• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 266 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.

ADS6444: Driving an ADC in the second Nyquist zone

Volodymyr Bendak
Prodigy 15 points
Part Number: ADS6444

Tool/software:

Hello

 

I am designing a digital receiver according to the topology shown in the figure. After the mixer there is a 5th order differential bandpass filter, the output of which is connected to the operational amplifier through 51 Ohm resistors that serve as a load for the filter. This filter has an input and output differential impedance of 100 Ohm and provides operation in the 2nd Nyquist band. The ADC clock frequency is 100 MHz.

After analyzing many application notes, I realized that between the ADC and the operational amplifier it is necessary to put an RC filter and resistors that shift the signal by DC. But the RC filter was considered for cases of operation in the 1st Nyquist band. For the 2nd Nyquist band, the capacitance value will be very small if we also take into account that the RC cutoff frequency must be 6.5 times greater than the upper operating frequency to ensure equal frequency response. I think that for cases where 2, 3, 4… Nyquist bands are used, slightly different approaches are used to solve this problem.

Since the ADC is 4-channel, it is not possible to place the operational amplifiers directly near the inputs, in my case the maximum length is approximately 37 mm.

So my questions are as follows

1) What should be the scheme between the operational amplifier and the ADC and how to calculate the values ​​of the components to achieve optimal parameters? What goal should be achieved to obtain optimal parameters?

2) How to correctly place the components in the case when the operational amplifier and ADC are separated on the board and what should be the line impedance?

3) For which cases is the value of the input resistance of the ADC used? It is given in the datasheet ADS6444 Figure 84

4) Is the ideology of my general scheme shown in the figure correct?

  • 0
    TI__Mastermind 38970 points

    Hi Volodymyr,

    See my comments to your questions below:

    1) What should be the scheme between the operational amplifier and the ADC and how to calculate the values ​​of the components to achieve optimal parameters? What goal should be achieved to obtain optimal parameters? RR: right now you need to create/design an anti-aliasing filter or AAF, this would go inbetween the caps and the ADC. For the other components, please review this app note to help. I would also put some small series resistors close to the outputs of the amp, 20ohms is fine to start.

    www.ti.com/.../slyt853.pdf

    2) How to correctly place the components in the case when the operational amplifier and ADC are separated on the board and what should be the line impedance? RR: this would be difficult to describe, as the impedance and interface gets tricky. I would advise to keep both components, the amp and ADC on the same board.

    3) For which cases is the value of the input resistance of the ADC used? It is given in the datasheet ADS6444 Figure 84 RR: use this impedance for interfacing with the AAF at the highest BW you plan to use in your application.

    4) Is the ideology of my general scheme shown in the figure correct? RR: yes, but you need termination resistors on both sides of the AAF and near the ADC inputs and near the AMP outputs. See the app note above which helps to describe this.

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    Prodigy 15 points

    Thanks for the quick reply

    I have read your app note. The document considers the case when an anti-aliasing filter is placed between the FDA and the ADC.
    In my diagram given in the previous post, the anti-aliasing filter is placed before the FDA just after the mixer.
    The filter after the mixer is necessary because it generates a lot of unnecessary spectral components that are not desirable. The filter also sets the noise band. Feeding the signal from the mixer output to the amplifier input without a filter will cause intermodulation distortion.

    1) What parameters (for example : SNR, SFDR, total dynamics of the receiver, etc.) will the placement of the AAF affect in the cases before and after the FDA? In which case will the system have better performance and why? Do I need to put an additional filter between FDA and ADC?

    2) I also understood that the input impedance of the ADC always has the input impedance specified in the datasheet, except for the moments when the sampling capacitor is connected during the acquisition time. Is this reasoning correct?

    3) If the previous reasoning is correct, then there must be sufficient capacitance at the ADC input to ensure a signal level drop of less than 0.5 LSB during the acquisition time. How is this achieved in a band-pass antialiasing filter topology?

    4) What is the resistor RTAMP for?

    5) "at the highest BW" Do you mean highest frequency of BW? For example, the Nyquist bandwidth is 50 MHz and I work in the second zone, then the frequency value will be 100 MHz?

    Best regards
    Volodymyr

  • 0 in reply to Volodymyr Bendak
    TI__Mastermind 38970 points

    Hi Volodymyr,

    Please see my comments below in RED.

    I have read your app note. The document considers the case when an anti-aliasing filter is placed between the FDA and the ADC.
    In my diagram given in the previous post, the anti-aliasing filter is placed before the FDA just after the mixer.
    The filter after the mixer is necessary because it generates a lot of unnecessary spectral components that are not desirable. The filter also sets the noise band. Feeding the signal from the mixer output to the amplifier input without a filter will cause intermodulation distortion. RR: you will still need an AAF inbetween the output of the FDA and input of the ADC in order to limit the noise of the FDA. If you don't, your ADC will degrade in SNR/SFDR performance.

    1) What parameters (for example : SNR, SFDR, total dynamics of the receiver, etc.) will the placement of the AAF affect in the cases before and after the FDA? In which case will the system have better performance and why? Do I need to put an additional filter between FDA and ADC? RR: yes, see above, you can also prove this by doing just a noise analysis of the FDA and ADC with and without an AAF. 

    2) I also understood that the input impedance of the ADC always has the input impedance specified in the datasheet, except for the moments when the sampling capacitor is connected during the acquisition time. Is this reasoning correct? RR: its more simpler to just use the rated input impedance, 100ohm, etc in the datasheet specifications.

    3) If the previous reasoning is correct, then there must be sufficient capacitance at the ADC input to ensure a signal level drop of less than 0.5 LSB during the acquisition time. How is this achieved in a band-pass antialiasing filter topology? RR: not sure I totally understand your question here, but the ADC has sufficient BW in order to settle to 0.5LSB from sample to sample, hanging a BPF on the input of the ADC's does not change this metric.

    4) What is the resistor RTAMP for? RR: sometimes you need to add in a separate resistor termination near the outputs of the FDA in order to get the impedance closer to what the amplifier wants to see for a load. Keep in mind the amplifier was characterized per the datasheet under specific load conditions. So you want to make the aggerate impedance that the amp "sees" near this load impedance.

    5) "at the highest BW" Do you mean highest frequency of BW? For example, the Nyquist bandwidth is 50 MHz and I work in the second zone, then the frequency value will be 100 MHz? RR: please give me more of a reference to this statement, what page, etc are you sighting?

    Regards,

    Rob