Adding 2 channel audio to a Gumstix

Walter,

Are you talking output-only, input-only, or input/output?  Your description makes it sound line output-only, but it's not clear what your need for 'split' entails.

How "independent" are the two audio channels?  Different sample rates / different number of bits / same sample rate but not synchronized? 

Could you use the tps65950 for one of the channels, and a pcm2902 for the other (assuming they really are independent)?

Regards,

Frank 

Sorry, let me try and provide a clearer picture.

It would be output only; we want to use the gumstix to drive two speakers.  One would be located on one side of a kiosk and the other on the opposite side.  So the gumstix will monitor lets say which keys are pressed on each side of the kiosk.  So if someone on side A for example presses key 1, a wav file would play with a voice saying "you pressed key 1".  On the other side someone could press key 2 and hear a different message.  So by independent I mean the audio is detached to where someone on each side can press a different button and they could hear a totally different message at the same, but still only using one gumstix.  I did find the cm108 chip set (we have USB), but since the gumstix is mainly TI parts, I would like to keep it that way.  The rates, bits all the same, but yes not synchronized.

I think I am on the path that you are describing.  Using the built in 65950 as one audio source and adding the pcm2902 as a second, since we have been unsuccessful in splitting the R and L with the 56950 and are new to working with audio at this level.

Thanks for your feed back Frank.

Walter

Walter,

If you play a stereo file to the 56950, do you get good separation between the Left and Right outputs?  If so, then it doesn't seem like the "unsuccessful in splitting the R and L" is really "with the 56950".  Do you need a software application that can interleave two mono channels real-time into a stereo pair to be sent to the 56950 stereo DAC?  Is the lack of such an application why you are looking for a hardware work-around?

Regards,

Frank

I would have to say yes, the lack in being able to use software to accomplish what we are trying to do with the current setup is why we are looking for a hardware work around.  It seems somewhat difficult to work with the audio in this embedded version of linux we are using.    Also, another factor is the time to get this done.  only have around a couple of months to get this done.  if the developer can easily operate to different sound devices then that would get it moving, instead of trying to work out a solution with the current setup.  If there is a software solution out there, we have just not found it yet.

Thanks for your feedback,

Walter

I was looking through the 65950 spec again and reading the inital section on the audio says

"The audio codec in the device includes five DACs and two ADCs to provide multiple voice channels and stereo downlink channels that can support all standard audio sample rates through I2S/TDM format interfaces."

I am assuming that the 5 DAC's are used in different modes since you only have the HSOL and HSOR coming out of the chip.

Walter,

I suggest that you ask your question regarding the five DACs via a 'New Post'. 

I'm not knowledgeable on this part, and I'm afraid that others won't look far enough into this trail to see it.

Regards,

Frank