[tulip] AT-2800TX, I give up
Fri Feb 22 19:46:01 2002
On Fri, 22 Feb 2002, Gert Doering wrote:
> On Wed, Feb 20, 2002 at 07:14:55PM -0500, Donald Becker wrote:
> Talking to A-T has actually led to something :-) - while not to a sample
> card (which would be impractical anyway, due to having to ship that card
> to the US then), I got some information about the hardware used:
Great -- at least we have a contact where we can get some information.
If they are U.S.-based and willing to talk, my phone number is below.
> --------- snip -----------
> Response (Customer Support) - 02/21/2002 02:35 PM
> The Phy we use on the AT-2800 (version 2) is a Broadcom BCM5201.
That appears to be a MII transceiver, although the description on the
Broadcom web page isn't clear on this point.
> Due to our agreement with Broadcom, we cannot supply the data
> sheet. You will have to contact Broadcom for that.
Broadcom has only password-protected access to the datasheets. Hmmmm, I
guess that means "direct customers only".
> Once you have the datasheet, you should be able to
> determine the registers used to activate the Phy.
No, the datasheet won't help, unless there is some magic init sequence
before the PHY starts responding to MII management requests. From the
press release on the 5201 the only magic power saving mode is that it
waits for a signal on the Rx data pair before activating the
> We believe the Phy on the AT-2800 is set to power-up in power-saving mode
> where link is not activated until a driver loads.
What we need to know is what pins on the 21143 are connected to the
PHY's to control the power mode. It isn't anything simple, link a
single signal, or the previous tests with setting all GPIO pins to high
and low would have activated the transceiver.
Donald Becker firstname.lastname@example.org
Scyld Computing Corporation http://www.scyld.com
410 Severn Ave. Suite 210 Second Generation Beowulf Clusters
Annapolis MD 21403 410-990-9993