Anonymous
10/30/2025, 2:26:37 AM
No.2954124
>>2954096
>An audio frequency PWM signal is differentiated by C7 and R3
unrelated but i think your differentiation drawing is missing half of the differentiation.
>The output is audio between 50 - 4000 Hz
no sir. the CLOCK may be 50 - 4000 Hz, but if you generate a 4 kHz sawtooth and filter out everything above 4 kHz, its just going to sound like a sine wave. the higher-order harmonics are precisely the difference between a sine wave, square wave, triangle, sawtooth, etc. you need to keep them. also if you generate any wave above about 10 kHz theyll all sound identical because the 2nd harmonic will get pushed out of human hearing range. give it a try https://onlinetonegenerator.com/
>But I understand that the square wave from the Pi will contain all these very high frequencies because the rise/fall time is measured in nanoseconds.
the sawtooth will contain the same harmonics. more, even. i really wouldnt work about it too much. if it helps you sleep better, know that a piece of metal works better as an antenna if it sees a high impedance load. pin 14, for instance, will never pick up interference, but pin 13 might. lowering R6 and R7 would help. and for gods sake do a fucking ground pour. you dont need a separate return path.
>And I know that interference is more of a problem with high frequencies.
thats a very strange thing to say for someone whos trying to build a synth circuit and lives in a world with 60 Hz power distribution and free automatic lowpass filters built into every human ear.
>I've heard a lot of stuff about separating signal ground and power ground
if youre below 1 amp just forget it.
>I also want to buffer and offset the square wave to send to the rest of the circuit
you might want to double check your wiring of U1B; you have feedback going to the non-inverting input. same for U1A and U1C... what the hell is this circuit even? did you build this? does it work?
>An audio frequency PWM signal is differentiated by C7 and R3
unrelated but i think your differentiation drawing is missing half of the differentiation.
>The output is audio between 50 - 4000 Hz
no sir. the CLOCK may be 50 - 4000 Hz, but if you generate a 4 kHz sawtooth and filter out everything above 4 kHz, its just going to sound like a sine wave. the higher-order harmonics are precisely the difference between a sine wave, square wave, triangle, sawtooth, etc. you need to keep them. also if you generate any wave above about 10 kHz theyll all sound identical because the 2nd harmonic will get pushed out of human hearing range. give it a try https://onlinetonegenerator.com/
>But I understand that the square wave from the Pi will contain all these very high frequencies because the rise/fall time is measured in nanoseconds.
the sawtooth will contain the same harmonics. more, even. i really wouldnt work about it too much. if it helps you sleep better, know that a piece of metal works better as an antenna if it sees a high impedance load. pin 14, for instance, will never pick up interference, but pin 13 might. lowering R6 and R7 would help. and for gods sake do a fucking ground pour. you dont need a separate return path.
>And I know that interference is more of a problem with high frequencies.
thats a very strange thing to say for someone whos trying to build a synth circuit and lives in a world with 60 Hz power distribution and free automatic lowpass filters built into every human ear.
>I've heard a lot of stuff about separating signal ground and power ground
if youre below 1 amp just forget it.
>I also want to buffer and offset the square wave to send to the rest of the circuit
you might want to double check your wiring of U1B; you have feedback going to the non-inverting input. same for U1A and U1C... what the hell is this circuit even? did you build this? does it work?