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.

DAC8562T: VREFOUT voltage dropping

Part Number: DAC8562T
Other Parts Discussed in Thread: DAC8562

Hello,

I'm encountering some issues with the DA8562T.

When I start sending data to the DAC, the Vrefout become crazy. It rise above 2.5V (to 3v3 which is our Vsupply), or fall to 0V. Unnecessary to say that the output isn't working properly :

Yellow is nSYNC / Blue is clock (~4.5MHz so not enough samples to see it here), and purple is Din. While I send zeroes, everything is ok, and when I start with data, Vrefout problems start (green).

Next to that, if I plug the evaluation board in parallel, with the same supply and same communication BUS in the same time, it's working and my output is ok on eval board.


The only differences are on nLDAC and nCLR which are not the same, but I tried different solutions (0v, pull-up, cpu controlled); nothing change.
And the component footprint (VSSOP on my board).

I did a lot of different tests to find the problem (replacing component, checking supplies, using external reference, send power-on reset command)

I disconnected both output, tried multiple capacitor (on vref and vdd) 

Schematic :

Data bus looks good enough :

An other vrefout drop :

Thank you, if somebody can help me to find a solution, or a way to find the problem...

  • Hi,

    From your schematics, I am assuming you have 200nF cap on RefOut pin which is OK.

    Whats the power up sequence?

    Can you please tell me whats the voltage on RefOut pin after power up before sending any command to DAC?


    Can you do the following for me to get a better understanding.

    Make sure CLR is tied to high

    1. Power up Device

    2. Send RESET command

     (cmd:101, Add:000, Data:0000_0001)

    3. Wait for 5-10ms

    4. Enable Internal Reference of DAC

    (cmd:111, Add:000, Data:0000_0001)

    5. Measure Vref

    If Vref reads 2.5V, then issue another command to update DAC Voltage and set DAC Gain 1

    (cmd:000, Add:010, Data:0000_0011)

    6. Send code to update DAC voltage (Any value is fine)

    Regards,

    AK

  • Thank you for your answer,

    This shot give the power up sequence. Supply is a bit slow but when I use it on eval board, it's ok.

    Yellow: 3.3V supply

    Blue: nCLR  (nLDAC drived in the same way)

    Purple: Vrefout. 0V on startup, rise to 2.5V after commands

    Green: nSYNC

    Commands sequences (I was already doing something near that you suggested):

     /* POR */
      0x28,0x00,0x01

      msleep(10);

      /* LDAC inactive */
      0x30, 0x00, 0x03,

      /* Internal ref */
      0xFF, 0xFF, 0xFF,

      /* Gain = 1 */
      0x02, 0x00, 0x03,

    I'm using a bitbang spi interface for configuration, and once done, I use I2S bus of my CPU. It send continuous clock on SCLK pin, with data and SYNC.
    (2 chan at 96kHz).

    I write first chan, then 2nd + DAC update (as you can see on previous post).

    It seems working when I write low speed values, but not with my I2S. I don't think that it is communication problems because it's working on eval board. Maybe with higher rate, it consume more power, and reveal some hardware problem ?

  • Hi,

    Interface supports upto 50MHz So I dont think this has anything to do with higher rate. whats the update rate you are using, SCLK Frequency?

    Could you please check one more thing, when reference is drooping whats the voltage and current levels on supply pin?

    Regards,

    AK

  • Hello,

    96kHz output sampling rate, on 2 channels. So 192kHz data input rate. SCLK is around 5 MHz.

    I just measured DAC current, with a 10 ohm resistor  between supply and dac8562T (R65 on previous schematic).

    In blue , voltage around resistor (10ohm, so 72mV peak correspond to 7.2mA, a bit high according to the datasheet).

    In purple, it's the VRFOUT. (/!\ measure compared to Vsupply, and not GND because I needed to isolate my oscilloscope from GND to measure between my resistor).

    So we can read 0V, which mean that VRFOUT= 3.3V, and when it drop, current rise.

    Than you for your answer,

    Romain

  • Hi Romain,

    Thanks for providing us with the details. 7.2mA current for this device seems to be on the higher side. Having said that when you tried with eval board, Did you use the same bit bang SPI interface or used the EVAL software to do the test?

    I am assuming your DAC output is isolated for this test. Can you please check with normal SPI interface other than bit bang scheme to isolate interface is not the issue?

    Regards,

    AK

  • I didn't use eval software. I use I²S (Inter-IC Sound) bus with eval board, and everything is ok.( SPI bitbang is just for configuration).

    And with my board, I can plug, in parallel, the eval board. So physically the SAME bus for talking to DACs is used for eval board and for my DAC8562

    Outputs are left unconnected.

    "Can you please check with normal SPI interface other than bit bang scheme to isolate interface is not the issue?"

    I did not really understand your question. Do you want me to use another bus to communicate with the device ?
    (My bus currently works well with eval board).

  • Hi,

    One quick check, did you happen to check with external reference as well? That will give us some clue regarding internal reference misbehavior.

    Regards,

    AK

  • Hi,

    Did you have look at the above suggestions?  Also I want you to check with different capacitor value on Ref Pin, Can you try with 1uF Capacitor?

    Regards,

    AK

  • On the first time, a reference was used, and because of output problems, we disconnected it to ise internal one.

    I also tried different capacitor values on Vref (10µF).

    But finally, we may have found the cause of the problem. The communication bus start from a CPU and go through an EPLD. Some unused pins (CPU side, on bus) were leaved unconnected, and it may be the source of the problem.

    However, we absolutely don't understand why its was perfectly working with the evaluation board plugged in parallel with our system, but not on ours.

    Thanks you for your support, I do not put the post as "resolved" right now, I'll use the system a bit more to be sure that the problem has gone.

    Sincerely,

    Romain

  • Hi

    Glad that your problem got solved. Please test system thoroughly and let me know any additional support you need on the same.

    I will close the thread now, should you have any questions in future, you can reopen the same.

    Regards,

    AK