• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 63 subscribers
  • Views 778 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.

Switched mode power supplies (LMZ21701), LDOs (LP3982) and RF circuitry

user3938377
Prodigy 20 points
Other Parts Discussed in Thread: LP3982, LMZ21701

Dear Sir,

I have designed a quadruple GNSS (Global Navigation Satellite System) Radio-Frequency Front-End, i.e. four RF super-heterodyne independent RF channels in a single Front-End. Then I use a four channel ADC to digitize the four IF signals.

The main supply voltage is 5v, and the total consumption is high (including RF circuitry, ADC, and FPGA board).

So, first I use a high efficiency switched power supply (LMZ21701) to generate 3.7v from the 5v. Then I use LDOs (LP3982) to generate 3.3v from the 3.7v. In fact I use 3 LDO (at 3.3v s per each channel, one for the LNA (Low Noise Amplifier), a second one for the LO (Local Oscillator), and a third one for the VGA (Variable Gain Amplifier), which yields a total of 3x4 = 12 LDOs!!!. I also use ferrite beads. The noise in the power supply is low. However I believe my design might be too redundant. My questions are:

Are LDOs a good solution to filter out the noise and glitches coming from the switched power supply? I guess that not really, as far as I know they behave as a High Pass Filter. They will filter out the low frequency oscillations, but the noise due to the switching inside the switched power supply is above their cut-off frequency. In fact ferrite beads should be more efficient for this purpose. Do you agree?

It is wise to supply RF circuitry (LNA, PLLs, VGA) from a switched power supply (LMZ21701) carefully filtered by ferrite beads and capacitors?

Using one independent LDO for each RF blocks will be helpful to reduce crosstalk among RF channels? Again I guess it will not, for the same reason aforementioned. Do you agree?

What about supplying the ADC with a second (LMZ21701) to generate 1.8v from the incoming 5v (again, carefully filtered by ferrite beads and capacitors) ?

BR

Orlando Peña 

  • 0
    TI__Mastermind 22610 points

    Since this is in 'Linear Regulators Forum' I will respond to the LP3982 questions.

    Q1: "... In fact ferrite beads should be more efficient for this purpose. Do you agree? ..."

    A1: For high frequency noise above the LDO cut-off frequency, generally Yes. The PSRR typical characteristic curve will show the typical roll-off at high frequency, where the output capacitor behavior dominates above the LDO cut-off frequency

    Q2: " ... It is wise to supply RF circuitry (LNA, PLLs, VGA) from a switched power supply ... carefully filtered by ferrite beads and capacitors?"

    A2: The operative words are '... carefully filtered ...'. The best I can say here is Perhaps.

    Q3: "... Using one independent LDO for each RF blocks will be helpful to reduce crosstalk among RF channels? ... Do you agree? ..."

    A3: One LDO per RF block should reduce crosstalk. However, the effectiveness will depend on capacitor selection and pc board layout. While there is a specification for noise/ripple at the input getting through the LDO to the output (PSRR), there is no equivalent specification for the other direction, so this is difficult to quantify.

    Q4: "... What about supplying the ADC with a second (LMZ21701) to generate 1.8v from the incoming 5v (again, carefully filtered by ferrite beads and capacitors) ? ..."

    A4: Same as answer #2