AD9276 CW operation question

Hi there,

We are evaluating AD9276 as medical ultrasound AFE. one question related to CW demodulation architecture on page27 of datasheet need to consult experts of ADI.

On page 27, it mentions two architecture of implementing CW sum and demodulation.

A. traditional method : V-I amp  --- passive delay line ---- summing (current) ---and  demod.----ADC

B. Better one as in AD9276: Demod each channel individually ---- then phase shift (delay) ----summing ----ADC

The difference is the sequence of demodulation and summing.

For A, “The dynamic range of the demodulator can limit the achievable dynamic range.”(quote datasheet page27),  is the disadvantage.

For B ,“Because the dynamic range expansion  from beamforming (i.e. the summing technique to enhance SNR ) occurs after demodulation, the demodulator dynamic range has little effect on the output dynamic range”(quote datasheet page27),  is better according to datasheet.

I may not understand this very well.

Firstly, a good demodulator (like AD8339) has very high Dynamic Range, seems not likely to limit the achievable dynamic range?

Second, would you explain with a calculation example on how could modulator affect the dynamic range if placed after summing? I do found the concept that "summing all N channels signal will improve SNR by 10*log(N) times", but I do not know how to take modulator parameter into calculation.

Any advice is appreciated, thanks!



add details
[edited by: XYanalog at 1:02 AM (GMT 0) on 8 Jul 2020]
Parents
  • 0
    •  Analog Employees 
    on Jul 7, 2020 4:20 PM 9 months ago in reply to XYanalog

    Hi XYanalog,

    I'm sorry for the delayed reply. I'll try to find the right person to help with your question.

    Doug

  • +1
    •  Analog Employees 
    on Jul 8, 2020 2:33 AM 9 months ago in reply to XYanalog

    Hi XYanalog,

    The traditional CW method : LNA --- V-I amp  --- passive delay line (Inductor) ---- summing (current) ---and  demod.---- precision ADC

    The delay is realized by using an inductor as the passive delay line, the inductor has multi taps output, each tap has different delay time. Typically, the designer will use a combination of these delay time to generate the different phases, example 16 to 64 phases. This is realized by summing all the current output from different taps together in the circuits. The disadvantage of the delay line (inductor) is the accuracy is not so good, typically bigger than 10%.  A typical delay line circuit is shown as below picture.

    After summing, usually a frequency mixer, like TUF-3HSM, is used to generate I and Q signal followed with filtering circuits to get the phase shifter signal. such kind of demodulator has limited dynamic range. 

    if you look into the details of architecture and application information in the datasheets, AD8339 and AD8333 are more like the CW demodulator which has been integrated in AFE, such as AD9276, AD967x. They are used for the most recent method as integrated AFE, which is LNA ---- Demod each channel individually (AD8333/9) ---- then phase shift (delay) ----summing ---- precision ADC.

    You can contact me directly to discuss more details if you want, I am covering the medical ultrasound application for Asia. 

    My email is hugh.yu@analog.com

Reply
  • +1
    •  Analog Employees 
    on Jul 8, 2020 2:33 AM 9 months ago in reply to XYanalog

    Hi XYanalog,

    The traditional CW method : LNA --- V-I amp  --- passive delay line (Inductor) ---- summing (current) ---and  demod.---- precision ADC

    The delay is realized by using an inductor as the passive delay line, the inductor has multi taps output, each tap has different delay time. Typically, the designer will use a combination of these delay time to generate the different phases, example 16 to 64 phases. This is realized by summing all the current output from different taps together in the circuits. The disadvantage of the delay line (inductor) is the accuracy is not so good, typically bigger than 10%.  A typical delay line circuit is shown as below picture.

    After summing, usually a frequency mixer, like TUF-3HSM, is used to generate I and Q signal followed with filtering circuits to get the phase shifter signal. such kind of demodulator has limited dynamic range. 

    if you look into the details of architecture and application information in the datasheets, AD8339 and AD8333 are more like the CW demodulator which has been integrated in AFE, such as AD9276, AD967x. They are used for the most recent method as integrated AFE, which is LNA ---- Demod each channel individually (AD8333/9) ---- then phase shift (delay) ----summing ---- precision ADC.

    You can contact me directly to discuss more details if you want, I am covering the medical ultrasound application for Asia. 

    My email is hugh.yu@analog.com

Children
No Data