• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 5 answers
  • Subscribers 46 subscribers
  • Views 642 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.

ADC083000: Full Power Bandwidth Questions

Joseph Murray
Prodigy 40 points
Part Number: ADC083000
Other Parts Discussed in Thread: ADS54RF63, ADC08D1520, , ADC12QJ1600

Hello -

We are designing a ADC unit that requires a very wide full power bandwidth. We are considering the ADC083000, ADC08d1520 or the ADS54RF63. In our application we don't require sampling above about 500MHz but we require about 4GHz of full power bandwidth. The 3 models we are looking at have 3 GHz, 2 GHz and 2.3GHz respectively. We are trying to determine how much attenuation happens above the listed full power bandwidth? Does any data exist above the listings in the data sheets?

Thanks -Joe.

  • 0
    TI__Mastermind 39330 points

    Hi Joseph,

    I can certainly take some data on this to understand the rolloff on one of these devices, but there are other concerns that can arise when using a converter past its full power BW rating.

    The variability of performance becomes much less predictable. Variations in process limit us from adequately predicting the performance of the converter. This is due to the design of the sampling switch within the converter. What ends up happening, is that the converter doesn't settle in time for the next sample, and this translates into linearity issues, ie - SFDR for frequency applications. For time domain applications, the variability would be how fast you can adequately capture a pulse and its characteristics, like slew and settling.

    Looking at this on one device, can give you an idea, but might not tell the whole story when going into production. Would it be possible to consider another device?

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    Prodigy 40 points

    Thanks Rob -

    Our application is very unusual. We only anticipate sampling the ADC at ~100 - ~400 MSPS. Our signal will be aliased - we are using an aliased sampling technique.

    The nature of our work does not allow me to elaborate too much on what we are doing with these devices. 

    We are open to really any device. We are strongly considering the ADS54RF63 that has 2.3GHz Full Power Bandwidth. The only disappointing aspect is that our SWaP-C is going go up dramatically considering that we would then cut from a 4GHz band down to a 2GHz processing band. We don't really need all of the extra dynamic range of this device. We are considering the following devices (but open to others): 

    1. ADC083000: operational configuration with 4GHz input bandwidth, sampling between ~100 - 400MSPS. Issue: beyond the published full power bandwidth of the device (3GHz). So the signal estimation at each sample epoch is not clear.
    2. ADC08D1520: operational configuration with 4GHz input bandwidth, sampling between ~100 - 400MSPS. Issue: even further beyond published full power bandwidth of the device (2GHz). However, the price point with the QTY required is more attractive.
    3. ADS54RF63: operational configuration with 2GHz input bandwidth, sampling between ~100 - 400MSPS. Issue: within the published full power bandwidth of the device (2.3GHz). However, SWaP goes up dramatically. The device price but the size of the system will drive cost up as well. 

    As you can see we are looking for an ADC that allows us to alias sample up to the 6th or 8th Nyquist zone - which is unusual. 

    Any thoughts are appreciated.

    Best regards,

    Joe.

  • 0 in reply to Joseph Murray
    TI__Mastermind 39330 points

    ,Hi Joe,

    I apologize for the delay, have you looked at the ADC12QJ1600? This can sample down to 500MSPS and has plenty of BW, 6GHz, for your requirement.

    I can also be placed in 8bit mode, if you don't need the dynamic range.

    However, I am not sure you need this many ADC channels for your application. Please let me know if you think this will work, otherwise, we can brainstorm a little more.

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    TI__Mastermind 39330 points

    Hi Joe,

    There is an alternative approach, which I am not sure you are open to. But if you place a track and hold in front of the converter, this is a good way to extend the BW to your requirement and beyond. Unfortunately, TI doesn't offer stand alone track and holds but if interested we can take this offline and I can email you additional info.

    The downside in this approach, is more parts, power, etc.

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    Prodigy 40 points

    Rob -

    Thanks for the suggestion. I had not found that part - it is very interesting. I need to review the data sheet more carefully. Is there the ability to adjust the channel to channel clock phase; i.e., can we offset the sample clock between the 4 channels? The ADC083000 has Extended Clock Phase adjust fine/coarse that allows the clock to be shifted by between 0 and ~900ps. If we can adjust the clock phase between the 4 channels this is a nearly ideal device. We actually need a lot of channels so a quad is actually really nice - but only if we can control the sample clock between the various channels.

    Thanks, -Joe.

  • 0 in reply to Joseph Murray
    TI__Mastermind 39330 points

    Hi Joe,

    The channel to channel clock phase is already aligned within the part. No need to do any additional alignment for data capture, unless you are implying to interleave the four channels.....and that option isn't available to my knowledge.

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    Prodigy 40 points

    Rob,

    Thanks - that is a really nice part and would work very well but we do need to control the channel to channel clock phase offset for our application. Would be happy to discuss your thoughts on track-and-hold front end offline. I will mark this as resolved but look forward to hearing from you via email.

    Thanks,

    Joe.

  • 0 in reply to Joseph Murray
    TI__Mastermind 39330 points

    Sounds good Joe,

    I will send you an email now.

    Regards,

    Rob