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.

DAC7574: I2C bus voltage minor mismatch - problem?

Part Number: DAC7574
Other Parts Discussed in Thread: PCA9518, TCA9535, TS5A23157, , REF5040

Hello

we have an I2C sub-bus within our OMAP embedded system, fed from a PCA9518 I2C bus hub.

One of the outgoing buses runs on 5V level, two pull-up resistors, each 4k7, feed the SDA and SCL lines. This bus runs in FAST mode, 400kbps, and serves one TCA9535 GPIO extender, which itself controls a couple of TS5A23157 analog switches and the gates of two MOSFETS inverters. The voltage range of the analog signals is 0 .. 5V single supply, and the MOSFETS switch two 8V supplies. Supplying the GPIO extender with 3.3 volts is not an option here.

Two further bus participants on this I2C branch are one TMP0102 thermometer and one low speed precision DAC7574.

These two chips are supplied from different voltages.

The TMP0102 gets a 3.3V supply through a 1k resistor, in order to filter its supply voltage from digital noise, using a 0.1u bypass capacitor. This method is described within the data sheet SBOS357F, figure 15. In the same data sheet is mentioned, that the I2C bus supply voltage may be higher than the chip supply voltage, if the digital i/o current is limited to less than 3mA, which is provided by the 4k7 I2C pull-ups in series with the 5V digital supply. The temperature sensor works well, the values read out are pretty exact, so there is no internal warm up by the quiescent current flowing in from the (higher than supply) I2C bus voltage level of 5 volts.

My question regards the DAC7574. This one is supplied with 4.096 V from a REF5040 reference chip, which is used as voltage reference for an ADC also. The DAC, as well as the ADC, perform some measuring functions in the 5V single supply analog circuitry, mentioned above. For matching the measuring ranges, it is very handy to supply the DAC with 4.096 volts - because there is no separate reference voltage input.

Because the DAC is fed from 4.096 volts, its two I2C digital bus terminals get a bit overdriven from the 5V quiescent I2C bus level. The current is limited by the pull-ups at approx. 200uA ( ( 5V - 4.096V ) / 4700 Ohms ) and the circuit operates pretty well.

Is this an acceptable application case for the DAC7574, or do we have to reckon with long term degradation or even damage of the DAC, and, if so, can you give a recommendation to circumvent the problem with an acceptable effort? Note: all four I2C branches of the PCA9518 are already in use and can't be used for an new branch running on a lower voltage for the DAC seperately.

Thank you in advance

Horst

  • Hi Horst,

    Thank you for your query. In my opinion, the best way to have different I2C slaves operating at different power supplies is by separating them either (1) through MOSFETs or (2) I2C level translators. Draining DC currents into I/O pins continuously will certainly degrade the life and hence, is not a recommended way of operation.

    Please see one of TI's I2C level shifters: www.ti.com/.../pca9306.pdf

    I hope to have answered your question.

    Regards,
    Uttam Sahu
    Applications Engineer, Precision DACs
  • Thank you very much for this quick answer! The idea is pretty, the part is really cheap and there's enough additional board space available!
    quickly scanning the data sheet, I understand, that the port 2 side is recommended for the higher voltage level, the 'outside' 5V i2c bus branch and the DAC should be connected to port side 1. Is this important to observe or are the two sides electrically equal, the part fully symmetric?
    Horst
  • Hi Horst,

    I am sorry to say that I am not the right person to provide any answer on the I2C part. You can ask the question in the respective forum. They should be able to resolve your queries.

    Regards,
    Uttam