[alsa-devel] Audio routing in the tlv320aic3104

[Rick Mann]

> On Oct 7, 2015, at 05:52, Caleb Crome <caleb at crome.org> wrote:
> 
> 
> 
> Sent from my iPhone
> 
>> On Oct 7, 2015, at 4:12 AM, Rick Mann <rmann at latencyzero.com> wrote:
>> 
>> I've managed to get a hacky, fragile config working, in the sense that it sends my tlv320aic3104 samples from speaker-test, and the clocks and data all look good on the oscilloscope, and it ping-pongs between the left and right channel.
>> 
>> But it's outputting the signal on the HPLOUT/HPROUT pins, not the L&R Line Out pins.
>> 
>> My DTS has this:
>> 
>>   fragment at 3 {
>>       target = <&ocp>;
>>       __overlay__ {
>>           sound {
>>               compatible = "ti,da830-evm-audio";
>>               ti,model = "DA830 EVM";
>>               ti,audio-codec = <&tlv320aic3x>;
>>               ti,mcasp-controller = <&mcasp0>;
>>               ti,codec-clock-rate = <12000000>;
>>               ti,audio-routing =
>>                   "Line Out",             "LLOUT",
>>                   "Line Out",             "RLOUT";
>>           };
>>       };
>>   };
>> 
>> The original had:
>> 
>>   "Headphone Jack",       "HPLOUT",
>>   "Headphone Jack",       "HPROUT",
>>   "LINE1L",               "Line In",
>>   "LINE1R",               "Line In";
>> 
>> I also have an asound.state file I got from the Audio Cape:
>> 
>>   https://raw.githubusercontent.com/CircuitCo/BeagleBone-Audio/files/asound.state
>> 
>> I don't really know how to interpret that file. One resource online said it held the current state of the system. I guess that suggests I should play with amixer. But how does the DTS routing come into play?
>> 
>> Thanks,
> 
> Check for the routing in the davinci-evm.c file.  It has other routes,  I kind of recall it says some outputs are not connected.    You might have to edit routs there too. 

Actually I was able to use amixer to switch the DAC muxes to the right setting, and the signal routed to the right place. I think it's the first time in three weeks that the first thing I tried actually worked as expected.

But I had hoped ALSA would enable only that route given the DTS; perhaps it's the asound.state file I downloaded that forced its hand the wrong way; I'll try that experiment later.

At least now I can populate the remaining parts on the board to see if they work, too.

I still am far from complete with the Linux config (need to get it working in 4.x, need to get my enable GPIOs working automatically, etc.), but I can finally make real progress.

Thanks!