• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 438 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.

OPA1632: Proper way to tie source and OPA1632 grounds?

Supira Medical
Intellectual 300 points
Part Number: OPA1632
Other Parts Discussed in Thread: TPA3251, , TPA3251EVM, TPA3255, TPA3250, OPA1637, OPA1633

Tool/software:

Hi,

I have a question regarding the correct way to tie the grounds of the source input and the OPA1632 front end as shown in the Audio-OPA1632EVM. The related power amplifier in my application is the TPA3251 and its EVM. This is specific to single-ended inputs using the RCA jacks.

In the TPA3251EVM, the source grounds (RCA jacks) are AGND1 and AGND2. These are net-tied to the TPA3251 ground pins adjacent to AVDD.

The Audio-OPA1632EVM says that C17 must be connected to source ground. How does one accomplish a ground connection between the source ground and OPA1632 ground? Is there a recommended net-tie scheme?

Secondarily, I’d like to confirm that C17 indeed stays in the circuit and it’s not R12 that gets connected to source ground.

Thanks for your assistance.

  • 0
    TI__Mastermind 23819 points

    Hello Supira Medical,

    Please allow me some time this week to review the EVMs and provide you a definite answer.

    Thank you for reaching out for technical support.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Intellectual 300 points

    Hi Alec,

    Thanks for looking into this. Grounding can be tricky!

    FYI - I took a look at the EVM's for the TPA3250 and TPA3255 and interestingly, neither of these designs bothers with separate analog grounds the way the TPA3251 does. Curious, indeed.

    Regarding my second question about the capacitor (C17), I think it does stay in the circuit, as shown in Figure 5 of SLOA072.

  • 0 in reply to Supira Medical
    TI__Mastermind 23819 points

    Hello Supira Medical,

    I imagine we are observing a difference in engineering preference between EVM designers.  I am looking into what is best for OPA1632 today.

    As for C17, I agree to keep it (also in reference to SLOA072).

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Intellectual 300 points

    Thanks, Alec. I may use the OPA1637 but I think the OPA1632 findings will be equally applicable.

  • 0 in reply to Supira Medical
    TI__Mastermind 23819 points

    Hello Supira Medical,

    Okay, I will send over my thoughts once I have them.  We also have OPA1633 as a newer device with upgraded performance.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Intellectual 300 points

    Hi Alec,

    I've looked at the OPA1633 but wasn't sure if it was better than the super-beta 1637. What I like in particular about the 1637 is the much lower current, which results in lower temperatures. I've read that the 1632/1633 can get very hot.

  • 0 in reply to Supira Medical
    TI__Mastermind 23819 points

    Hello Supira Medical,

    My understanding of the differences in grounding is this: for OPA1632, there is a single L channel and single R channel.  These can use the same GND connection, and have a common GND net (no split GND plane or separate ground pours).  For TPA3251EVM, this EVM has two separate signal paths, each with their own L and R channel.  AGND1 corresponds to the GND for the first stereo path, while AGND2 corresponds to the GND for the second stereo path.  This may be due to the current return path present in both circuits: keeping the AGNDs separate reduces crosstalk and noise from one signal path affecting the other.

    On the OPA1632, there are only two inputs, rather than four.  You may follow your intuition to use the single GND net here.

    Your reasoning for OPA1637 is good, you may use this device if you have concerns about OPA1632/33 heating.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Intellectual 300 points

    Hi Alec,

    Thank you for looking into it. The OPA1632 is not configured as having a single L and single R channel. Two OPA's are used, one per channel since it's a SE to diff converter here. Therefore, the configuration issue for grounding remains the same. What I see as being different is that for either EVM, be it the TPA3251 or TPA3255 that have the AIB connector, the use of the OPA1632 input board changes the grounding.

    If you use the direct inputs on the TPA EVM, the AGND's come into play but if you use the OPA1632, the AGND's get taken out of the circuit. Perhaps out of an abundance of caution, I should retain the AGND scheme for the OPA1632 and put them on the TPA PCB instead of a separate input PCB. It's a loose end rattling around in my brain however as I'd be doing something without really understanding it.

    Thanks for taking the time to dig into it!