• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 440 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

DAC121S101-Q1: DC OUTPUT impedance DAC 12Bit DAC121S101QCMK/NOPB

Antonio Faggio
Expert 7130 points
Part Number: DAC121S101-Q1

Hi All,

This email to ask you technical information regarding DAC121S101QCMK/NOPB.

 

My customer encountered a problem on DAC121S101QCMK / NOPBI, for which they wanted to carry out tests.

While doing so they realized that in normal use conditions some values read by the instrumentation are not in line with what is reported in the table Electrical Characteristics of the DS.

In particular, customer realized that the measured output impedance is about 8 Ohm, compared to 1.3Ohm reported in DS.

In this regard, I would like to ask you your opinion, why customer found this difference between declared value and measured value?

Many Thanks,

Antonio

  • 0
    TI__Guru 53193 points

    Hi Antonio,

    I think we must first confirm the exact test conditions in which they are testing the output impedance.

    https://e2e.ti.com/support/data-converters/f/73/t/852659?DAC121S101-DC-output-impedance-measure

    "

    The PDS does not give specific conditions for the ZOUT specification, which in my opinion is a minor oversight. The output impedance is really only valid when the DAC (or really the integrated output buffer) is in its linear operation ranges. 

    Let's consider the source of full-scale and zero-scale error: the output buffer is really just an op amp.  Like all op amps, they have some output limitations.  Some are specified with rail-to-rail output operation, but some have a common mode output/input limitations.  Really what you are seeing is the output buffer running into headroom limits with its supply.  With devices like the DAC121S101 were the reference and VDD are connected, it will by definition have 0V headroom at full-scale. 

    Devices like the DAC80501, which have a dedicated reference (internal) and separate VDD pin can get around this as the reference can be 5V and the supply can be 5.5V, so there is 500mV of headroom at full-scale. 

    If you are trying to confirm the ZOUT measurement, I suggest you do the same experiment when the DAC is at mid-scale (2.5V).  This means the device will have plenty of headroom (and "foot-room" if you try sinking current). Also, note that small things like series resistance of the PCB and meter can add error to this measurement.  In our lab, we usually use a kelvin style connection to sense the voltage on the pin so it does not share the same current path.

    "

    Hope this helps,

    Paul

  • 0 in reply to Paul_Frost
    Prodigy 210 points

    Hi Paul,

    I try with mid-scale output (and also with different value of current) but I have the same result.

    I use Kelvin connection during my test because I use SMU (Source Measure Unit) to sink current and DMM to read output voltage.

    Thank you and best regard

    Francesco B.

  • 0 in reply to Francesco Buonomo2
    TI__Guru 53193 points

    What was your test current values?

  • 0 in reply to Paul_Frost
    Prodigy 210 points

    Hi Paul,

    output current are 60uA and 10uA.

    Thank you and best regard

    Francesco B.

  • 0 in reply to Francesco Buonomo2
    TI__Guru 53193 points

    Hi Francesco,

    Does the output impedance look different when the device sinks current as well? For instance, if you source -60µA and -10µA? 

    This device shares a reference with the power supply, so any series resistance on VDD would also scale the output with load. For example, if you had a 5V VDD with 10Ω in series with the VDD pin, and drew 60µA from the output at midscale, you would see the VDD pin reduce to 5V-600µV, and you output would reduce to 2.5V-300µV (not including the actual output impedance of the DAC).  

    This would look like you had 5Ω of resistance on the output.

    Thanks,
    Paul

  • 0 in reply to Paul_Frost
    Prodigy 210 points

    Hi Paul,

    In the circuit there isn't series resistance on Va, I have the impedance of cable between PCB and power supply but is in the order of mΩ.

    I describe below my setup and my test method.

    Setup:

    - PXI 4130 channel 1 (UTY) on Va;

    - PXI 6552 (DIO) on Din, SCLK, SYNC;

    - PXI 4130 channel 2 (LSMU) and PXI 4071 (DMM) on Vout .

    Test method:

    - Power on the DUT with UTY;

    - DIO send on SPI the code 0x7FF in Normal Mode;

    - LSMU sink I2=60uA;

    - DMM read output voltage V2;

    - LSMU sink I1=10uA;

    - DMM read output voltage V1;

    - Zout = (V2-V1)/(I2-I1)

    - Switch off the DUT.

    Thanks for your support and best regard

    Francesco B.

  • 0 in reply to Francesco Buonomo2
    TI__Guru 53193 points

    Hi Francesco,

    Do you have your PXI SMU in remote sense mode? In addition, it is critical on SMUs to not change the output range during these measurements, as the gain error of the various ranges can impact the results.  Does the resistance value change if you source current instead of sinking it?

    Thanks,
    Paul

  • 0 in reply to Paul_Frost
    Prodigy 210 points

    Hi Paul,

    I use SMU in local mode because the DMM measure the output voltage. With my setup I measure also FSO and ZSO with Va=5V and Iout=100uA and the value are in the range of datasheet.

    From my point of view is not clear the condition of the test for DC output impedance, in other post Uttam Sahu tell me for the same measure:

    "You can also do this by using two precision resistors one by one, in place of an SMU. As the DAC output impedance is very small, the measurement accuracy will be important."

    Which is the test condition that TI use to measure the DC output impedance? Which current or output resistor?

    Thank you for your support and best regard

    Francesco B.

  • 0 in reply to Francesco Buonomo2
    TI__Guru 53193 points

    If you use a DMM to measure the voltage at the pin, then you should be good, assuming the DMM is accurate, etc.  

    I do not know the exact method used when validating this product, but generally, we use a device like the PXI-4132 in remote mode, using a signal static range and force a few different currents, both sourcing and sinking.  We use the SMU's internal voltage measurement to derive the output impedance.  

    Occasionally I will sweep a wide range of currents to see if there is varying impedance.  For example, some output amplifiers have a very limited lineary current region.  You are using a pretty small current (10µA) but I wonder what the output impedance would look like if you used something like -10µA, -2µA, 2µA, and 10µA.  

    It is possible to do this test with a few resistors, but usually, a precision SMU is easiest.