Hi all!
For my first audio project I'm using the Tlv320aic1106 codec, connected to a FPGA chip. I’m supposed, for the first phase of the project to capture a microphone input and transmit the data on ethernet, receive it and put it on headphones.
I inspired my schematics from the ones presented in : http://www.ti.com/lit/ug/slau328a/slau328a.pdf on page 14. I removed all the jumpers and unnecessary additional connectors. The codec is used in uLaw conversion mode. Also, after few testes I removed the connection from MIC+ to R2 (the captured sound was all the same when the mic was connected on the differential ports of the chip, and removing that connection made the “reading” possible). Also, I’ve used R5(R6) = 470kohms and R3(R4)=39kohms.
In the first attempt I’ve used the same power supply( 3.3V ) for analog and digital power inputs. The result was a noisy one. So, after consulting with some colleagues, I’ve used different power supplies. In this point, I forgotten to connect the AGND to DGND, and this resulted in burning the codec(few ohms between DGND and 3.3VD ) and 2 ports on my FPGA, the ones corresponding to PMCO and PCMSYNC, so an input and an output port. I’ve replaced the codec and reconnect those signals to other ports and the whole thing was almost working. The noise was barely perceptible but the voice was saturated. I’ve changes the R5 and R6 resistors to 235kohms , and the result was almost good (I believe that I have some problems with the input impedance of the circuit, I used a 4,5Kohms microphone).
But after some time (the board was unplugged and plugged few times and was functioning ) I encountered the same problem as before. The codec IC was burned (32 ohms between PCMSYNC(18) and PCMO(17) ) and the same FPGA ports were burned.
Any ideas?
Thanks
