Computer USB Sound Card / Guitar-to-PC Audio Interface

Thread Starter

Westted

Joined Jul 26, 2026
7
I just need to know if the circuit itself is going to work, because I'm currently not able to simulate it and I need confidence to continue with the PCB design. It is a simple audio interface for a guitar, it consists of a preamp (Hi-Z), the codec (PCM2900C), and an amplified output with a headphone driver connected directly to the output jack. Oh, and everything is powered by the PC's USB port, except for the codec's analog section, for that, I used an adj regulator to ensure greater stability. I ask that you ignore the separation of the GNDs; it's mostly for clarity. On the PCB, everything will share the same ground, and I'll figure out how to deal with the noise (I can't use multiple layers). I'm a beginner at design, so I would love to hear y'all! thanks in advance.
 

Attachments

cmartinez

Joined Jan 17, 2007
8,828
Ground separation of a mixed digital/audio circuit is an important issue. Of course, all grounds have to meet at a point somewhere in the PCB, but where that point is located makes the difference between a clean and a noisy audio output. As for the circuit itself, I suggest you use LTSpice (it's free) to at least simulate the the parts of your circuit that can more easily be simulated, That is, everything but the PCM2900 codec, for which I honestly have no idea where to find the spice model.
 

MisterBill2

Joined Jan 23, 2018
27,952
I see no flaws in the circuit design, but I have not verified the pin connection numbers.
As for the "ground noise" consideration, the goal is to not have any audio signal currents flowing thru the same conductors as the noise currents. THAT is seldom simple.
 
Last edited:

eetech00

Joined Jun 8, 2013
4,723
I just need to know if the circuit itself is going to work, because I'm currently not able to simulate it and I need confidence to continue with the PCB design. It is a simple audio interface for a guitar, it consists of a preamp (Hi-Z), the codec (PCM2900C), and an amplified output with a headphone driver connected directly to the output jack. Oh, and everything is powered by the PC's USB port, except for the codec's analog section, for that, I used an adj regulator to ensure greater stability. I ask that you ignore the separation of the GNDs; it's mostly for clarity. On the PCB, everything will share the same ground, and I'll figure out how to deal with the noise (I can't use multiple layers). I'm a beginner at design, so I would love to hear y'all! thanks in advance.
I hope you know you can buy an excellent one for $79 USD
 

MisterBill2

Joined Jan 23, 2018
27,952
BUT here is a question: Just what is the circuit supposed to do ?? I am not familiar with the codec (PCM2900C) as to what function it provides. I see a USB connection, but the only control is a gain adjustment in the input circuit.
I SEE ANALOG IN AND two channels of ANALOG OUT.
So it would be a favor to al of us who are not familiar with the codec (PCM2900C) to explain what the purpose or goal actually is.
 

dashstofsk

Joined Jun 16, 2013
7
You seem to be trying to connect your circuit to the audio input of your computer?

I don't like your 470uF capacitors connected to the audio jack 3. When you switch your circuit off the charge on these capacitors will have to go somewhere and because the soundcard is a low impedance device you might get a forward current rush ruining your soundcard. Or even a back rush and ruin your TDA1308. Or am I wrong?

Also the audio interface on many new computers is not TRS but TRRS. Are you definitely sure it is TRS? If TRRS you will need to add a suitable resistor between GND and MIC so that your computer will recognise the external input.
 

Thread Starter

Westted

Joined Jul 26, 2026
7
Ground separation of a mixed digital/audio circuit is an important issue. Of course, all grounds have to meet at a point somewhere in the PCB, but where that point is located makes the difference between a clean and a noisy audio output. As for the circuit itself, I suggest you use LTSpice (it's free) to at least simulate the the parts of your circuit that can more easily be simulated, That is, everything but the PCM2900 codec, for which I honestly have no idea where to find the spice model.
Glad you said that, because yesterday i've tested the preamp part and everything went perfect, and today i'll be testing the amplified output for headphones. Oh and in case ur curious yes i've been using LTspice.
 

Thread Starter

Westted

Joined Jul 26, 2026
7
BUT here is a question: Just what is the circuit supposed to do ?? I am not familiar with the codec (PCM2900C) as to what function it provides. I see a USB connection, but the only control is a gain adjustment in the input circuit.
I SEE ANALOG IN AND two channels of ANALOG OUT.
So it would be a favor to al of us who are not familiar with the codec (PCM2900C) to explain what the purpose or goal actually is.
Yeah, you're right, I should have mentioned that. Basically, it's a pretty simple audio card or sound interface between an electric guitar and a PC. The PCM2900C is the codec; it has a preamp, an amplified output with a headphone driver, and that's it. There are plenty of filters to reduce port noise and a voltage regulator to provide a more stable power supply to the codec's analog section.
 

Thread Starter

Westted

Joined Jul 26, 2026
7
You seem to be trying to connect your circuit to the audio input of your computer?

I don't like your 470uF capacitors connected to the audio jack 3. When you switch your circuit off the charge on these capacitors will have to go somewhere and because the soundcard is a low impedance device you might get a forward current rush ruining your soundcard. Or even a back rush and ruin your TDA1308. Or am I wrong?

Also the audio interface on many new computers is not TRS but TRRS. Are you definitely sure it is TRS? If TRRS you will need to add a suitable resistor between GND and MIC so that your computer will recognise the external input.
Thanks for your comments, but I need to clarify a few things. The capacitor needs to have that specific capacitance to block the DC component coming out of the amplifier (2.5 V). If I used a capacitor with, say, 10 μF, it would cut off all the bass (high-pass filter), which is how I prevent the headphone coils from burning out (they’re usually low impedance, 32 ohms). Also, the 10k resistors slowly discharge the capacitors, and the 47-ohm resistors limit the maximum current (Imax) to 53 mA (this doesn’t damage the op-amp; it even has short-circuit protection). Regarding the last point, the output jack doesn’t go to the PC; it’s the headphone output. The PC connects to the circuit via USB and recognizes it as an audio card.
 

Thread Starter

Westted

Joined Jul 26, 2026
7
Update: I simulated the headphone amplifier at the output, and everything worked as expected. I replaced the coupling capacitors on pins 15 and 16 of the PCM2900C with multilayer ceramic capacitors, and the 47-ohm resistors at the output with 10-ohm resistors to get a wider volume range on the headphones (plus, the TDA1308 already has short-circuit protection, so the 47-ohm resistors aren’t really necessary). I changed the values of R10 and R11 because I noticed that, together with the capacitors, they formed a filter that didn’t quite cut off the codec’s ultrasonic frequencies—and now it does.
 

Attachments

Thread Starter

Westted

Joined Jul 26, 2026
7
Update:
I used op-amp B of the LMC6482 as a voltage follower to act as a buffer for the virtual ground of the TDA1308, thereby achieving greater stability since I now have about 2.5 V at low impedance.
 

Attachments

Top