• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 48 subscribers
  • Views 482 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.

Vref question on Parallel ADS1278's

Don Ransbury
Prodigy 70 points

Other Parts Discussed in Thread: ADS1278, OPA350, OPA1632, THS4521

I want to parallel 10 ADS1278's across two boards.  The phase relationship of the data is important for my digital processing.  I am concerned about driving the Vref input for that many IC's, and also across the added impedance of the board connector.  Can you tell me the input characteristics of the Vrefp/Vrefn inputs? 

  • 0
    TI__Guru**** 182601 points

    Hi Don,

    We're on break for the Thanksgiving Holiday here in the US through Monday next week.  We'll get back to you on this question as soon as possible.

  • 0
    TI__Mastermind 43915 points

    Hi Don,

    Sorry for the delay. This got lost in the shuffle with Thanksgiving.

    The ADS1278 does not have a high impedance reference input explaining that their is not an internal buffer. I have done some testing on the ADS1278 to see how the device samples off of the reference to understand how important the external reference drive circuitry is. The conclusion was the the Vref is double sampled with each modulator clock and the amount of instantaneous current that it pulls is dependent on how many of the eight channels are powered up. When all eight channels are powered up, the current required at 2xfmod is at its highest. Long story short, the conclusions were that a 10uF cap right at the the reference pin of the device is recommended to help restore the charge each time the reference is sampled. I would probably have a buffer for every 2-3 ADS1278s as well. Since you need 10uF at each Vref of the ADS1278, you will need a buffer than can drive a good amount of capacitive load which is why we highlight the OPA350 in our datasheets. You cannot afford to have any droop on the reference line during the sampling process when having 10 ADS1278s sample in parallel which is why using one REF device, maybe the REF5025, with 3-4 separate op amp buffers, OPA350s, to drive 2-3 ADCs with 10uF caps, would be what I would recommend.

    Let me know if I can help further.

    Regards,

    Tony Calabria

  • 0 in reply to Tony Calabria
    Prodigy 70 points

    Hi Tony,

    Thanks for the input. It's very helpful.  I plan on running the ADS1278 at 144Khz Fs (with a 37Mhz CLK).  I like the REF5020 (with Vout=2.048), but since I need to also drive DVdd at 2.1, I was thinking of tying the DVdd input to the Vref input.  It's unclear to me how the Vref and Vcom work together to support the input buffer.  The Vcom just looks like a resistor divider from the DVdd input, and Vref looks like it acts as an offset to the input voltage conversion.  Since I'm a bit worried about my DC term (I would really like it to go to zero if possible), it just seems like I could tie the two together to null out any DC "better"(?).  Then I could run Vcom to an input DiffAmp Vocm input and match the diffamp to the ADC. 

    I like your suggestion of using the OPA350 to buffer Vref, is there a Differential buffer equivalent?

    -d

  • 0 in reply to Don Ransbury
    TI__Mastermind 43915 points

    Hi Don,

    We do not really have an equivalent differential out amplifier to the OPA350. Out of the PA line, our differential output amplifiers fall in the OPA16xx family. I am not exactly sure why you would need a differential out amplifier with the ADS1278 reference. You should be able to use the OPA350 to drive the reference. A differential amplifier such as the OPA1632 or THS4521 would be differential options for the ADC drive circuitry.

    Vcom is a resistor divider off of the analog power supply used to set the common mode point for the drive amplifier. We used the analog power supply as it is the limiting factor for the common mode input range. It is completely independent of the reference voltage and designed to be routed into the Vcom in of a differential amplifier.

    I probably would not tie Vref directly to DVDD. You want to keep Vref as clean as possible and should be independent of the digital power supply. You could use a REF5020 and use a separate amplifier to isolate it in parallel with the other amps (OPA350s used to drive the reference pins) to drive the DVDD of all the ADCs.

    Regards,

    Tony Calabria