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.

THS4520: Stability issues THS4520

Part Number: THS4520
Other Parts Discussed in Thread: TINA-TI

I am having trouble using the THS4520 as an MFB differential high-pass filter. I am trying to use the device to stack the same active-filter for hard roll-off, schematic below.

I am finding that a single stage functions as expected when isolated from the 2nd stage. However, as soon as I populate C10 and C11 on my board the circuit becomes unstable. An oscillating signal of fixed amplitude, in the range of 120MHz to 140MHz appears at both complementary outputs of the first stage. This is then amplified by the second stage and so on…

Any ideas what could be causing the instability and how to solve it?

 

  • Hi Carlos,

    Could you attach your TINA-TI file? I think you may need some isolation resistance for driving the input capacitance of the second stage. You can measure the closed loop input impedance of the first stage (inductive), and graph this against the or second stage input impedance (capacitor), over frequency and look for a resonant intersection point. Then you can determine what minimum isolation resistance to add.

    Best regards,

    Sean
  • Hi Carlos,

    Sean is totally right. Take care, MFB high pass filters are notoriusly instable and should be avoided whenever possible!!! Because a direct capacitive path exists between each input and output (C4, C1, C6, C2, C10, C7, C11 and C8) all involved OPAmps see huge capacitive loads which results in an erosion of their phase margins.

    I have seen designs where the instability was attempted to fix by adding resistances in series to some of the capacitors of MFB high pass filter. In your application these would be C4, C1, C6, C2, C10, C7, C11 and C8. You can try that procedure. But I would heavily recommend to avoid this MFB high pass filter topology at all!

    Kai