• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 16 subscribers
  • Views 307 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

SN74TVC3306: I2C level translation to very low voltage

Philippe SAUVAIN
Prodigy 30 points
Part Number: SN74TVC3306

Hello !

I am using the SN74TVC3306 dual-votlage clamp to interface an I2C master at 1.8 V, to an I2C slave powered down to 0.5 V. It is working correctly down to 0.7 V (on the slave side), but my slave seems to be unable to pull-down the SDA signal when the power supply is lower than 0.7 V. There is only one master and one slave on the I2C bus.

My question is :

- can I use the SN74TVC3306 down to 0.5 V voltage supply on the slave side ? If not, is there any solution in integrated or discrete form ?

Best regards,

Philippe Sauvain
R&D Engineer
SG R&D Ltd

  • 0
    TI__Mastermind 23373 points

    Philippe,

    Is the 0.5V supply on your slave side the Vref point you're look to set your voltage translator to as the 'voltage limiting point'?
    Could you show a block diagram or schematic of your setup?

    Rami

  • 0 in reply to Rami Mooti1
    Prodigy 30 points

    Hello,

    You find below the schematics of the level shifter in my application. The signal ENA_GG was intended to switch off the level shifter, but is currently not used.

    My slave device is connected to the signals XCL_T and XDA_T (specific names for SCL and SDA). On the slave side, the VREF clamping voltage is set to the same voltage as my DUT.

    Any suggestions ?

    BR,

    Philippe

  • 0 in reply to Philippe SAUVAIN
    TI__Mastermind 23373 points

    Philippe,

    Your set up looks fine. Nothing egregious is sticking out to me. Could you try shifting Vref @ Vdd down to 0.4 or 0.3 and see what happens to the clamping voltage?
    Have you also probed the voltages to ensure they're stable at 3.0V/0.5V at the chips inputs. Seeing as how you're 0.1V off, I wonder if the source itself is just a bit off by the time it reaches the supply path for the 3306.

    Rami

  • 0 in reply to Rami Mooti1
    Prodigy 30 points

    Hello Rami,

    Sorry for our peaceful dialogue, that's because I hade only a few time per day to look at this problem ...

    I looked at the clamping voltage and both power supply voltages, they look fine and stable, with minimal votlage offsets down to 0.1 V.

    As my slave device is a custom low power IC, with integrated 1k pullups and relatively weak pulldown transistor, the XDA line cannot be pulled down and I only see the zero bits as small steps millivolts below the power  supply votlage; this effect become worse when the power supply is lower, as the internal pulldown transistor becomes weaker vs the TVC3306 strong pullup.

    Am I wrong when I deduce that the SN74TVC3306 shows a strong pullup behavior ?

    I just made an experiment by replacing the IC with a variable pulldown resistor, and varying it until I got half the Vref/VDD clamping voltage on the XDA_T line; that experiment shows me that the pullup made by the SN74TVC3306 is strong, on the order of 100 ohms.

    BR,

    Philippe

  • 0 in reply to Philippe SAUVAIN
    TI__Mastermind 23373 points

    Philippe,
    If i'm understanding correctly, you're saying that you just tried to replicate it as a voltage divider with resistors? and it showed that the 'effective On-Resistance' was 100ohms? Could you clarify this more if i'm misunderstanding?
    If the pullup you're referring to is the resistance of the chip itself, then I would say that the On-Resistance of the device with your setup should be much lower than that. With your setup it should be nearly 5ohms. Do you have a source measuring unit available that you can use to run a test on the chip by supplying a known current to the input and measuring the voltage drop across? Then use R = V/I to solve this resistance?

    Additionally, a simple albeit very hopeful question, can you vary that internal pull down on your custom device at all? Maybe a quick fix there.

    Thanks,
    Rami