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.

TINA/Spice/THS4631: Ask problems about applications of high-speed op amp

Part Number: THS4631
Other Parts Discussed in Thread: TINA-TI, , THS3062, THS4021

Tool/software: TINA-TI or Spice Models

Dear ,

I have some problems about application problems about high-speed op amp from TI.

In my last post, https://e2e.ti.com/support/amplifiers/high_speed_amplifiers/f/10/t/652730, I had doubt about the big mismatch between datasheet and the simulation results in terms of THS4631’s capacitance load drive performance. And I’m replied that the experiment data should be followed. I have 2 problems to consult you.

1.About capacitive load drive performance. In my application, I’ll connect op amp’s output to NPN-PNP pairs (emitter-follower configuration). Since NPN-PNP pairs' base can be modelled as a capacitor up to 10nF and TINA-TI’s simulation results can't be trusted, what can I do to evaluate the capacitance driving performance (compensated by isolated resistor)? Is it necssary that I buy an evaluation board?

2.I totally have no experience in high-speed op amp use. But due to the easy implementation and pure resistive external components of noninverting and inverting gain configuration, I have confidence in using them. As to analog PI(proportion-integration), which involves the use of capacitor in the feedback network and maybe other compensation capacitors for circuit stability, I fear that I can’t do this job. Can you please offer instructions to help me with PI application with high-speed op amp (THS3062, THS4631 and so on)?

Best regards!

Yatao

  • Hi Yatao,

    I will help you on this issue. Unfortunately TINA is not capable of capturing full effects of capacitive loading in this case and an EVM would be the best way to test the device with your transistor loads.

    For an integration circuit you can use a standard integrator circuit with a high speed amplifier. As long as you choose an amplifier that is unity gain stable, then it shouldn't present any additional stability issues due to the circuit configuration. However, you do need to make sure you have high speed PCB layout that doesn't add additional parasitics that can effect stability.

    Regards,
  • Hi Jacob,

    Thanks for you answer and sorry to reply late.

    I have my 2 problems and I’ll wait for your answer.

    1.Can you tell me, in common applications, how large capacitance high-speed op amp can drive? I see such descriptions as Fig.1 in nearly all high-speed op amp’s datasheet and Fig1 is screen captured from THS3062’s datasheet, pp17/33. I always doubt, to ensure stability and bandwidth, high-speed op amp can drive capacitors only in the range of tens of pF(note the 10pF in Fig1)?

    Fig 1.

    2.You say that due to the proportion-integration circuit configuration, with high-speed PCB layout and unity gain stability, the high-speed op amp shouldn’t display stability issues. So I can use THS4621(VFA, unity gain stable) in the configuration in Fig2 (noninverting configuration, and except R1, Rc1, Rc2,Cc, no other R or C are needed for this PI regulator?), but not THS4021(VFA, not unity gain stable)? And what about the CFA THS3062? I would refer to the datasheet about the high-speed PCB layout, is this just ok? Finally, assuming that my PCB is produced, how can I know whether Fig2’s PI regulation meets its design function  -  Proportion and Integration of its input Vi? Because only when I’m sure it works, then I can run the whole closed-loop control system.

    Fig 2.

     Best regards!

    Yatao

  • Hi Jacob,
    Can you spare some time to ask my problems?Thx.

    Best regards!
    Yatao Ling
  • Hi Yatao,

    Sorry for the delay in reply.

    To answer your questions:
    1. The large capacitance drive of a high speed op-amp primarily depends upon the stability and slew rate (V/us) of the op-amp. In Fig 1 you posted, it says that the THS3062 certainly can drive greater than 10pF load cap as long as there is enough series isolation resistor between the op-amp output and the cap so as to not causing any peaking or oscillation. You should be able to drive a 1nF cap using the THS3062 by adding an 8-ohm series isolation resistor.
    2. Yes, you are right. As Jacob mentioned, it is recommended to use the THS4621 which is unity gain stable instead of the THS4021 which is not unity gain stable for the PI regulator configuration. I would not recommend using the CFA THS3062 because the CFA stability depends upon the feedback impedance which changes across frequency from the circuit you have posted in Fig 2. You can use suggestions in the datasheet for high speed PCB layout. The working of the PI regulator to meet its design function totally depends upon your application. I am not sure we can help there, unless you provide more information about what you are trying to achieve with this circuit. I would recommend that you simulate your circuit in TINA-TI before implementing on the pcb to understand how the closed loop control system works.

    Best Regards,
    Rohit
  • Thanks for your reply.
    Then, you mean that CFA isn't fit for analog proportion-integration application?
  • Hi Yatao,

    Yes, I think it is usually not recommended to use CFA for analog proportion-integration application. However, I would recommend you to simulate your circuit in TINA-TI before completely rejecting the idea.

    The reason for not recommending CFA is primarily because in PI application circuit you have shown, if the Rc2 value happens to be very small then you might end up running into stability problems at high frequencies because the Cc looks like a short at high frequencies. For a CFA, the feedback impedance (Rc2 + 1/sCc) is what's going to determine the stability of the circuit. You may want to take a look at the attached pdf for loop gain stability analysis of a CFA: CFB_LoopGainAnalysis.pdf

    Best Regards,

    Rohit

  • Hi Rihit,
    Thx