• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 34 subscribers
  • Views 1213 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.

PCA9517: about pull up of PCA9517 LOADING

Vera Mao
Expert 5065 points
Part Number: PCA9517
Other Parts Discussed in Thread: TCA9548A, TCA9517, TCA9509

Hi dear supporting team,

customer used PCA9517 and PCA9548 in their system,  but meet issue for the pullup resistor, the topology is as below,  the questions is:

1.  if they have 1k pull up at 9517 side, should they add pull up at the loading side?

2. for 9548 output, they found if they add 2k pull up, the load may not be recognized, if they add 4.75k, then some load could be recognized.  the SCLK is 400k.  and the load cap is unknown, assume 10pF for each.  so how to choose the Rpu? tks a lot!

  • 0
    TI__Guru 73651 points

    Hey Vera,

    I would like to recommend the TCA family as potential replacements for our PCA. (TCA9517 and potentially TCA9548A) The TCA line up is pin to pin with the PCA, are more cost competitive, fix some bugs from the previous family, and some support wider voltage ranges.

    "1.  if they have 1k pull up at 9517 side, should they add pull up at the loading side?"

    -Yes, pull up resistors need to be present on all channels of the PCA9548.

    "2. for 9548 output, they found if they add 2k pull up, the load may not be recognized, if they add 4.75k, then some load could be recognized.  the SCLK is 400k.  and the load cap is unknown, assume 10pF for each. "

    I would like to see scope shots (SDA/SCL) for the devices that are not recognized. From what you are describing, your VoL may be too large and causing issues to either the master or slave.

    "so how to choose the Rpu?"

    Here is a guide:

    Thanks,

    -Bobby

  • 0 in reply to BOBBY
    TI__Expert 5065 points

    hi Bobby,

    thank you for the reply!

    yes, I am also promoting them to do the replacement, while it may take a long cycle.

    as for the SCL/SData scope shot, pls see below, the Vol is too high.

  • 0 in reply to Vera Mao
    TI__Guru 73651 points

    Hey Vera,

    I think the VoL may be large because of the PCA9517. The B side has a static voltage offset when A side is driving it. This voltage is about 0.52V but can become larger with stronger pull ups.

    This should be okay for devices using a 3.3V logic though as the ViL for them is about 1V (30%*3.3V). If the customer is using legacy I2C devices then maybe this could be an issue.

    From the scope shot you have provided, I cannot really tell what is going on. Can you provide the SDA AND SCL waveforms. The SCL with the SDA will help me look at what is going on bit by bit. It looks like right now the device is not seeing an ACK because I don't see another VoL besides the one from the PCA9517.

    Rise time looks okay to be me.

    One thing we could try is to use TCA9509 as a pin to pin replacement and see if this resolves the problem (if it is related to VoL.....). The static voltage offset of TCA9509 is on the A side of the device so the slave devices won't see a large voltage on the B side. One thing about this device is that the A side requires that there are no pull up resistors on that side so you will need to depopulate them.

    Thanks,

    -Bobby

  • 0 in reply to BOBBY
    TI__Expert 5065 points

    hi Bobby,

    thank you for the reply!

    unfortunately, they are using TSSOP package of PCA9517, Seems 9509 has no such package. any other choice?

    and below is I2C plot: blue is SCL, yeLlow is SDA.

  • 0 in reply to Vera Mao
    TI__Guru 73651 points

    Hey Vera,

    "unfortunately, they are using TSSOP package of PCA9517,"

    You mean the VSSOP?

    Because our TCA9509 also comes in the VSSOP package:

    We should be able to replace the two fairly easily.

    ---------------------------------------------------------------

    From your waveform I can see that the slave is NACKing but I can also see that it looks like PCA9517 isn't doing anything wrong.

    Are you able to tell me what the slave NACKing is? (device #/datasheet?) I would like to check to see if ~600mV offset is acceptable for it. It may be that VoL of the buffer is too large for it to recognize the signal is low.

    Thanks,

    -Bobby