• TI Thinks Resolved

LMK04808: LMK04808 - command/new register processing time

Genius 12010 points

Replies: 4

Views: 31

Part Number: LMK04808

In the scenario we assume that we have already powered up and programmed the LMK04808. When we try to re-program through MICROwire (just to send one register), how long does it take to process? Is there a specification for this process delay?

For example:

At the beginning, we programmed LMK04808 chips CLKout0_1_PD bit as power down.

Later we re-send this register to the Chip to power it up.

After sending this register to chip, how long does it take to clockout0_1 output toggle?

Is it always fix delay or variable delay?

Is there specifications for delay process? Can it be range stated for this delay?

Can the time be made of Processing time + n Clock cycles?

Or the processing time will be always smaller than the clock time or be also n clock cycles?

Thanks for the follow up,

SunSet

  • Hi SunSet,

    Thanks for posting on E2E. I have assigned your post to the responsible engineer. He will respond to you soon.

    Kind regards,
    Lane

    ____________________________________________________________________________________

    More information Clock and Timing System products: http://www.ti.com/clock-and-timing/overview.html

    Learn more about clocking fundamentals by watching the TI Precision Labs: https://training.ti.com/ti-precision-labs-clocks-and-timing

  • Hello Sunset,

    Before I begin answer your questions, please note that the following timing diagram and timing requirements were used to answer your questions. These diagrams and requirements can be found in the LMK04808 data sheet.

    When we try to re-program through MICROwire (just to send one register), how long does it take to process?

    Process time = tECS + tDCS + tCDH + ((tCWH + tCWL)* 32)

    32 is the number of bits in the register.

    Is there a specification for this process delay?

    There is no exact specification, but an approximation can be found from the timing requirements above.

    After sending this register to chip, how long does it take to clockout0_1 output toggle?

    The time it takes for the clockout0_1 to toggle is the time it takes to latch the contents into the register. This time is represented by tEWH in the timing diagram. The value for tEWH is a minimum of 25 ns.

    Is it always fix delay or variable delay?

    The minimum delay is fixed.

    Is there specifications for delay process? Can it be range stated for this delay?

    There is no exact specification, but an approximation can be found from the timing requirements above.

    Can the time be made of Processing time + n Clock cycles? Or the processing time will be always smaller than the clock time or be also n clock cycles?

    The clock time is not made of the processing time + n Clock cycles.

    Regards,

    Kia Rahbar

  • In reply to Kia Rahbar:

    Thanks for the info.

    Let me add some comments to clarify.

    In addition, we are checking the timing below as a way to ON/OFF outputs with register programming.

    Are there other ways to do it? 

    When we try to re-program through MICROwire (just to send one register), how long does it take to process?

    Process time = tECS + tDCS + tCDH + ((tCWH + tCWL)* 32)

    32 is the number of bits in the register.

    I think this is the time that we need to spend to send register to Chip. But I was asking after that. I was actually asking that the time between LE signals falling edge and chip clockout.

     

    Is there a specification for this process delay?

    There is no exact specification, but an approximation can be found from the timing requirements above.

    After sending this register to chip, how long does it take to clockout0_1 output toggle?

    The time it takes for the clockout0_1 to toggle is the time it takes to latch the contents into the register. This time is represented by tEWH in the timing diagram. The value for tEWH is a minimum of 25 ns.

    Are you sure about tEWH timing representation? I think it is the Latch Enable signal that selects the chip for the register programming. It is like Chip select on SPI Bus. It does not say much about the internal process of the chip in order to latch the new register content. I also agree with your words about ``The time it takes for the clockout0_1 to toggle is the time it takes to latch the contents into the register.``  I am actually looking for this delay and I think it is not tEWH. If your assumption is correct, can I also assume that after 25ns later sending register they will start toggling.  

  • In reply to Sunset:

    Hello Sunset,

    When we try to re-program through MICROwire (just to send one register), how long does it take to process?

    Process time = tECS + tDCS + tCDH + ((tCWH + tCWL)* 32)

    32 is the number of bits in the register.

    I think this is the time that we need to spend to send register to Chip. But I was asking after that. I was actually asking that the time between LE signals falling edge and chip clockout.

    Apologies for the misunderstanding. This is the time it takes for the contents of each register to clock in every bit. The time between the LE signals falling edge and chip clock out would be tECS (25 ns). tECS is the LE to clock setup time, which is another way of saying the time between the LE signals falling edge and the chip clock out.

     

    After sending this register to chip, how long does it take to clockout0_1 output toggle?

    The time it takes for the clockout0_1 to toggle is the time it takes to latch the contents into the register. This time is represented by tEWH in the timing diagram. The value for tEWH is a minimum of 25 ns.

    Are you sure about tEWH timing representation? I think it is the Latch Enable signal that selects the chip for the register programming. It is like Chip select on SPI Bus. It does not say much about the internal process of the chip in order to latch the new register content. I also agree with your words about ``The time it takes for the clockout0_1 to toggle is the time it takes to latch the contents into the register.``  I am actually looking for this delay and I think it is not tEWH. If your assumption is correct, can I also assume that after 25ns later sending register they will start toggling.  

    In order to latch the contents of a register, the LEuWire signal should be toggled low-to-high-to-low (as shown below). The time it takes for the LEuWire signal to toggle (latch the contents into the register) is tEWH (25 ns). 

    Regards,

    Kia Rahbar