Need Ultra Low Noise Power Solution for the Requirement

shyam.sunder91 said:

the output then taken to a LDO again the LDO output is maintained with a CLC filter

Hi,

it is not a good idea to use an CLC-Filter at the output of a LDO.

I suppose that your noise on the power rails comes from ground noise, that is generated from your LT8471 DC/DC.
Try to locate your noise source. But be aware that with a scope you can get fooled and see ghost signals, because of ground bounce.
(You can check that by connecting the crocodile clip to circuit ground and the probe tip directly to the crocodile clip.
Any signal you still see, ist from ground HF-noise)
A superior layout is absolute mandatory. for the layout, look at the demo board.
Additionally:
I would suggest that you insert ferrite beads + Cs: Converter ground to board ground and pos. and neg. rail.
Pay attention to the frequency of your noise. Choose a ferrite bead that has high impedance in that frequency region.
Be aware, that inductors have a self resonance frequency. When noise frequency comes close to that frequency it doesn't filter anymore.
Also the CLC-filter has its own resonance frequency, that should be damped ore avoided.

Better choose a ultra low noise, high PSSR LDO like LT3042 (positive) and LT3093 (negative). For your load current you would have to parallel two of it.
Or you look for stronger LDOs.
These regulators have high PSSR even at high frequencies. That eases the filter cutoff-frequency in front of it.
And they have so low noise, that an output filter will not improve that significantly .
Probably you'll better choose a C - ferrite bead - C filter in front of LDOs. Because they are better for filtering up to high frequencies.
You must choose a part with sufficient current handling.

shyam.sunder91 said:

the output then taken to a LDO again the LDO output is maintained with a CLC filter

Hi,

it is not a good idea to use an CLC-Filter at the output of a LDO.

I suppose that your noise on the power rails comes from ground noise, that is generated from your LT8471 DC/DC.
Try to locate your noise source. But be aware that with a scope you can get fooled and see ghost signals, because of ground bounce.
(You can check that by connecting the crocodile clip to circuit ground and the probe tip directly to the crocodile clip.
Any signal you still see, ist from ground HF-noise)
A superior layout is absolute mandatory. for the layout, look at the demo board.
Additionally:
I would suggest that you insert ferrite beads + Cs: Converter ground to board ground and pos. and neg. rail.
Pay attention to the frequency of your noise. Choose a ferrite bead that has high impedance in that frequency region.
Be aware, that inductors have a self resonance frequency. When noise frequency comes close to that frequency it doesn't filter anymore.
Also the CLC-filter has its own resonance frequency, that should be damped ore avoided.

Better choose a ultra low noise, high PSSR LDO like LT3042 (positive) and LT3093 (negative). For your load current you would have to parallel two of it.
Or you look for stronger LDOs.
These regulators have high PSSR even at high frequencies. That eases the filter cutoff-frequency in front of it.
And they have so low noise, that an output filter will not improve that significantly .
Probably you'll better choose a C - ferrite bead - C filter in front of LDOs. Because they are better for filtering up to high frequencies.
You must choose a part with sufficient current handling.

LDOs are ok, but what about a 12V to lower voltage conversion bucks / DCDC ?

i currently use a C 10uH C filter bank, i need to try measuring noise as you said.