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.

TCA9800: TCA9800 B side powered down

Part Number: TCA9800
Other Parts Discussed in Thread: REF3430,

See attached schematic.

We are using the TCA9800 to interface an MSP430 running at 3.3V to an ADC121C02 running at 3.0 V. This 3.0V is coming from a REF3430. The 3.0V is also being used to power other circuitry.

To prolong battery life, the micro will keep the REF3430 disabled and only enable when required. The TCA9800 VCCA is 3.3V and VCCB is 3.0V.  VCCA is always present but VCCB is enabled only when required as mentioned earlier. 

TCA9800 enable pin is left unconnected so that it is pulled high at all times.

According to the datasheet,

"The SCLA and SDAA pins enters a high impedance stte when either VCCA or VCCB fall below their UVLO voltages".

"When VCCB falls below UVLO, the current source turns off, and a weak pull-up is connected to prevent the B  pins from floating. This is intended behavior, because no external pull-up resistors are to be used on the SDAB or SCLB pins. This behavior prevents the bus pins from floating, and allows it to follow VCCB.".

In our circuit, when the REF3430 is disabled, VCCB will be 0V which means

 1) The SCLA and SDAA pins enter high impedance state preventing any leakage on the A side.

   - and -

  2) SDAB and SCLB will also be at 0V and since the ADC is also powered off, there should not be any scope for current leakage. 

Do you see any issues with this circuit ? 

  • Hello Ram,

    I agree - the loss of the 3-V rail will put the device into UVLO on the B side.  This would make the SDAA/SCLA lines high impedance, allowing the bus to continue to function using the external pull-up resistance.  It sounds like this is what you would want in this application.  One thing you may want to double-check is the power-down behavior of the 3-V components.  For example, if the ADC's SDA/SCL lines are expected to go low before the TCA9800 reaches a UVLO state then it may propagate a glitch to the "A" side when powering down.  If that were the case, you could see if the 3.3-V side were capable of tolerating such a glitch or could consider using the EN control as a workaround.


  • Hi Max,

        Thanks for your response.

       We do not want to use the enable pin since it will consume 3.3V / 250K = 13.2 uAmps when disabled (logic 0) which is very high rendering the enable pin useless in our application (low power wireless sensors).

       I do not think the "glitch" you mentioned is an issue in our case. The ADC (via the TCA9800) is the only slave on the bus and is not being accessed when we power down the 3V supply.

       I am going to wire up a test circuit to make sure we do not see any issues (current consumption etc) before finalizing the PCB.


  • Kris,

    Thanks for the clarifications - that sounds good.  If you see anything strange when you are testing the circuit please just let us know.