Questions about building a USB Type-C microphone

[Oblivion1407]

Hi everyone,
I'm new to this forum, and I feel that I suck at EE even after I have graduated from college with a decent GPA...
Anyways... I recently would like to learn how to use EDA tools to make circuits on PCB, and one of the project ideas was building a microphone and interfaces with pc using a USB module. The schematic is attached. I have several questions regarding to this design, which is just an assembly of reference designs...
First of all I learned that the electret condenser mic module requires some level of biasing voltage for the built-in JFET, but how do I determine the appropriate value for that? Is an arbitrary 5V good?
The second question is that after the AUDIO-IN, in INA217 reference design, there will either be a phantom power or switch to ground. AFAIK the ECM doesn't really need a phantom power to work, are these resistors pulling down to ground before the 47uF coupling capacitors still necessary?
The third question is how to determine the proper gain for the INA217? Do I need to run tests on the average and maximum output voltage from the ECM powering circuit and choose accordingly? What is the typical voltage range for ADCs? (The CYM-4032-6P module is a USB Type-C module with a Realtek ALC4032 chip on it, which I'd assume the chip should include ADC)
The last question is that would DC-DC converters be sufficient for powering these ICs? Can I just take the 5 V coming out of the USB Type-C port and use commercial DC-DC booster ICs to power the INA217 and Op-amps (and maybe the ECM powering circuit)? Are DC-DC booster ICs stable enough for this application? Do I have to use a separate battery instead?
Any help is strongly appreciated...

[Oblivion1407]

The sensitivity of the omnidirectional capsule I'm getting is about -42 dB, which is about 8 mV at 1 Pascal. 316 mV RMS meaning 447 mV on average so I assume the VREF that needs to be applied on the INA217 amplifier should also be that value? Then should I disconnect the feedback loop (The one with OPA237, 1MOhm and 0.1uF, suggested by TI) and put a voltage reference circuit in combination with the op-amp as voltage-follower? So the gain for the amplifier should be 20*log(447/8) = ~35dB?
Regarding to the chip itself, I actually cannot find the datasheet for the chip in the same number. According to the vendor, the chip does support USB interfacing, has ADC and DAC supporting either 16/24 bit with sample rate of up to 96 kHz(ADC) and 192 kHz(DAC), but the gain option is never mentioned. I would assume such chip would come with gain settings from 0 to 30 dB, but I doubt it would be as good as building a standalone preamp? Or maybe because they are much more pro and my effort could be just pointless lol. I think I will just build an amp with around 35 dB and just never use the gain setting on the driver then.
For the LDOs, I have always wondered "does providing higher voltage to the IC chips better?" I never dived deep into the internals of these power-related ICs so I'm pretty curious about it. Does giving a higher voltage makes it draw less current to do the same job? or does it make the performance better?
Also, thanks a lot for providing so much help and learning material that I couldn't find on Bing, Bing sux.

[Oblivion1407]

This is the new schematic, I wonder if this setup is okay...

[Oblivion1407]

I plugged in the USB audio module, and from the sound card driver settings I don't see a gain setting, so I would assume that the ADC does not have a built-in gain. So the preamp should have about 80 dB gain? I expect this to be used for speech rather than recording music. I find it very hard to design working with black-box modules, since the Realtek ALC4032 audio chip is quite mysterious and its datasheet isn't available anywhere AFAIK.
Since DC-DC boosters create mV level noise in general, is it a good idea to add LDOs at the output of these boosters?

[NiHaoMike]

For the analog voltages, you could use a charge pump running from the ADC sample clock so any feed through would become a small offset that can be nulled out.

Since audio ADCs are pretty much always going to have at least 2 channels and you're only making a mono microphone, you could have one set so that it would clip after the microphone itself would and the other set so that the microphone's noise floor would be just above the ADC's noise floor. If you're using a 24 bit ADC, there's a good chance it would have a wider dynamic range than the microphone itself and the gain would only need to be in the ballpark.

[Oblivion1407]

After printing out the PCB and assembling the PCB, The thing isn't working as intended. The power module by Mournsun E05155-1WR3 outputs only +-~0.5V under load, and outputs 17.3V with no load. So... maybe the circuit is taking too much power than the module can handle... Also the USB VBUS spikes to something like 20-ish volts when probed by multimeter, and then comes back down to 5.07V.
How to design around USB power? It seems the power budget is just too small.
Theory of operation: The circuit is sharing the same USB type-C port with the ALC4032 realtek audio codec module. VUSB from port will go through L78M05 and becomes VECM (used for biasing the electret condenser mic). The VUSB also go through the Mournsun module to produce a +-15V each capable of delivering 67mA or 1W, each output would go through a corresponding LDO such as L7815CV and KIA7915F to become more stabilized since the Mournsun module itself doesn't provide clean output. These power would supply the INA217 instrumentation amplifier and OPA137 op-amp. The mic generates signal to AUDIO_IN, which then goes through some ac coupling circuit and current biasing circuit to the INA217 input. The output feeds back into the reference terminal (which I am quite confused about but it's a recommended TI design so IDK). The AUDIO_OUT would then be fed into the ALC4032 audio codec and ready to be sent to the PC. But at the end, all I could see and hear was only periodic "metal-hits" sounds coming from the circuit (ding, pause, ding, pause, ding, and so on.), also nothing from the mic is recorded and added to these hitting sounds.
  I might have thought all of this as too straightforward and probably missed a lot of nuances in designing such circuit. Any suggestions of feedbacks are very appreciated.