G'day y'all.
Venting my frustration a bit and looking for a bit of light as well.
I have a design and boards where I get audio into a PCM1804, to a µController, and out to a PCM1793. The µC - 1793 combination works a treat. I used it with an Amanero combo board as well, and it's nice and transparent. No fancy parts on the design, either OPA2134 or NE5532 (don't mind either) at the output stage and it's all good.
The ADC brother however is a different story altogether. When I listen to it, there's always some white noise in the background. I wrote a small Python script in which I can show the digital data coming into the µController. So, debugger connected, read 8192 samples of audio from memory into the PC and show them. the noise floor is barely around -100dB, whereas the 1793 does easy -130. I also measured the inputs just by the 1804 and they're a good -120 or so dB noise floor after two input op amps.
Clock is generated by the µC, 24.576 MHz to generate 96KHz sampling rate. The signal on the scope looks pretty stable.
I've attached the schematic, and the memory content with a 1VRMS signal input to the 1804 and the µc. I controlled all settings (OSR, FMT, M/S) a couple of times, and can't seem to find a fault (OSR is set to dual rate 96KHz slave and FMT to I2S).
If any of you have any suggestions, happy to entertain them.
Cheers,
Albert
Venting my frustration a bit and looking for a bit of light as well.
I have a design and boards where I get audio into a PCM1804, to a µController, and out to a PCM1793. The µC - 1793 combination works a treat. I used it with an Amanero combo board as well, and it's nice and transparent. No fancy parts on the design, either OPA2134 or NE5532 (don't mind either) at the output stage and it's all good.
The ADC brother however is a different story altogether. When I listen to it, there's always some white noise in the background. I wrote a small Python script in which I can show the digital data coming into the µController. So, debugger connected, read 8192 samples of audio from memory into the PC and show them. the noise floor is barely around -100dB, whereas the 1793 does easy -130. I also measured the inputs just by the 1804 and they're a good -120 or so dB noise floor after two input op amps.
Clock is generated by the µC, 24.576 MHz to generate 96KHz sampling rate. The signal on the scope looks pretty stable.
I've attached the schematic, and the memory content with a 1VRMS signal input to the 1804 and the µc. I controlled all settings (OSR, FMT, M/S) a couple of times, and can't seem to find a fault (OSR is set to dual rate 96KHz slave and FMT to I2S).
If any of you have any suggestions, happy to entertain them.
Cheers,
Albert
Attachments
Alright, very little update.
Sent all details to the TI forum, see if they can come up with answers or any help. There's more pictures of the signal on the link just in case.
Cheers,
Albert
https://e2e.ti.com/support/data_converters/audio_converters/f/64/t/559002
Sent all details to the TI forum, see if they can come up with answers or any help. There's more pictures of the signal on the link just in case.
Cheers,
Albert
https://e2e.ti.com/support/data_converters/audio_converters/f/64/t/559002
- Status
- Not open for further replies.