Hello -
The ADC on the codec on the C6713 DSK multiplexes "mic" and "line"
inputs. That is, you generally set the codec to grab samples from the mic
input or the line input, which gives you two analog inputs - the left and
right channels. Is there a fast way to "strobe" the line and mic inputs -
using the existing HW on the C6713 DSK board - so that I can effectively get
4 analog inputs?
Thanks in advance for any suggestions!
Bill
Strobing inputs on DSKC6713
Started by ●September 4, 2006
Reply by ●September 5, 20062006-09-05
Bill wrote:> Hello - > > The ADC on the codec on the C6713 DSK multiplexes "mic" and "line" > inputs. That is, you generally set the codec to grab samples from the mic > input or the line input, which gives you two analog inputs - the left and > right channels. Is there a fast way to "strobe" the line and mic inputs - > using the existing HW on the C6713 DSK board - so that I can effectively get > 4 analog inputs? > > Thanks in advance for any suggestions!I don't know the board, but I'd be surprised if there were a way. Consider the input (anti-alias) filter. Is is built in? Can you bypass it and use separate ones on the separate inputs? The filter's settling time is bound to be longer than the sample rate, so it can't be multiplexed. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●September 5, 20062006-09-05
"Bill" <BillTurnip@ACSM.net> wrote in message news:gt6dnRUx8uNEDmHZnZ2dnUVZ_sydnZ2d@comcast.com...> Hello - > > The ADC on the codec on the C6713 DSK multiplexes "mic" and "line" > inputs. That is, you generally set the codec to grab samples from the mic > input or the line input, which gives you two analog inputs - the left and > right channels. Is there a fast way to "strobe" the line and mic inputs - > using the existing HW on the C6713 DSK board - so that I can effectively > get > 4 analog inputs? > > Thanks in advance for any suggestions! > > BillAdding to what has been said about the inevitable discontinuities from switching, the AIC23 mic in is mono. Almost all codecs I'm aware with mic inputs only provide mono mic inputs. Regards, Howard
Reply by ●September 6, 20062006-09-06
Howard and Jerry -
Ahhh, thou art wise. The mic. is indeed mono, so at best I can get 3
inputs, which is OK. But the settling time between switching the ADC to and
from mic input <> line input could be an issue - if I can even do it in
pseudo-realtime. I found the C6713 DSK structure that sets up the AIC23
CODEC, and I know which register/bits to change to toggle between mic input
and line input. By changing the "analog audio path control" register in the
structure, I can set up the board to accept either mic or line input - prior
to compilation. However, I am not sure how to change only this register
during runtime. For example, if switch 3 is 0 on the board, the input is
line input, but if switch 3 is 1, the input is mic input. Once I can do
this, I can measure the latency caused by the multiplexer switching, and see
how much longer this is than the sampling time, settling time, etc. There
has to be a way, but I didn't write all the *.h and *.c support files that
came with the board, so am still fuguring this out.
Thanks guys!
Bill
"Jerry Avins" <jya@ieee.org> wrote in message
news:q7idnUoAe5d8XGDZnZ2dnUVZ_uqdnZ2d@rcn.net...
> Bill wrote:
> > Hello -
> >
> > The ADC on the codec on the C6713 DSK multiplexes "mic" and "line"
> > inputs. That is, you generally set the codec to grab samples from the
mic
> > input or the line input, which gives you two analog inputs - the left
and
> > right channels. Is there a fast way to "strobe" the line and mic
inputs -
> > using the existing HW on the C6713 DSK board - so that I can effectively
get
> > 4 analog inputs?
> >
> > Thanks in advance for any suggestions!
>
> I don't know the board, but I'd be surprised if there were a way.
> Consider the input (anti-alias) filter. Is is built in? Can you bypass
> it and use separate ones on the separate inputs? The filter's settling
> time is bound to be longer than the sample rate, so it can't be
multiplexed.
>
> Jerry
> --
> Engineering is the art of making what you want from things you can get.
> �����������������������������������������������������������������������
Reply by ●September 8, 20062006-09-08
"Bill" <BillTurnip@ACSM.net> wrote in message> By changing the "analog audio path control" register in the > structure, I can set up the board to accept either mic or line input - > prior > to compilation. However, I am not sure how to change only this register > during runtime. For example, if switch 3 is 0 on the board, the input is > line input, but if switch 3 is 1, the input is mic input. Once I can do > this, I can measure the latency caused by the multiplexer switching, and > see > how much longer this is than the sampling time, settling time, etc. There > has to be a way, but I didn't write all the *.h and *.c support files that > came with the board, so am still fuguring this out.Take a look at the AIC23.C file, part of the dsk_app DSK example. On the DSK, the MCBSP0 channel is allocated to run an SPI channel for codec control. Howard
