• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 47 subscribers
  • Views 682 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Original question:

OPA855: Question About Using OPA855 for Differential TIA Design

OPA855: Question About Using OPA855 for Differential TIA Design

문 찬희
Prodigy 40 points
Part Number: OPA855
Other Parts Discussed in Thread: OPA856, TINA-TI

Tool/software:

Hello,

I am currently designing a differential transimpedance amplifier (TIA) using the OPA855. My design is based on the paper "Wide-Bandwidth Low-Noise Simple Differential Transimpedance Amplifier Ideal for Detecting High-Frequency Faint Optical Signals" by Zhiyuan Wu, Zeyang Wen, Wenqiang Li, and Gongliu Yang.

However, I am facing an issue where oscillation occurs at the output. I used the feedback capacitor calculator provided by TI as a reference, but no matter how I adjust the capacitor value, the oscillation (around 500 MHz) persists. The only change observed is in the frequency range where peaking occurs in the output FFT.

When I simulated the same circuit in TINA, it was stable, as shown in the attached image.

and the image below shows the FFT of the output when only the supply is applied to the current board. It confirms that oscillation exists in the hundreds of MHz range.

To identify the root cause, I looked into TI’s forum and came across a response from kai klaas69 mentioning that the gain stability condition (7V/V) of the OPA855 was violated in a similar case. The link to the post is as follows:
TI forum link: e2e.ti.com/.../opa855dsgevm-high-gain-peaking-on-evm

In my current TIA board, the two input terminals are separated by approximately 8mm, but this gap is not purely copper traces. It also includes a DC block capacitor (0805 size) for APD bias isolation, as well as the feedback resistor and capacitor(0402 size).

To ensure the noise gain stability condition (>7V/V) is met, I intentionally added C_cm between both input terminals and tested values ranging from 1pF to 5pF in 1pF increments. However, the oscillation remained unchanged.

I intentionally increased C_cm, which is included in C_TOT, to ensure that the noise gain exceeds 7V/V.

Additionally, I tried reducing the OPA855-APD path length by physically moving the APD closer, but the results were the same (this was only for testing, not intended for actual use). This led me to believe that violating the noise gain stability condition may not be the root cause of the oscillation.

Interestingly, when I replaced the OPA855 with the OPA856, which has a lower GBWP but is stable at 1V/V, the oscillation disappeared. I believe this was simply due to the reduction in bandwidth.

So far, I have attempted:

  • Adjusting the feedback capacitor
  • Ensuring the noise gain stability condition is met

Despite these efforts, the oscillation persists, and I am unsure of what to try next. I would greatly appreciate any insights or advice from experienced engineers.

Best regards,

Moon

  • 0
    TI__Intellectual 2065 points

    Hi, will look into and get back to you.

    Regards,

    Aditya Gosavi

  • 0 in reply to Aditya Gosavi
    TI__Intellectual 2065 points

    Hi Moon,

    I read the document that you sent above, it mentions the differential TIA circuit as below-

    However, I could not understand what circuit configuration you have wired up in TINA-TI. I see that you have also AC coupled the TIA.

    Stability wise, you can try to add a small (1-5pF) capacitance directly at IN- to GND of OPA855, and check the stability.

    Regards,
    Aditya Gosavi

  • 0 in reply to Aditya Gosavi
    Prodigy 40 points

    Hi Aditya Gosavi

    Thanks for your reply.

    In the actual implementation, a 105V reverse bias is applied to the APD, and a capacitor was added across its terminals to AC-couple the bias.
    Could this affect the simulation results?


    Regards,
    Chanhui Moon

  • +1 in reply to 문 찬희
    TI__Intellectual 2065 points

    Hi Chanhui,

    I couldn't understand why you added a capacitor across the APD terminals, because you already have AC coupling caps C1 and C2 (100nF) in your TINA schematic. I also believe the 1.2pF C5 cap in the schematic is the APD junction capacitance.

    With the original circuit in your post above, I conducted some simulations and it turns out that it is unstable.

    Circuit stability cannot be fully judged based on AC response. As you mentioned in the post above "When I simulated the same circuit in TINA, it was stable, as shown in the attached image.", this is not fully correct, because the AC response show some peaking, which is a sign of instability. To check if a circuit is stable, you need to perform the open-loop stability analysis in TINA, which will fetch us phase and gain margin values.

    Please refer these resources for stability analysis- www.ti.com/.../4080235259001

    From the stability analysis, you can see the bode plot for closed-loop gain curve (Aol*Beta curve) above. The system has negative phase and gain margin, thus the system is unstable.

    You can see below transient response also, which is oscillating-

    To improve the stability I would suggest you to add an input capacitor at the IN- or inverting input of the op-amp to GND. I see you have tried to add a Ccm cap in your circuit, but that will not work because there are other circuit components like resistors in that path, which will not increase your Ctot cap. So I would suggest you to add a 2pF cap (or even higher) from the IN- pin to GND directly, and check the stability.

    With 2pF input cap added the phase margin is 46 degrees. I have attached the transient response and closed-loop gain bode plots below-

     

    Please find the attached TINA schematics below-

    opa855 diff tia e2e.TSC

    opa855 diff tia stability e2e.TSC

    Regards,
    Aditya Gosavi

  • 0 in reply to Aditya Gosavi
    Prodigy 40 points

    Thank you very much for your kind and detailed explanation. It was very helpful for my understanding.

    The capacitor labeled C5 corresponds to the APD junction capacitance.

    I also reviewed the stability analysis material you shared.

    One part I still don’t fully understand is the reason for adding the red-circled section in the schematic during the open-loop stability analysis.
    Is it correct that this part was added to create a separate AC analysis loop while maintaining the actual feedback loop?

    Best regards,
    Chanhui Moon

  • 0 in reply to 문 찬희
    TI__Intellectual 2065 points

    Hi Changui,

    The red part is added so that the op-amp’s output sees the feedback network during AC analysis. During AC analysis, the inductor is open circuit and the op-amp is cutoff from the feedback network.

    Regards,

    Aditya