• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 33 subscribers
  • Views 1036 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.

TCA9517 use for optional alternative connection

Andrew Cottrell
Prodigy 100 points
Other Parts Discussed in Thread: TCA9517, TS3A24157

Hi


I'm trying to design a circuit which does the following:

One single I2C slave, running at 3v3. Controlled by either a master which is always there, or if in a configuration, ignore that master and use a separate one. Both masters run at 3v3 however its 3v3 cannot be connected together.

My current design can be seen like this: 

The power from the not always present master enables one TCA9517 and disables the other TCA9517. 

The SDA/SCL lines have appropriate pull ups. The enable line has an appropriate pull down for when the master is not present. 

My question is, will this work? How does the IC behave when you only power one side of the TCA9517?

  • 0
    TI__Intellectual 960 points

    Andrew,

    This looks like it should work. A simple and cheaper solution since you arent using the 9517 to translate voltages might be to use a basic 2:4 mux such as the TS3A24157. You can use the same detection scheme as listed above, have the slave as the common ports, and tie the select pins together.

    Something to be aware of though is what the masters and slaves are doing during this hot swap procedure. The always present master might get hung-up if the 2nd master gets plugged in in the middle of a transaction, never seeing an acknowledgement from the slave. Also in this situation, it would be best to put the pull-ups on the master side rather than the slave side.

    Do you have some external controls to reset the masters and slave if something weird happens during a transaction?

  • 0 in reply to James Smith
    Prodigy 100 points

    Thanks, and there's no risk to the device by only having one side powered?

    I will look into the Mux too, it sounds promising, my main reason for using this part is that we already have it used on other products and it seems to do what I was looking for.

    The procedure won't be actively hot swapped, the devices will be in a reset stage or similar during the changeover, as well as the enable disable signal will be sent to the always present master as a signal with some pre existing circuitry. 

     

  • 0 in reply to Andrew Cottrell
    TI__Intellectual 960 points

    Andrew,

    If one side is unpowered, all SDA/SCL pins on the device will be high-impedance.

  • 0 in reply to James Smith
    Prodigy 100 points

    James Smith said:

    If one side is unpowered, all SDA/SCL pins on the device will be high-impedance.

    Perfect! Thank you!