[En-Nut-Discussion] Internet radio with inbuilt AT73C213 DAC on AT91SAM9260-EK?
Harald Kipp
harald.kipp at egnite.de
Mon Sep 10 10:09:32 CEST 2007
Hi Ajit,
ajitn at inventionlabs.in schrieb:
> I noticed that the Calypso board is required for the DAC and user
> interface. I also noticed that the AT91SAM9260-EK board already has a DAC
> (AT73C213), and a couple of push buttons. I was wondering why the Internet
> Radio code cannot be made to use the on-board DAC instead of using the
> external DAC so that the Calypso board can be done without for this
> project. Has anyone tried this before? Are there any problems with this
> approach?
>
One reason for adding a DAC to Calypso was, that the board should work
with the AT91SAM7XC256-EK as well, which doesn't include a DAC. In the
initial release we used the AT73C213 on the Calypso. We experienced a
problem though.
When looking at chapter "7.4 Timing Specifications" of the AT73C213
datasheet, there is a maximum specified for td1 (max. delay between
rising edges of MCLK and BCLK) of 7.5ns. When the application is
started, this is maintained more or less well. However, after
re-programming the PLL for new sample rates, the delay between both
rising edges is often far beyond the limit. As a result, the audio
output is overlaid by white noise.
As far as I understood the SAM9 hardware, PCK (MCLK at the DAC) and SSC
TK (BCLK at the DAC) are driven by two independent dividers. I have no
idea how to make sure, that both clocks are in sync. We tried a really
large number of different initializations, clock enables and disables,
SSC resets, power mode controls etc. etc. None of them provided any
reproducible phase between these two clocks. As far as I can say,
nothing except an MCU reset clears the internal dividers.
> If anyone has tried this before (created a driver for AT73C213), or anyone
> has created a similar SPI interface for a DAC, I would be happy to know
> about it.
>
So do I,
Harald
More information about the En-Nut-Discussion
mailing list