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.

Glitch observed on DAC7760 output at Module Power ON

Other Parts Discussed in Thread: DAC7760, DAC8760

Hi 

We are using DAC7760 in our Analog Card design

Analog Card is placed on power supply module ( this will provide 5V to DAC ) and separate 24V given to Analog card ( through this +/- 15V is generated fro DAC )

Now, when both Supplies are ON simultaneously, Glitch of -10V of 40ms is observed at DAC Output

What could be the reason?

Thank you,

Saurabh

  • Saurabh,

    Could you provide an oscilloscope capture of this glitch?

  • Kevin,

    Let me inform you, when glitch is observed -

    1. 5V and 15V both supplies simultaneously made ON

    2. Supplies made ON -> Analog Card generates output -> Both supplies are made Off -> again both supplies made ON  

    Please refer the attached Glitch wave forms

    ( wave forms names are provided at extreme left of each wave form )

  • Hello Saurabh,

    It's difficult to 100% nail down the source of this transient since this can have dependencies on the environment in which the component is installed in addition to the component itself.

    For the DAC8760 family of device specifically, there is a recommended power supply sequence described in the datasheet. This supply sequence isn't required but it is recommended to help provide a smooth power ramp sequence for the part. Most ideal is DVDD, AVDD, then AVSS to reduce power up glitch. On a power down sequence the best sequence I can recommend is to keep DVDD high until AVDD/AVSS have dropped. This keeps the analog switches controlled by the digital core in a known state during power down rather than DVDD disappearing and unpredictable things happening inside the device.

    Voltage output stages in general, including the output signal chain of the DAC8760, have some parasitic capacitance to both supply rails. On a power-down event this capacitive coupling feeds through to the output and can create these output glitches. A method you can use to minimize the magnitude of this output glitch is to include a relatively large resistor on the output stage to allow for this output glitch energy to flow to ground. A 10kOhm resistor on the voltage output stage will have minimal impact on the voltage output accuracy and help minimize this output glitch considerably.

  • Hi Kevin

    Thank you for the solution

    We have few queries regarding the solution provided - 

    1) To add 10K resistor at output - do you mean it like Pull Down the Output Pin of DAC 7760

    2) We also use to generate Current Output, will this resistor cause any effect at that time 

    At our end we have implemented one solution,

    We have added a RC Delay Circuit ( R=15E & C=10uF ) in Input Supply path of AVSS supply

    Please refer the wave forms below -

    a) Output without RC 

    b) Output with RC

     

    Please share your comments on the same

  • Saurabh,

    Yes - my proposed solution to reduce the glitch on the voltage output (short of the recommended power supply sequence) was to include a 10kOhm pull-down resistor on the output pin. For applications that share the voltage and current outputs on a single output terminal this is most likely not an option because the pull-down resistor to ground will create a current divider with the load and adversely affect the accuracy of the current output.

    If modifying the power-supply sequence has delivered acceptable output glitch for your needs, this is the best solution. This does not impact the accuracy of the voltage or current outputs and is a reliable solution. Basically the AVDD/AVSS supplies are used for the bipolar output stages while the AVDD/GND supplies are used to power the input stage. If AVSS comes up at the same time or before AVDD the output stage may be active without anything driving the input stage so power on behavior is more or less random. If the input block powers up first the behavior is more predictable and should display lower output glitch on power up.