• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 46 subscribers
  • Views 178 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.

ADS9227: VCMOUT Buffering

Gordon McKenzie
Intellectual 610 points
Part Number: ADS9227
Other Parts Discussed in Thread: THS4541, , OPA320

Tool/software:

Looking at the ADS9227 datasheet, the application circuits in figs. 9.1, 9.4 and 9.7 all show the VCMOUT pin voltage being buffered though an unspecified op-amp, rather than being connected directly to the common mode input of the fully differential driver (THS4541). Two questions from this:

1. What benefit does the op-amp buffer give? I assume the VCMOUT voltage cannot drive much current but don't see any value listed in the datasheet.

2. If the buffer is needed, does it have any particular requirements? I assume VCMOUT is fairly steady and that something with low GBW would be OK but wondered if, perhaps, this pin sees any effects of sampling glitches. If I sample at 5MSPS, would you expect it to be safe enough to use something pretty basic with GBW of about 1MHz and decent noise spec, as long as it's unity gain stable?

  • +1
    TI__Expert 3850 points

    Hello Gordon, 

    Thank you for posting on TI's E2E forum! 

    We do recommend that the vcmout have a buffer when used externally. To answer your question:

    1. The VCMOUT buffer is only able to drive up to 5uA, having a buffer is a good way to ensure that limit is not reached.  
    2. As far as buffer needed, that should be more tailored to the external circuitry it is being used in,  but typically we recommend the OPA320.
      1. Like you mentioned, the VCMOUT should remain stable and it also is closely tied to the precision of the measurements, the main concern would be precision and stability. 
      2. The OPA320 is a good option because its a precision OpAmp with low noise, unity-gain-stable, high CMRR, low-offset, output current drive, amongst other great features. 
      3. Overall, Yes, 1MHz should be sufficient if its unity-gain stable, low-noise/offset, can drive the current and capacitance of the intended driver (and anything else tied to it). 

    Best regards, 

    Yolanda 

  • 0 in reply to Yolanda Gonzalez
    Intellectual 610 points

    Thanks Yolanda.

    That all makes sense. Not quite sure how I missed the current spec from the table you've screenshotted as I've been back and forward to that a few times today to look at the voltage limits!

    From that, definitely worth using a buffer. I'll find something suitable, and will keep the OPA320 in mind.

  • 0 in reply to Gordon McKenzie
    TI__Expert 3850 points

    Hello Gordon, 

    It happens! It definitely has to me every now and then (ctrl + F) is a life saver.  Glad the concern was answered though. 

    If you would like to share more information about your system I would be happy to make additional suggestions for vocm drivers or anything else involving the signal path to the ADS9227. 

    Best regards, 

    Yolanda

  • 0 in reply to Yolanda Gonzalez
    Intellectual 610 points

    Hi Yolanda,

    I just notied your latest reply now. I'm looking to use the ADS9227 in a design update, to deal with a single-ended signal that has bandwidth up to 1MHz. I have a low-noise front end, mux and variable gain stage which I'll reuse from the existing design. The application circuit in figure 9.7 of the datasheet looks like it would be OK - I was planning on giving that a quick simulation in spice to check for stability but it should be usable as shown. The GBW of the THS4541 that it suggests is probably a bit higher than necessary but I'm not hugely concerned about current consumption when the board is awake so I don't think it'll matter too much.

    However, the ADC and driver and new and not fixed yet, so if there's a more appropriate way of driving it, happy to consider.

    Thanks.

  • +1 in reply to Gordon McKenzie
    TI__Expert 3850 points

    Hello Gordon,

    Very happy to hear that you are considering the ADS9227 for a design update! 

    The circuit in Figure 9-7 would be a great option for a 1MHz bw input signal, the only thing I would advise to ensure to consider adding a step for impedance matching since it is single ended and there is a mux and variable gain stage prior.  

    If you would like any help confirming your full signal chains, please feel free to share any schematics and system needs and I'd happy to help review them with you. 

    Best regards, 

    Yolanda  

  • 0 in reply to Yolanda Gonzalez
    Intellectual 610 points

    Hi Yolanda, that all sounds fine. The signal chain before the ADC driver is being reused and was previously used at higher frequencies so I think it'll be OK.

    Thanks for your help on this, much apprecieated.