Re: MPC823E USB controller - delay gates necessary?

To follow up on my original message, it appears that the delay gates are
indeed required for proper host mode operation on the 823e.  Using the
PDIUSBP11A transceiver, we measured about 6-8ns of skew between the single
ended and differential receive lines.  This is apparently too close to the
+8ns limit stated in the MPC850 device errata.

So you may get a lucky transceiver/processor combo that will work, but most
likely you'll need the delay gates on the differential receive pair to get
host mode to work reliably.

-Josh

Josh Horvath wrote:

The device errata for the MPC850 includes the CPM17 bug that states delay
gates are necessary on the USB pins.  The MPC823E errata sheet does not
mention this bug.

I recently read the message by Christian Schroeder about getting the USB
controller working on the TQM823L board (with an 823E processor).  He states
that the delay gates are necessary.  Is this true?

I've got a custom 823E board that doesn't include these gates and I'm having
trouble getting the controller working, and it's not totally clear to me if
it's a hardware or software problem.

Thanks,
-Josh

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/