• Locked Locked
  • Replies 5 replies
  • Subscribers 31 subscribers
  • Views 1610 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.

Mistake in "A Differential Op-Amp Circuit Collection" sloa064a.pdf

Steve Jeffrey
Prodigy 30 points

Hi, I found a mistake in Section 3.2.2 Akerberg Mossberg Filter.

In Figures 9 through 12 the integrator capacitor C1A is connected from inverting output to inverting input, C1B is connected from non-inverting output to non-inverting input, and the result is positive feedback.  You can see that by inspection but I also simulated it and found the AC analysis looked correct but the transient analysis quickly saturated at the rails as expected.

When you fix the mistake you end up with the Differential Biquad Filter shown in Figure 13 of the next section.  Calling  Figures 9 through 12 Akberg Mossberg filters is stretching the definition anyway; with a fully differential op-amp the non-inverting integrator is exactly the same as the inverting integrator.

Steve Jeffrey

  • TI__Guru**** 189399 points

    Hi, Steve,

    First off, welcome to e2e, and thank you for your interest in our products!!!

    Thank you for the feedback on this document. In the next revision cycle, we will include your fixes.

    -d2

  • in reply to Don Dapkus
    Prodigy 20 points

    Steve & Don,
    Also found a problem with fully differential filters, both in this application note and FilterPro. In all configuration of a HighPass diff. filter, there are a ~ +30-40dB pole in the MHz region. (My simulations using 1 stage / 2 order).
    In all of my configurations - before and after this HighPass filter -, 1MHz is clear, straight and definetly needed. Any advise about how to solve this problem without using other filters / stage.

  • in reply to Thomas Littmann
    Prodigy 30 points

    Thomas,

    I was able to duplicate your resonance problem with the fully differential high pass filter, and then I duplicated it using a single ended filter because it's easier to analyze.  It looks like it's a problem with filter topology caused by the non-ideal performance of the op-amps, but I don't know exactly what.

    SLOA049A describes the non-ideal performance of  Sallen-Key and MFB low pass filters, so I tried adding an RC filter to the output of the amplifier as shown in SLOA049A.  That helped, but not enough.  I then found that adding a small resistor in series with the input damped out the resonance very well, while the RC filter on the output rolls off the response at higher frequencies.

    This was all using ideal capacitors, once you put in the actual ESL and ESR values it could get worse.

    Steve

  • in reply to Steve Jeffrey
    Prodigy 20 points
    Hi Steve,

    Thank you for your answer, i also made these simulations in LTSpice and had the same results and conclusions.
    Thanks again.

    Thomas
  • in reply to Don Dapkus
    Prodigy 25 points

    Hi Don, 

    So any news on when this app note will be revised? It has been over three years now since Steve brought it to your attention, and I'm sure dozens of people have been bitten by these errors (including myself). I found at least two other websites calling out these errors. It would be nice if you either pulled that section or otherwise addressed this.

    Thanks for listening.

    -George