Very cool! I'm still waiting for the boards from fab but I've written the mcu code and basically did that so I'm glad that it works. How's the performance? Is it quiet?PhilHaw wrote: ↑Wed Nov 03, 2021 5:53 pm We've also got our AK4619 working. As sebotron says, most of the default values in the control registers will work ok. We just needed to set up the analog input mode to single-ended and you also need to set the Power Management register to power up the ADCs and DACs as they default to off.
Codec alternatives?
Re: Codec alternatives?
Re: Codec alternatives?
We're still at the breadboard stage (PCB design almost finalised) so I wouldn't like to comment on any noise due to the spaghetti wiring which is probably picking up some clock noise and other garbage from my PC! Mick has told me his setup is as quiet as the FX Core Dev Board with the AK4556VT codec.
Philip Hawthorne
Blue Nebula Development Team
Blue Nebula Development Team
Re: Codec alternatives?
I didn't dare try to breadboard high speed clock signals - I'm surprised you guys got it working! I'm running a final check on my layout and schematics and will order proto boards today. I'm deathly afraid of having forgotten somethingPhilHaw wrote: ↑Wed Nov 10, 2021 5:54 amWe're still at the breadboard stage (PCB design almost finalised) so I wouldn't like to comment on any noise due to the spaghetti wiring which is probably picking up some clock noise and other garbage from my PC! Mick has told me his setup is as quiet as the FX Core Dev Board with the AK4556VT codec.
Re: Codec alternatives?
I know the feeling!
Adding series resistors in the clock lines (33ohm or 51ohm) is recommended to help reduce clock problems - see the MCLK connection on the FXCore Dev Board schematic for example.
Good luck with your boards
Phil.
Philip Hawthorne
Blue Nebula Development Team
Blue Nebula Development Team
Re: Codec alternatives?
This^^^
Put the resistor on the driving side of the line (i.e. close to FXCore for MCLK) and not the receiving side. This will cut down on reflections, ringing, false clocking, etc. Also helps FCC/CE compliance testing.
Most times values in the 30 to 50 ohm range work fine but you may have a board with complex routing and have an issue getting the right value. In that case remove the resistor, put a POT across the pads and adjust it till you get rid of the reflections then remove the POT and measure it to find the value.
I also put them on the SDO lines, I should have them on SDI as well but missed them and since they didn't ring I did not notice.
Re: Codec alternatives?
Thanks for that advice Frank.
Phil.
Phil.
Philip Hawthorne
Blue Nebula Development Team
Blue Nebula Development Team
Re: Codec alternatives?
Frank wrote: ↑Wed Nov 17, 2021 8:12 amThis^^^
Put the resistor on the driving side of the line (i.e. close to FXCore for MCLK) and not the receiving side. This will cut down on reflections, ringing, false clocking, etc. Also helps FCC/CE compliance testing.
Most times values in the 30 to 50 ohm range work fine but you may have a board with complex routing and have an issue getting the right value. In that case remove the resistor, put a POT across the pads and adjust it till you get rid of the reflections then remove the POT and measure it to find the value.
I also put them on the SDO lines, I should have them on SDI as well but missed them and since they didn't ring I did not notice.
Thank you guys! I haven't gotten my proto boards back yet but I'm definitely adding that to my design. I used the devboard schematics as a reference but I totally missed that 33 ohm resistor.
Re: Codec alternatives?
Hi guys!
Quick update - I did get my PCBs last week but then I'm having a bunch of issues I need to take care of before it works. Side note, hand soldering qfn is horrible. That is all
Quick update - I did get my PCBs last week but then I'm having a bunch of issues I need to take care of before it works. Side note, hand soldering qfn is horrible. That is all
Re: Codec alternatives?
I feel your pain buddy! I can barely see the QFN package never mind solder it by hand
Phil.
Philip Hawthorne
Blue Nebula Development Team
Blue Nebula Development Team