I have posted complete schematics for moderate_power audio amplifiers, here on stackX. Try searching. "amplifier"
And in the www.diyAudio.com site, search for "simplistic njfet RIAA" for the discussion thread, moderated by "salas", on excellent vinyl preamp for moving_coil phonograph circuits.
=======================================
About 3 years ago, a diyAudio poster named "Riccardo"? developed a crucial PCB design with two large areas and a rather thin region between them that served to GROUND the RIAA time_constant capacitors. That long thin region is important in preventing ECHO; since the RIAA has a 50Hz pole (3 millisecond time constant), crosstalk at the 1/100,000 (10 PPM, -100dB) may be audible. You want to have the Preamp output current NOT CIRCULATE around the edge of the PCB and slightly modulate the gate_source voltage of the input JFET.
See this answer for some thoughts on this crosstalk.
Why are the advantages of JFET over MOSFET, or why are JFET still used?
===========================================
Here is thinking on WHY have the Preamp PCB far away from the Power Amplifier PCB/power_supply.
What's the point in a preamp?
This style of thinking ---- minimize currents in the Ground Plane for precision (echo-less) circuits ---- is important for instrumentation and telemetry and any high_bit_count measurements even if your Signal Chain is only a 22bit or 24 bit or 18 bit ADC.
Often these ADCs have a VREF (voltage reference) pin, along with expectation for a CLEAN QUIET externally-generated voltage.
However, if you have TRASH or SPIKES or simply power_supply_currents that flow in your GROUND PLANE, that VREF will not be constant.
That means your 18/20/22/24 bit ADC will have LOTS OF MEASUREMENT ERRORS.
Thus, for success in the audio world, and success in the precision measurement world, you need to think about GROUND CURRENTS.
A square of standard thickness PCB copper foil has resistance of 0.000500 ohms (that is 1/2,000 ohm or 500 microOhms). A ground current of 0.1 amps, flowing through that square of foil, using Ohms Law V = I * R, produces 0.1 amp * 500 microOhms === 50 microVolts error.
In a 2.5volt FullScale system, that is 20 ppm (parts per million) error. Thus your 24_bit ADC now is only good for 16 bits.
Depending on how that 0.1 amp flows in the GROUND PLANE, this 20ppm might affect the Full Scale or the ZERO POINT, or both.
Learn to sketch the current flows. A grid of resistors in SPICE, maybe of size 10 by 20, or 200 tiny squares, is a good initial thinking tool.