AD8251 in CN0385 does not work when measuring the ac signal across a resistor at 1MHz

Hi, I try to measure a AC signal across a resistor at 1MHz using reference design CN0385, where the difference signals (with relatively a large common mode voltage) are converted to single-ended using AD8251. I build this schematic using LTSpice.

However, I found that,

1. The voltage at the Vp node might be affected by the installation of AD8251. With AD8251 removed, V(vp) is a normal ac voltage signal with Vpp = 1.6V. However, when AD8251 is connected, V(vp) becomes,

which is centered above 0V and is around 2V. Although, V(vp) - V(vn) is the same regardless of AD8251.

2. The output of AD8251 is always tied to -5V. 

3. I found that the slew rate of AD8251 is only 20V/us, can AD8251 still be used in my applications where the frequency of ac signal is 1MHz, and the Vamp is around 3V to 7V ? It not, is there alternative ICs that can be used in this application ? (note that there is a large, high frequency common mode voltage when measuring ac signal across a resistor out of many resistors)

Thanks all in advance!

  • 0
    •  Analog Employees 
    on Jun 23, 2017 12:14 AM over 3 years ago


    Apologies if we missed this thread.

    Can you clarify how high is the common mode voltage that you expect on your circuit?

    Note that the output voltage swing of the instrumentation amplifier (3 opamp inamp architecture) is dependent on the input common mode voltage as shown here. This means that you might not get your expected output voltage once the input common mode voltage saturate the internal node of the first stage of the inamp.

    Also the slew rate of the AD8251 is not enough with your requirements in terms of Vpeak and frequency. One of the fasters inamp that we have in terms of slew rate is the AD8421.

    Best regards,


  • we are injecting a small ac current (around 1mA, 1MHz) into the human head and measure the differential voltages of two locations (denoted by m, n) on the human head, where

    Vm = (R_mn + Rn)*I

    Vn = Rn*I

    Rn is the resistance of n to ground, Rn is approximately 3k Ohm and R_mn is the equivalent resistance between m and n, which is approximately ~100 Ohm. We want to measure R_mn as precision as possible.

    The common mode voltage, which is Vcm = (Vm + Vn)/2 = (R_mn/2 + Rn)*I which is relative large and Vcm is also depend on frequency. The difference voltage, which is the one we want to actually measure, is Vdiff = R_mn*I.

    Vdiff is way too much smaller than Vcm. This is why we need a INA with high CMRR at high frequency for our device.

    AD8421 has comparable Slew Rate to AD8251, IMO.

  • 0
    •  Analog Employees 
    on Jul 12, 2017 2:47 AM over 3 years ago

    Hi Benyuan,

    Based on the value of the excitation current and series resistance of R_mn and Rn that you mentioned, I computed the needed slew rate of the inamp and it is around 20V/us. This is already the minimum guaranteed slew rate of the AD8251 at G=1 and you may experience some sort of distortion here. You can solve it by increasing your gain.

    Also note that the AC CMRR of the inamp decreases with frequency. The AC CMRR for 1MHz is only around 60 dB.

    Best regards,


  • 0
    •  Analog Employees 
    on Aug 2, 2018 3:47 PM over 1 year ago
    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