• State
  • Replies 6 replies
  • Subscribers 48 subscribers
  • Views 182 views
  • Users 0 members are here
Support feedback
Options
Options
Related

OPA657: OPA657 output noise/oscillation

yi yang Hu
Prodigy 10 points
Part Number: OPA657
Other Parts Discussed in Thread: OPA656

Tool/software:

Later I tried various methods, increasing the mask and reducing the noise to 1.5mVpp.

I've designed an interface circuit using an OPA657, but I'm experiencing some noise during actual measurements. The circuit has two inputs. I initially connected both inputs to ground using SMA shorting capacitors. I managed to detect approximately 30mVpp for a 60Hz signal. I assumed it was due to the room voltage, so I used aluminum foil as a shield, reducing the noise to 4mVpp.

         

Later I tried various methods, increasing the mask and reducing the noise to 1.5mVpp.

I'm interested in the noise behind my oscilloscope. Can I reduce the noise outside of the chamber? What's the best way to reduce it? Below is my schematic.

I would also like to ask about your knowledge of layout?

  • 0
    TI__Mastermind 29670 points

    Hello Yi Yang Hu,

      Thank you for sharing your debugging steps! Quick questions for your setup: 

    1. On board, your U1 and U2 are both shorted to ground/resistor to ground at non-inverting input?
    2. Do you see the noisy signal in scope shot 3 (below) when inputs are grounded at nodes VA1_out and VA2_out?  
    3. If first scope shot and third scope shot are without input as suggested in 1) and 2), then it looks like oscillations due to amplifier instability. It would usually show up as very high frequency in hundreds of MHz:

       Looking at your gain setting, it looks the device to be configured in a 5.7V/V signal and noise gain. This is slightly lower than the stability requirement for this device of 7V/V. The OPA657 is a unique voltage feedback amplifier where it is decompensated to achieve a high bandwidth. The OPA657 (decompensated, only stable at noise gains of 7V/V) has a GBW of 1.6GHz, compared to its unity gain stable version OPA656 which has a GBW of 230MHz and UGBW of 550MHz. If you would like to learn more about decompensated amplifiers here is a quick app note: https://www.ti.com/lit/an/slyt174/slyt174.pdf?

       Would you be able to try setting the OPA657 in a gain of 7V/V: RF = 453Ohms and RG = 75.5Ohms, and let us know if that fixed your issue?

    Thank you,
    Sima

  • 0
    Prodigy 10 points

    Hello Sima ,

    I'm currently redesigning my component layout. The documentation suggests placing backflip resistors and capacitors below the components, but most sources recommend placing them on the top layer. Could you give me some advice?

  • 0 in reply to yi yang Hu
    TI__Mastermind 29670 points

    Hello Yi Yang Hu,

       You are correct. You should always place resistors and capacitors on the top layer. The only capacitors that can be placed on bottom layer are your decoupling capacitor to your power supplies with the smaller value capacitor right at the amplifier supply pins if there is no room on top layer or will conflict with your important signal path components such as your feedback resistor/feedback capacitance.

    Thank you,
    Sima

  • 0 in reply to Sima Jalaleddine
    Prodigy 10 points

    I understand. So I should place the backflush resistor and capacitor as close to the -in position as possible, even though this will make the wire connected to the out position longer, is that okay? Do you think this layout is acceptable?  

       

  • 0 in reply to Sima Jalaleddine
    Prodigy 10 points

    HI,Sima

    If there's very little space on the top layer today, would you suggest I place the backflush resistor and capacitor on the bottom layer? I have a differential amplifier set today with very little space, and I can't think of a good way to place it on the top layer.

  • 0 in reply to yi yang Hu
    TI__Genius 17775 points

    Hi Yi Yang Hu,

    Sima is out of office, but I was able to look over the circuit you are trying to implement using the OPA657 and I am concerned with the stability of the circuit in its current form. As this is a decompensated, the feedback capacitor of 1.5nF would almost certainly be an issue as this would affect your noise gain at higher frequencies leading to instability. I am not sure if you have simulated this circuit, but I am not sure this would work with this device.

    Best Regards,

    Ignacio