• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 48 subscribers
  • Views 1404 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.

LMP7708: Oscillation seen on the band pass filter

Amit Ashara
Guru**** 244380 points
Part Number: LMP7708
Other Parts Discussed in Thread: LMP7702

Hello Team,

We are working on a band pass filter for an accelerometer. What we see on the output is that for steady state output there is a ~4.8MHz oscillation. We checked the supply and it is a steady state 3.3V supply to the filter and sensor. We removed the DC-blocking caps C7, C10 and C13 and checked the output of the sensor and it shows a clean DC output. However when we connect the scope probe to the any of the cap's we see the oscillations.

The spice-sim we ran on the circuit did not show any such issue.

Attached is the snapshot of the Schematic.

  • 0
    TI__Mastermind 23050 points

    Hi Amit,

    The LMP7708 has a second pole in its open loop gain curve that occurs just before the 0dB crossover, and as such is not a unity gain stable device.

    It is recommended to run the device in gains greater than 10. In this case, the addition of the 1800pF feedback cap on the first amplifier stages causes the amplifier gain to fall back to unity in the upper frequency range, before crossing the Aol curve. This results in instability.

    If you want to maintain the bandpass filter functionality then you will have to add some components to compensate the op amp. One approach would be to add output pin compensation by adding an RC directly on the op amp output as shown below. This adds a pole and zero to pull Aol down so that the second pole in Aol doesn't interact with the feedback network. Something similar would need to be added to stabilize the buffer if you wish to continue using the LMP7708.

    LMP7708 BP Filter.TSC

  • 0 in reply to Zak Kaye
    TI__Guru**** 244380 points
    Hello Zak

    OK, let me try to add the Compensation circuit as shown. Are there any other parts that I can look at?
  • 0 in reply to Amit Ashara
    TI__Guru 53036 points
    Amit,

    The LMP7702 is the "compensated", unity-gain stable version of the LMP7708.
  • 0 in reply to Paul Grohe
    TI__Guru**** 244380 points
    Hello Paul, Zak,

    An additional question: Addition of the 1800 pF cap in the feedback at high frequency should cause the Rf to be 0 and thus there should not be any gain at 4.8 MHz. Is there something I am missing!

    Thanks for the alternate part number.
  • 0 in reply to Amit Ashara
    TI__Mastermind 23050 points
    Amit,

    I apologize for the confusion. It is true that the closed loop gain continues to roll-off indefinitely in an inverting configuration, however the feedback factor of the op amp does not increase indefinitely and will approach unity as the impedance of the cap decreases with frequency. From the perspective of the op amp, once the cap shorts out it is operating as a unity gain buffer to the voltage on the non-inverting pin. This is the gain I was referring to for stability analysis, NOT the closed loop gain from input to output. If you are interested in learning more about how to evaluate op amp stability, there are some useful presentations on the subject that can be found here: training.ti.com/ti-precision-labs-op-amps-stability-1
  • 0 in reply to Zak Kaye
    TI__Guru**** 244380 points

    Hello Zak

    Thanks for the link. This is a very useful presentation. Also I would be doing a board mod to place the compensation circuit and will verify.

    Hello Paul,

    I replaced LMP7708 with LMP7702 and the ~4.8 MHz oscillations are no longer present and the response of the sensor has improved. I really appreciate the help here.