• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 1116 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.

DAC8871 Output not working

Dave Boechler
Prodigy 70 points
Other Parts Discussed in Thread: DAC8871, OPA277, DAC8760

I am working on a frequency voltage adjustment project for a Frequency Standard and not getting the output expected from a DAC8871.

SPI output and relevant part of the circuit can be found here:  https://plus.google.com/112981640120383921331/posts/9VFK56QYZ3n

I have a REF5010 feeding the DAC8871 and see 10.00 volts on the VREFHF/S and ground on VREFLH/S. The VCC shows 12.06 volts and VDD show 4.87 volts. RSTSEL and LDAC are tied low and CS and RST are tied high except during SPI transmission. (RST is bounced prior to CS going low)

noInterrupts(); // disable interupts to keep them quiet
digitalWrite(RSTPin,LOW); // Reset the DAC before transmission
delayMicroseconds(5); // reset time
digitalWrite(RSTPin,HIGH); // Reset complete
delayMicroseconds(5); // pause to wait for reset
digitalWrite(CSPin,LOW); // Chip Select pin low to indicate start of transmission (DAC)
delayMicroseconds(1); // pause before sending data
SPI.transfer(HighByte); // send in the High Byte
SPI.transfer(LowByte); // send in the Low Byte
delayMicroseconds(1); // pause before pulling CSPin high
digitalWrite(CSPin,HIGH); // Chip Select pin high to indicate end of transmission (DAC)
interrupts(); // Enable interupts

SPI Transmission seems to be working properly as the DAC does change voltage on the Output, but only by a small amount even trying full swing from 0 to 65535. I can't get more than 750mv full swing on the output before the output filter.

Not sure what I did wrong here. Ideas?

  • 0
    TI__Genius 10615 points

    Hello Dave,

     

    Dave Boechler said:
    The VCC shows 12.06 volts

    The minimum VCC voltage should be +13.5V.

     

    Your SPI sequence and schematic seem correct.

    Why are you resetting the DAC before every new code write? It won't affect the performance of the DAC, I'm just curious.

  • 0 in reply to Eugenio Mejia
    Prodigy 70 points

    Hi Eugenio

    I added a voltage regulator for the OCXO, which requires 12v and cranked up the input voltage to the DAC and Op Amps (OPA277) to 14.5v and it made no difference. The DAC output read as follow:

    .745v when SPI shows FFFF (65535)

    .654v when SPI shows 4E94 (20115)

    .004v when SPI is set to 1

    I am resetting the DAC out of desperation to see if there was some issue with the power up sequence as described in the datasheet page 16 "POWER-SUPPLY SEQUENCING" 

    I don't think that is the issue, but thought it would not hurt to try. I can take it out of the code.

    Any other suggestions?

    Dave 

  • 0 in reply to Dave Boechler
    TI__Genius 10615 points

    Dave,

     

    I did not notice this earlier but you have VSS tied to GND. If you refer to page 4 of the DAC8871 datasheet the VSS voltage should be below -13.5V.

    If you do not have a negative supply available, I can try to recommend another single supply device to suit your needs.

    What are the most important specifications that made you choose the DAC8871?

    Dave Boechler said:
    I am resetting the DAC out of desperation to see if there was some issue with the power up sequence as described in the datasheet page 16 "POWER-SUPPLY SEQUENCING" 

    The reset pin cannot be a reliable substitute for the power supply sequencing. The problem is that if the power supplies are powered-on out of sequence, the device's internal power-on circuit can be corrupted. The reset pin resets the DAC registers only, it does not reset the power-on circuitry. This can cause the DAC to behave erratically. Some devices may power-on correctly even with the wrong power supply sequencing, but this is not guaranteed across all devices.

  • 0 in reply to Eugenio Mejia
    Prodigy 70 points

    I misread the specs when I picked the part then as it shows on page 2 the differential voltage from VCC to VSS to be –0.3 to +39.6 V. I took that as you could use a single supply as long as VSS differed by at max, the amounts shown.

    What I need in a DAC is:
    - 16 to 24 bit resolution.
    - Wide Unipolar supply capability. (i.e. min 15v VCC and analog ground for VSS)
    - Similar low noise, drift and thermal characteristics of the DAC8871.
    - Need to drive with a 10v reference like the REF5010 so the 16bit range swings is 0 to 10v, or as close as possible.
    - SPI (or I2C) 5V uC input with an option of 3.3v if possible.
    - Speed is not critical, stability is.

    Do you have such a product?

    Dave

  • 0 in reply to Dave Boechler
    TI__Genius 10615 points

    Hi Dave,

     

    If you are interested in looking at another device I would recommend looking at the DAC8760. We have a very limited portfolio of high voltage DACs (>5V) and most of them require a negative rail as well, the DAC8760 is the exception.

    It is a device with more features than you are interested in, but I think it may fit your application well. It can be used with a +15V AVDD and 0V AVSS and it generates a 0V to 10V output signal using its internal reference.

    The downside is the noise. The spectral noise density is 10x higher (185 nV/√Hz).

     

    Staying on the original issue, I tested the DAC8871 in the lab with your original conditions.

    I was able to get normal functionality. (I did not test accuracy and performance).

    Measurements:

    • 0V @ 0x0000
    • 5V @ 0x8000
    • 10V @ 0xFFFF

    Test Setup:

    • AVDD = 12 V
    • AVSS = 0 V
    • VDD = 3.3 V
    • VREFH = 10 V
    • VREFL = 0 V
    • \LDAC = 0 V
    • Digital Inputs = 3.3 V
    • RSTSEL = 3.3 V
    • \RST = 3..3 V

    I am still curious as to what is going on in your circuit.

    In the lab I also tried violating the power-on sequence. This seemed to have the biggest impact on functionality, sometimes the device would function and sometimes not at all.

    So just to make sure I have a few questions:

    1. What is the power-on profile that the device is currently subject to?
    2. You are latching data on the rising edge of the CLK correct? At least thats what I can tell from your capture. (SDI being the yellow line)
    3. Where are you connecting AGND and DGND?

     

  • 0 in reply to Eugenio Mejia
    Prodigy 70 points

    Eugenio

    1) LDAC, RSTSEL are permanently tied to digital ground. RES, CS are tied to 5v at startup through a 10k resistor connected through the voltage regulator of the microcontroller. (Arduino Nano for the prototype) I have not measured the delay on VDD. VCC is on when power is applied to the circuit, which supplies the REF5010 and DAC8871 with 14.5 volts (currently) I suspect then that VREFH has little or no delay at startup as well. That was one of the reasons I tried a reset in the code before sending data to the DAC.

    2) In the analyzer data Cyan is Chip select, Red is SCLK and Yellow is SDI as you stated.

    3) AGND and DGND are kept separate until they reach the voltage input connection for the board itself.

    Based on your lab testing I am suspecting my power up sequence could be part of the issue. The prototype is an etched board and there are no jumpers, wires, etc. The only other difference I see in your test is the use of 3.3v for the digital power.

    Dave

     

  • 0 in reply to Dave Boechler
    TI__Genius 10615 points

    Hello Dave,

     

    Have you been able to solve the issue?

    An oscilloscope capture of the power supply ramps would maybe let us know if there is an issue at power up.

    Also, I'm having problems seeing the images you provided originally. Can you attach them to a reply?

  • 0 in reply to Eugenio Mejia
    Prodigy 70 points

    Had to go a different route for now. When I have more time I will revisit this and see if I can get a better understanding of the issue, which I will post.

    Thank for your help

    Dave