Post Go back to editing

AD8510 and capacitive loads

I use the AD8510 in unity gain configuration as DAC buffer. The resistive load at the output is 10kOhm and there is also a capacitive load due to cable capacitance in the range of 1 - 2nF. Which compensation method for the capacitive load would you recommend? To preserve DC accuracy I won't use a output series resister larger than 5Ohms. The compensation method should preserve DC accuracy and it should be able to cope with 1nF

as well as 2nF capacitive load.

I have an additional question about snubber networks to compensate for capacitive loads: is the order of the capacitor/resistor series connection relevant? Does it have to be op amp output --> resistor --> capacitor --> ground, or is op amp out --> resistor --> capacitor --> ground also ok?

Regards,

Steffen

  • Steffen,

      With 5 ohms and 2 nF, you have a pole at about 15 MHz which will give you some phase shift, etc.

    Op amps are not perfect voltage sources with zero ohm output impedance, so all op amps have

    trouble with cap loads, it's only a matter of degree.

      The only way you have a chance is with a bigger Rs.  Close the loop for DC at the cap load,

    and close the loop for AC before the resistor.  See the last schematic in:

    http://www.analog.com/library/analogDialogue/archives/31-2/appleng.html

    (The + and - inputs are mislabeled)

    For snubber compensation, it relies on dropping the gain quickly before the phase has shifted too much.

    This is most effective with op amps that have high output impedance, such as CMOS rail to rail outputs.

    but is somewhat useful with bipolar RRO.

    Obviously, it should not make any difference with a series circuit, but I had a customer last year that

    said it did.  We did not explore.  Could have been stray C.

    Harry

  • Dear Harry,

    I have a question to the "in-the-loop compensation scheme" you proposed. This compensation method is described in figure 3 of the following article https://ez.analog.com/message/82888#82888

    But how to use this compensation method in a unity gain configuration with an infinite R_IN?

    Due to R_x=R_0*R_IN/R_F this would lead to an infinite value for R_x.

    Regards,

    Steffen

  • Steffen,

      I missed the fact you are connected as a buffer.  One of the designers here said a snubber is the way

    to go;  values on the data sheet.

    Harry

  • This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

    Thank you,
    EZ Admin

Before You Switch


Switching languages will make ADI Explorer unavailable. Resume your session by switching back to English and reopening ADI Explorer.