• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 47 subscribers
  • Views 1666 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

DAC7568: Output gain is doubled but waveform is Fine(spi data write)

vik
Intellectual 500 points
Part Number: DAC7568

Hi all
My DAC output data is doubled for reason not known to me.

I am sending below spi Dac commands at MOSI in my init function:
0x07000000 //reset cmd
0x05000000 //zero scale
0x08000000 //Ref off (since 5V external ref is provided)
0x060000FF //Ignore LDAC on all DAC channels

I am running the below command in to write 0x400 to Channel A.
0x00040000
which is supposed to produce 1.25V (0x400/0xFFF * 5V) But produces 2.5V. Doubled!!
and 0x00080000 produces 5V instead of 2.5V(0x800/0xFFF * 5V).
Is there any gain change command to fix this issue?

Note:
- same function is used to send 32bit spi data to DAC7568.
- Chip select pin(sync pin) is selected throughout the 32 bit transaction.
- SPI frequency is at 100k.
- Only one device is connected on the SPI bus.
- Tried with 1K SPI clocks too, same issue persists.
- SPI mode is set MODE1 (data sampled at clock high-to-low edge).

Refer Below snapshot of oscilloscope when sending 0x00040000:


Refer Below schmatics:

  • 0
    Intellectual 500 points
    Hi all,
    I observed that when a gpio is used as chip select in my linux board, that dac works fine(no gain double issue). To attain better speed I need to use spi chipselect(which caused the double gain issue apparently).
    The only difference that I found in waveforms between chipselect gpio and slave-select pin is that, the gpio is latched(low) much earlier to the spi transaction and released much later after the spi transaction, where as the spi slave-select pin latches(low) immediately before the spi transaction and releases immediately after the spi transfer is over.
    I tried with 1k spi frequency with spi slave-select signals, yet no luck(assuming there should be slave select to spi clock delay requirements). How can I make this work with spi slave-select signal??
  • 0 in reply to vik
    TI__Genius 14220 points
    Hi Vik,

    Thank you for your query. I will get back to you by tomorrow.

    Regards,
    Uttam Sahu
    Applications Engineer, Precision DACs
  • 0 in reply to Uttam Sahu
    TI__Mastermind 46290 points
    Hello Vik,

    I apologize for the delays on this topic.

    Could you share oscilloscope captures showing the differences in the timing details you mentioned? We can check this against T3, T5, and T8.
  • 0 in reply to Kevin Duke
    TI__Mastermind 46290 points
    Hi Vik,

    Just checking back in on this topic. As I mentioned I think the best thing we can do right now is take a look at the timing with the SPI peripheral controlling CS versus the GPIO.
  • 0 in reply to Kevin Duke
    TI__Mastermind 46290 points
    Hello,

    I will be closing this thread due to inactivity. If follow up is posted prior to the thread locking, it will unlock. Otherwise, please click on "Ask a Related Question" to continue the discussion.

    Thanks.