and after searching further I found PCA9517 has an errata on Clock stretching.

    All I expected from this thread was to figure out at what voltage RTA (like TCA9406) trigger.

  • 0 in reply to Hithesh
    TI__Mastermind 23226 points
    Hello Hithesh,
    For the TCA9803 that voltage would be very small and would likely not be an issue because the RTA are typically at 0.3*Vcc which would likely be much higher voltage, therefore we aren't likely to have the RTA be triggered. The TCA980x family doesn't have a static voltage offset and it directly makes decisions based on current draw from that node.
    -Francis Houde