• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 18 subscribers
  • Views 907 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LSF0108: optional voltage translation of PCI bus from 3.3V to 1.8V

Troy G
Prodigy 20 points
Part Number: LSF0108
Other Parts Discussed in Thread: TXS0108E

Hi,

Is the LSF0108 a practical choice for performing 3.3V to 1.8V translation of a PCI bus operating at 33MHz? 

This assumes the total load/trace capacitance meets the LSF0108 datasheet requirements. 

This would be for 2 PCI down devices.  No add-in card or connector. 

It is understood that PCI is not spec'd for 3.3V.  This is for an optional custom solution where one end device operates at 3.3V and the other at 1.8V.

Thanks!

  • 0
    TI__Guru* 96591 points

    LSF0108 can definitely handle 33MHz translation from 1.8V to 3.3V. The device should be placed as close as possible to the 3.3V side to ensure minimized capacitance. 

    I would recommend using 10kΩ pull-up resistors on the 1.8V side, and 330Ω pull-up resistors on the 3.3V side.

    My calculations assume that the 'high side' parasitic capacitance is <= 5pF and the drivers on the line are capable of sinking ~10mA. If either of these conditions are not met, the signal integrity will be compromised.

  • 0
    Prodigy 20 points
    Thanks for your input.

    I have a few follow-up questions.
    1. Do you consider the LSF0108 as the best option for such an application?
    2. To meet the parasitic capacitance requirements (PCB trace, device pin, PU resistor pin), this suggests the LSF0108 should be placed within 1" of the hi-side device. Is that true?
    3. Is an IBIS or Spice model available? Historically, there have been issues with modeling bi-dir translators and quickswitches, is this still the case?
  • 0 in reply to Troy G
    TI__Guru* 96591 points

    1. Do you consider the LSF0108 as the best option for such an application?

    This depends on a few things. A unidirectional or direction controlled translator would produce a better signal - but if you can't use them, then you have to go with an auto-bidirectional translator.

    TXS0108E would probably be the best choice for this translation. It operates very similarly to the LSF and has similar limitations, however it has internal pull-up resistors and rising-edge accelerators that will improve signal quality. I would also recommend minimizing high-side capacitance on the TXS0108E, however it will tolerate larger capacitance than the LSF translator.

    LSF0108 provides flexibility in the pull-up/down resistor values as well as the switching threshold, but requires external components to function.

    2. To meet the parasitic capacitance requirements (PCB trace, device pin, PU resistor pin), this suggests the LSF0108 should be placed within 1" of the hi-side device. Is that true?

    Yes - we recommend placing the LSF or TXS translators as close as possible to the high-side device. The closer, the better.

    3. Is an IBIS or Spice model available? Historically, there have been issues with modeling bi-dir translators and quickswitches, is this still the case?

    An IBIS model is available for both devices mentioned above. They are located in the "Tools & Software" tab in the respective device's product folder.  Links below:

    LSF0108 Tools & Software

    TXS0108E Tools & Software

    We are currently working with TI's modelling team to produce usable spice models for these devices, however it will be some time before they are available.