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.

TCA9544A: how to estimate Max VOL(output low voltage) ?

Part Number: TCA9544A

Hi team

I want to estimate Max VOL(output low voltage) to check I/O level compatibility with Slave device.

may I get your quick reply ?

[1] using VCC = 5V, may I take Max Ron = 16 ohm to estimate Max Vb for Max VOL at below diagram ?

[2] using VCC = 3.3V, may I take Max Ron = 20 ohm to estimate Max Vb for Max VOL at below diagram ?

[3] and what Vo = 0.4V and Io = 15mA mean at that Ron spec ?

  • Hey Paul,

    I've assigned a colleague of mine to take a look at your request, I believe he will get back to you sometime tomorrow.

    Sorry for the delay
    -Bobby
  • Hey Paul,

    Things are a bit busy here so I'll be taking over the question instead of my colleague.

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

    I've redrawn the diagram you made with some adjustments, I will be focusing on only I2C line (this will work for either SCL or SDA so it doesn't matter)

    When the host turns on the FET, current will flow through the FET (current 1 and current 2 in the diagram). The total voltage across the pull down FET will be the VoL1 generated by the host BUT NOT WHAT THE SLAVE SEES. (I will show you how to estimate that later but we need to calculate VoL1 of master first)

    1) To Calculate VoL1 the equation V=I*R still applies but we need to calculate I and R here:

    Lets first make an assumption on the pull down impedance of the host's pull down FET. I2C standard states the Max allowed pull down voltage is 0.4V if you are sinking 3mA of current. Below is an excerpt from the I2C standard:

    So this can help us estimate the max impedance of the master's FET.

    V/I=R

    0.4V/3mA=133 ohms

    This value here tells us the maximum the FET's impedance is allowed to be but keep in mind in a real life use I would expect this to be much smaller and is dependent on the voltage at the FET's gate.

    For these calculations we will just assume the impedance is 133ohms as a worst case scenario but I believe a more likely impedance would be around 20~40 ohms.

    2) Lets calculate current 2:

    I will use your 10k pull up resistor value as R2 like in your diagram and use 5V as Vcc.

    I=Vcc/(R2+Zpulldown)

    Current 2 = 5V/(10kohms+133ohms)=~0.5mA


    3) Lets calculate current 1:

    I will use your 10k pull up resistor value as R1 like in your diagram and use 5V as Vcc.

    Current 1 = 5V/(10k+Ron+Zpulldown)

    Ron from the datasheet for 5V states a max of 16 ohms and a typical of 10 ohms. Please note that this is assuming that the current passing through the pass FET is 10mA, I suspect the impedance to be much less (maybe around 3 ohms or less at room temperature). We can assume worst case though and just use 16 ohms as I don't think it will affect our measurements much.

    Current 1 = 5V/(10k+16+133)=~0.5mA


    4) We can now calculate VoL1:


    VoL1=Zpulldown*(Current1+Current2)

    VoL1=133ohms*(0.5mA+0.5mA)=~133mV

    5) We can now calculate VoL2 which is the voltage the slave sees:

    VoL2=VoL1+(current1*Ron)

    133mV+(16*0.5mA)=133.008mV


    From this we can see that the Ron of the pass FET does not contribute much deviation from what the master's VoL sees compared to the slave for this scenario. 

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

    "what Vo = 0.4V and Io = 15mA mean at that Ron spec ?"

    This means if 15mA of current is flowing through this FET and the voltage difference between the FET is 0.4V then you can expect the Ron to be what is specified in the datasheet.

    Lastly, keep in mind these calculations made an assumption that only one channel of the device was on and not multiple (this would change VoL1 that the master generates).

    -Bobby