• TI Thinks Resolved

PCA9548A: Is PCA9548A a Pass-through FET or does it re-drive signal?

Part Number: PCA9548A

Hi,

Looking at the Functional Block Diagram on page 10 of datasheet it appears that the connection between master and slave is a pure pass through. However, the "Vpass" spec has me concerned. This somehow allows voltage translation by limiting the voltage passed through to the other line. My main concern is whether this affects the Vol.

Basically, in my application, the master has a specified Vol. Is this Vol passed through the device or does the PCA9548A re-drive the signal in someway that creates its own Vol or in some way increases the master's Vol? Is the Vol of the PCA9548A output determined by the input voltage low level?

Thank you!

  • Hey Cassidy,

    "Basically, in my application, the master has a specified Vol."

    -Are you saying the master generates a static voltage each time it pulls low? (example: no matter what Vcc or pull up resistor it will basically generate a low of 500mV) Or are you saying that somewhere in the datasheet it states master's VoL is something like 0.4V with 3mA as IoL?

    "Is this Vol passed through the device or does the PCA9548A re-drive the signal in someway that creates its own Vol or in some way increases the master's Vol?"

    -Your question here is if the device device is actually using a purely pass FET architecture or if it is actually buffering the input and generating it's own output. In both cases of this, it is possible for the master's VoL to be increased (will do it every time for pass FET and for a buffer with a static voltage offset it can vary).

    Below I made an example using the pass FET design which PCA9548A uses, due to the Rds(on) of a pass FET the VoL seen by the slaves is a little bit larger (due to voltage divider)

    "Is the Vol of the PCA9548A output determined by the input voltage low level?"

    Yes, V(low) seen by the slaves will be V(seen by slaves) = VoL(master) + Current(throguh pass FET)*Rds(on)

    Thanks,

    -Bobby

    Extra: I would like to also highlight that we have a TCA9548A which the TCA lineup fixes bugs with the POR of the PCA family, generally has a wider Vcc range, is more cost competitive, and is pin to pin with the PCA.

  • In reply to Bobby Nguyen:

    Thank you for your response!

    So it looks like it is a pass FET design, and yes certainly the Rdson will cause a rise.

    The Vol from the master does vary with current draw but I can calculate what that will be given the pullups. The TCA9548A have a pass-FET implied that my master is responsible for sinking the current of the pullup resistors on both sides the the 9548, correct?

  • In reply to Cassidy Aarstad:

    "The TCA9548A have a pass-FET implied that my master is responsible for sinking the current of the pullup resistors on both sides the the 9548, correct?"
    Yes, this is correct. The master will have to sink all the current of all the enabled channels when it pulls low. The pull up resistors need to be calculated in accordance to how many channels will be on when the master is pulling low.

    (more channels ON means more current from each channel through the master's pull down FET thus a larger VoL generated by the master, you need to think of the pull up resistors on all enabled sides as parallel like you implied in your question here)

    -Bobby