• Resolved

TMP435 IBIS Model Question

According to the TMP435 datasheet, SDA signal need using  an external  pull-up resistor as it is an open drain buffer.

But when I use the ibis model (download from TI website) ,the GND_clamp V/I curve of  SDA_3 model have 2.5mA current in 0V.It just like an internal pull-up resistor exist.

So my question is whether an internal pull-up resistor use in the pin 9(signal SDA)?Is the IBIS model accurate?

  • Shen,

    A pull up resistor is required for SDA and SCL as they are open drains. I will look further into your question about the IBIS model and get back to you shortly.

    Best Regards,

    Chris Featherstone
    Linear Characterization Engineer
    High Performance Linear Products

  • In reply to Chris Featherstone:

    Hi Shen,

    Let me update information for your question.

    Actual device does not have internal pull up (SDA pin).

    For IBIS model, it is under survey and get back to you shortly.

     

    P.S. As SCL(clock) pin is an input pin, it is not an open-drain for this device.

    Best regards,

    Ichiro Itoi 

  • In reply to Ichiro Itoi:

    Hi Shen,

    This is the update of the IBIS simulation for your question.

    When I sweep the SDA pin with a bias from negative to positive, as shown in the graph, It shows 2.7mA at 0V, 0mA at 40mV.

    I think this replicates your simulation. It would be appreciated if you could send me your simulation schematic and condition. (also include X-axis and Y-axis on your graph. )

    Best regards,

    Ichiro Itoi 

  • In reply to Ichiro Itoi:

    Hi Ichiro Itoi,

    Thanks for you answer!

    I just do the model checking work. So I haven't the schematic for simulation. 

    Through the Ground Clamp Curve, I can see 2.4mA at 0V.  Could you tell me the origin of the 2.4mA.Leakage current or the pullup resistor current ? 

    If we do the pre-simulation use the Star-topology(One Driver to 8 tmp435).

    2.4mA is  large enough to affect the simulation result.It will increase the bias voltage.You can see in the following picture.

    One TMP435 reciever,

    Eight TMP435 receiver,

     

    Thank you very much!

    Best regards,

    Shen feiqin

     

     

     

     

     

  • In reply to feiqin shen:

    Hi Shen,

    Thank you for sharing your simulation result.

    If I understand your simulation correctly, the sum of the SDA currents from eight TMP devices raises the low level about 750mV.

    Let me clarify the situation as follows.

    1. Actual device has no pull up circuit internally and there is no current that appears when SDA = 0V (in the condition where no pull up resistor is on the bus).

    2. IBIS model of the device shows SDA = 2.7mA when at  0V  (in the condition where no pull up resistor is on the bus) and it is under investigation.

    3. IBIS model should work with pull up resister on a bus.

    My proposal, at the moment, is to use a pull up resistor in your simulation.

    If there is any reason why there is no pull up resistor on the bus for your simulation or anything I can do, please let me know.

    Best regards,

    Ichiro Itoi

  • In reply to Ichiro Itoi:

    Hi Shen,

    Could you provide us the simulator name and version, so that I could prepare for the IBIS re-modeling.

    Best regards,

    Ichiro Itoi

  • In reply to Ichiro Itoi:

    Hi Ichiro Itoi,

    The simulator is Mentor Hyperlynx 8.2.1.

    Thank you!

    Best Regards,

    Shen Feiqin

  • In reply to feiqin shen:

    Hi Shen,

    Thank you for the information.

    Let me proceed re-do IBIS modeling. I will contact you within two weeks.

     

    Best regards,

    Ichiro Itoi

     

  • In reply to Ichiro Itoi:

    Hi Shen,

     For re-do IBIS modeling, it seems a lot of queues at the moment.

    Let me update status within a week.

     It would be appreciated if you could use pull up resister on your simulation.

     

    Best regards,

     Ichiro Itoi

  • In reply to Ichiro Itoi:

    Hi Ichiro Itoi,

    Thanks for the updata.

    I used the pull up resistor in my simulation.

    Best regards,

    Feiqin