• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 396 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.

OPA858: Can the NC pin be connected to ground

Shawn McConomy
Prodigy 10 points
Part Number: OPA858


Tool/software:

I am trying to reduce the capacitance across the feedback resistor.  I have read that placing a ground trace between the feedback resistor pads would do this.  I would also like to continue that trace between the inverting input and the feedback pins.  between those pins is the NC pin.  It would be useful to be able to connect that NC pin to ground to imp[prove the shielding.

  • 0
    TI__Mastermind 23819 points

    Hello Shawn,

    If you are able, could you draw or provide an image of the exact trace placment you are describing?

    For high speed devices in general, including our OPA85x family, we recommend GND and power plane cut-outs along the input and output signal paths.  This includes our Rf feedback resistor and the inverting input pin of the OPA858.  The Layout Guidelines section of the datasheet (Section 12.1) shows a cut-out for GND and power underneath Rf.  I m not sure what best practice you are referring to; I would like to learn more.

    We design evaluation kits/boards, known as EVMs, for many of our devices & device families.  The OPA858 has an EVM, the OPA858DSGEVM, which can be sample and ordered off of TI.com.  The EVM has a companion user's guide which includes PCB layers and cut-outs as well as other supporting information.

    Placing a single trace could introduce board parasitics or provide an opportunity for noise to couple into the input circuitry.  I would be happy to help review your layout and help debug any behavior you are observing if this is a revision discussion.  If you are doing a new design, I can review the schematic and layout if you would like me to do so.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Prodigy 10 points

    The main part of my question is can I connect Ground to the NC pin of the OPA858?

  • 0 in reply to Shawn McConomy
    TI__Mastermind 23819 points

    Hello Shawn,

    From a strictly mount & bond diagram perspective, the NC pin is not bonded to the die and has no electrical connection to the OPA858.

    The NC pin exists to create isolation between the IN- and FB pins; this is the reasoning for why we strongly recommend to leave the pin floating.  The NC pin helps reduce capacitive coupling between IN- and FB pins.  By increasing the spacing between IN- and FB pins by an additional in-between 'pin', the NC pin helps minimize unwanted effects and creates a distance suitable for a small package surface-mount RF resistor.

    Section 9.3.2 states what I provided above in datasheet language/terms.

    From my understanding & TI's support of the OPA858 I cannot recommend connecting the NC pin to GND.  I would have concern if your goal is to improve capacitance across the feedback resistor (IN- and FB pins), as floating the NC pin already reduces capacitive coupling. If you do have a source and support that details how the proposed GND trace will help reduce capacitance, you may take that into account for your decision.

    I would be curious to know the background to this method; I do apologize as I do not have firsthand experience with adding/removing a GND trace to OPA858 NC pin.  As such I can only offer our TI recommendations for layout & understanding the NC pin. 

    In summary, there is not an electrical connection between NC and OPA858 die.  This lack of electrical path does not mean there would not be a capacitive or inductive effect from the decision to use a GND trace.  The potential negative effect may outweigh the proposed benefit.

    Please let me know what you find if you decide to utilize additional techniques for reducing capacitance and capacitive coupling.  It would be good to learn from the experience, as the boards I use in-lab are designed to EVM spec.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Prodigy 10 points

    Thank you for your reply.  I am basing the reduction on capacitance with the ground trace on the article link below.  There are other conversations online that talk about a ground ring that works similar

    https://www.analog.com/en/analog-dialogue/articles/how-to-cancel-ambient-light-for-lidar-receivers.html

    We have used the OPA858 on other TIA designs but are experiencing more noise than our models predict.  We are trying multiple things to see what if anything would help with the noise problem.

  • 0 in reply to Shawn McConomy
    TI__Mastermind 23819 points

    Hello Shawn,

    Ahh, thank you!  I can help support with investigating noise contribution and other effects, if you would like to share more about your design and system.  We can always connect over email for privacy if you need to share schematics/etc.

    May I ask what amount/level of noise you are measuring?  

    I will look over the article, thank you for that also!

    Overall my team and I are happy to help support your continued use of OPA858.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    TI__Mastermind 23819 points

    Hello Shawn,

    I just spoke with our team on this.  There is actually a bit of a contrasting approach here; since the pin is truly not connected, we do often recommend grounding the NC pin for shielding and capacitance reduction.  The datasheet language is overly harsh on restricting the NC pin use to 'must be left floating'. 

    My prior messages were true from the datasheet & product intent, but in practice good grounding and isolation will help a design.  As the NC pin exists to improve noise and capacitive coupling, adding a ground stub (your GND trace) to the circuit may help resolve the observed noise increase.

    I apologize for the confusion, it turns out high speed amplifiers are complicated Slight smile.

    Please connect the NC pin to GND and proceed with your design; I am curious to hear back on an improvement.   Thank you for bearing with my investigation today.

    Best,

    Alec