I have a customer who is trying to understand the mechanisms in place to syncronize USB audio.
His design requires that the 1ms USB frames to be syncronized to an external clock.
In review of the TRM, several questions have arised related to trying to determine this.
One option that was discussed is possibility to synchronize the HFCLKIN to his external clock via a clock synthesizer. This would create the HFCLKIN from his external clock guaranteeing them being synchronous.
We would like to understand the clocks related to generating the 1ms SOF, and if above approach is the best method with these design constraints.
I have included customer comment and questions below:
------------
I'm trying to
determine how the 1 ms USB Start-Of-Frame (SOF) packets are generated in the
OMAP USB host controller and which clock is used to generate this
timing.
Section 23.1.2.1.1
of the OMAP3530 TRM specifies that the functional clock for the USB
interface is the 60 MHz clock generated from the external ULPI
transceiver.
In
our products, we plan to use the TPS65950 USB PHY. Figure 3-2 and
3-3 of the TPS65950 TRM show that the ULPI 60 MHz clock is derived from the
26 MHz_USB clock generated from the HFCLKIN source. It is not clear if
this is synchronized to the 26 MHz clock.
We require the 1 ms
USB frames to be synchronized with our E1/T1 line rate clock.
My analysis seems to
indicate that if I
synchronize the 26 MHz HFCLKIN clock to my E1/T1 line rate clock
using some sort of external synthesizer/synchronizer, then the USB 1 ms packets
will be synchronized to that clock.
Can you please give
me some feedback on this clock synchronization approach?
I was reviewing the Clock description in the TRM for the
TPS65950 (Chapter 3) again and I have another question. Figure 3-3 shows a
64kHz clock that is sent from the Voice 26 MHz clock divider to the ULPI
interface and then out somewhere else. Can you please find out the
underlying design intent for this 64 KHz clock. Maybe there is already
some sort of synchronization of USB being done between the Audio Codecs and the
USB PHY.
----
Thanks & Regards,
Wade
