• State TI Thinks Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 4 answers
  • Subscribers 20 subscribers
  • Views 5510 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.

TXB0108: Bi-direction level shift/buffer question

Arthur Huang
Expert 6065 points
Part Number: TXB0108
Other Parts Discussed in Thread: LSF0108, , TXS0108E, TXB0304

Hi Team,

customer is looking for the 8 bits bi-directional buffer for their LPC application, can you please advise if we have any suitable device that can suggest to customer?

Thanks.

Regards,

Arthur

  • 0
    TI__Intellectual 2100 points
    Hey Arthur,

    What voltage levels are you looking to translate between? What are your current requirements? How much load/line capacitance do you have? How fast is your signal. Any other information you could provide would be useful in addition to the above questions, i.e. package type.

    Best Regards,
    Nick
  • 0 in reply to Nicholas Mariani9
    TI__Expert 6065 points

    Hi Nick,

    customer already implemented the LSF0108 for 1.8V to 3.3V level shift for LPC bus(low pin count bus at 33.3Mhz frequency), however due to their system design that routed these LPC signals from main board to the daughter board through PCB trace length about 10 inches with 4 vias and one connector, so they are consider to implement another level shift or signal buffer for extending the signals reach.

    Please advise if we have such 33.3Mhz directional 6bits or 8bits small package buffer or line driver to approach this application requirement?

    Thanks.

    Regards,

    Arthur

  • 0 in reply to Arthur Huang
    TI__Intellectual 2100 points

    Hey Arthur,

    As you figured out, the LSF family of devices has no drive strength inherent to itself; see this app note here. This doesn't mean that it can't be used to drive large traces, you just need to ensure that the supply it's hooked to can provide a strong enough pull for it, more about this can be learned in the aforementioned app note.  

    Some solutions to your situation could be:

    1. Attach a supply of sufficient drive strength to the LSF pull ups. This should be able to drive the loads you want and may be the most straight forward solution. 

    2. Add a buffer after the LSF0108. You hinted to this and this could be a great option with minimal effort for install into the circuit. Due to the drive characteristics of the LSF, adding a buffer could give you the drive that you need in order to power your signals across the trace length. A possible candidate could be to use the SN74LV244A. This is an octal chip, can run from2 to 5V for your LPC bus, supports 100MHz, and has a pretty small footprint in the VQFN package of 3.5 x 4.5mm. 

    Additional options for a buffer can be found here on TI's website.

    Let me know if any of these options work for you and please verify the answer if it answered your question!

    Best Regards,

    Nick

  • 0 in reply to Nicholas Mariani9
    TI__Intellectual 2100 points

    My apologies, I just realized that I forgot about your need for bidirectional operation with the buffer I gave you. Everything else I told you still stands however.  One option that could be used instead of the LSF0108 is the TXS0108E. It has internal drive and can supply a continuous 50mA. The TXB0108 would also have its own drive but struggles with high capacitive (~70pF or more) loads and might not perform well with such a long trace and then the actual load after that.  The TXS can run at the 70pF and past it if need be with slower data rates.

    If you have push pull drivers then the max data rate is 60Mbps at 1.8V Vcca, meaning that you might have some bandwidth issues so I'll see if I can track down some additional parts with greater bidirectional data throughput for you.

  • 0 in reply to Nicholas Mariani9
    TI__Intellectual 2100 points
    Arthur,
    I've looked around and I think the solutions menitoned above will be your best bet. Let me know if something changes or one doesn't work for you.
    Best Regards,
    Nick
  • 0 in reply to Nicholas Mariani9
    TI__Expert 6065 points
    Hi Nick,

    Thanks for your information feedback. However, since customer already implemented the LSF0108 on their system board, so it's hard to switch the LSF0108 to TXS0108E unless they are p2p solution. That is the reason why customer was looking for the bi-directional buffer or line driver that interfaces to the original LSF0108 for extending the long trace length reach.

    Any your advise for p2p solution or additional buffer solutions are very appreciated.

    Regards,
    Arthur
  • 0 in reply to Arthur Huang
    TI__Intellectual 2100 points
    Hey,

    Unfortunately, I don't know of another way other than to just increase the drive strength of your supply if you don't want to have a part change.

    Nick
  • 0 in reply to Nicholas Mariani9
    TI__Expert 6065 points

    Hi Nick,

    ok. I see. And do you think is there any concerns if customer insist to use one TXB0304 device (3.3V to 3.3V) connect to LSF0108 for signal buffer?

    Your kindly comments is very appreciated.

    Regards,

    Arthur

  • 0 in reply to Arthur Huang
    TI__Intellectual 2100 points
    Hey Arthur,

    The TXB series of parts could work for you but you could get more drive strength from strong pull-ups on the LSF than the TXB could provide.I would try something from the TXS family first since they will both drive your signals but the TXS has far better output drive capabilities over the TXB family which has a very weak output drive. I know you say that you are already committed to using the LSF part and want to buffer it but with some minor route changes, the TXS can completely replace the LSF part.

    Nick
  • 0 in reply to Nicholas Mariani9
    TI__Expert 6065 points

    Hi Nick,

    Attached are the customer application schematics, can you please review it and have your comments if any.

    Note, customer had a pull high resistors on LSF0108, however, due to the long layout trace length and the load capacitor is high, that caused the signal high level is only 2.2V, low level is only 0.6V, this is the reason they added a TXB0304 after LSF0108 to revert the voltage level become high: 3.3V, low: 0V.

    Thanks for your support and we are looking forward to hearing from you soon.

    Regards,

    Arthur

    Carrier-TXB0304.pdf

    Module-SF0108.pdf

  • 0 in reply to Arthur Huang
    Prodigy 10 points
    Dear Sir,

    I'm Portwell PM, Neil.
    Firstly, thnaks for our supports on this issu. We are still waiting for your reply. Did you have any comments ? if there is any suggestions, please let us know. We need to fix the issue A.S.A.P..

    If needed, you can contact me by email. My email address is neil.su@portwell.com.tw.

    Best regards,

    Neil
  • 0 in reply to Arthur Huang
    TI__Guru* 96591 points
    Hi Arthur,
    Using LSF0108 with TXB0304 is a contradiction - the TXB family of translators cannot work with pull-up resistors, and the LSF series requires pull-up resistors. The correct fix is to move the LSF0108 as close as possible to the higher-voltage device.