Hello,
For a while now, I have been looking at the option of using the
AT91SAM with the ethernet option. To do so, I have had to glue a chip
which can now be hooked onto the actual ethernet socket/ cable. I
could not find the DM9161 or something which has a driver for the
AT91SAM, but instead I am using the 8201 which is also 10/100. While I
had to put in enough time to wire the chip to AT91SAM7x, I have been
unable to put enough time to bring it up via some kind of a modified
driver, and I cannot find a way to get to respond.
So to those of you out there, my question is, that is it actually
possible to make this 8201 work? Has anybody done some kind of work
with the driver to get it work? What other similar devices can be used
instead of 8201, and is there a driver for any other chip to be used
with the AT91SAM7x ?
Thank you and regards.
about AT91SAM and ethernet
Started by ●November 2, 2008
Reply by ●November 2, 20082008-11-02
tasbihmr wrote:
> What other similar devices can be used instead of 8201, and is
> there a driver for any other chip to be used with the AT91SAM7x ?
I've used the Davicom DM9161A (on the evaluation kit), the Intel
LXT972A and the SMSC LAN8700i with the AT91SAM7XC256. All these PHY
chips are similar; there are a few differences between them, but
nothing too tricky. There's not much of a driver required - I have a
simple state machine that detects LINK, loss of LINK, speed, duplex,
etc. and maintains the EMAC_NCFGR.
HTH, Nick
> What other similar devices can be used instead of 8201, and is
> there a driver for any other chip to be used with the AT91SAM7x ?
I've used the Davicom DM9161A (on the evaluation kit), the Intel
LXT972A and the SMSC LAN8700i with the AT91SAM7XC256. All these PHY
chips are similar; there are a few differences between them, but
nothing too tricky. There's not much of a driver required - I have a
simple state machine that detects LINK, loss of LINK, speed, duplex,
etc. and maintains the EMAC_NCFGR.
HTH, Nick
Reply by ●November 2, 20082008-11-02
Hello,
Thank you for sharing that Nick, it seems that you have good idea of
how to interface the PHY chip with the SAM7x. There are differences as
you say, as I have not been able to simply compile the evaluation
software and make it work with the 8201, even though I had the idea
that since the interfaces are universal, it should be easy to do so.
In short, I do't know to put the time and change the present driver of
DM9161 to work with the 8201, or just try to look for DM9161. What are
the differences between the chips that you have used Nick? I mean
surely the reset functionality is unique in each case.
Thank you and regards.
--- In A..., Nick Howes wrote:
>
> tasbihmr wrote:
>
> > What other similar devices can be used instead of 8201, and is
> > there a driver for any other chip to be used with the AT91SAM7x ?
> I've used the Davicom DM9161A (on the evaluation kit), the Intel
> LXT972A and the SMSC LAN8700i with the AT91SAM7XC256. All these PHY
> chips are similar; there are a few differences between them, but
> nothing too tricky. There's not much of a driver required - I have a
> simple state machine that detects LINK, loss of LINK, speed, duplex,
> etc. and maintains the EMAC_NCFGR.
>
> HTH, Nick
>
Thank you for sharing that Nick, it seems that you have good idea of
how to interface the PHY chip with the SAM7x. There are differences as
you say, as I have not been able to simply compile the evaluation
software and make it work with the 8201, even though I had the idea
that since the interfaces are universal, it should be easy to do so.
In short, I do't know to put the time and change the present driver of
DM9161 to work with the 8201, or just try to look for DM9161. What are
the differences between the chips that you have used Nick? I mean
surely the reset functionality is unique in each case.
Thank you and regards.
--- In A..., Nick Howes wrote:
>
> tasbihmr wrote:
>
> > What other similar devices can be used instead of 8201, and is
> > there a driver for any other chip to be used with the AT91SAM7x ?
> I've used the Davicom DM9161A (on the evaluation kit), the Intel
> LXT972A and the SMSC LAN8700i with the AT91SAM7XC256. All these PHY
> chips are similar; there are a few differences between them, but
> nothing too tricky. There's not much of a driver required - I have a
> simple state machine that detects LINK, loss of LINK, speed, duplex,
> etc. and maintains the EMAC_NCFGR.
>
> HTH, Nick
>
Reply by ●November 2, 20082008-11-02
Hi
The standard registers of all PHY are defined by the Ethernet
specification and so (as long as advanced features of a specific chip)
are not used a single driver should work.
Where this is a big difference is the way that they default to a
standard mode and obtain their address - this is usually over
pullup/down resistors on certain pins (sometimes MII bus lines and
sometimes LED lines).
Often the driver will need to have a little help to override the
default settings.
The following give more details specifically for working with the SAM7X:
http://www.utasker.com/forum/index.php?topic1.0
Regards
Mark
http://www.uTasker.com
http://www.uTasker.com/forum/
http://demo.uTasker.com
>
> Hello,
>
> Thank you for sharing that Nick, it seems that you have good idea of
> how to interface the PHY chip with the SAM7x. There are differences as
> you say, as I have not been able to simply compile the evaluation
> software and make it work with the 8201, even though I had the idea
> that since the interfaces are universal, it should be easy to do so.
> In short, I do't know to put the time and change the present driver of
> DM9161 to work with the 8201, or just try to look for DM9161. What are
> the differences between the chips that you have used Nick? I mean
> surely the reset functionality is unique in each case.
>
> Thank you and regards.
The standard registers of all PHY are defined by the Ethernet
specification and so (as long as advanced features of a specific chip)
are not used a single driver should work.
Where this is a big difference is the way that they default to a
standard mode and obtain their address - this is usually over
pullup/down resistors on certain pins (sometimes MII bus lines and
sometimes LED lines).
Often the driver will need to have a little help to override the
default settings.
The following give more details specifically for working with the SAM7X:
http://www.utasker.com/forum/index.php?topic1.0
Regards
Mark
http://www.uTasker.com
http://www.uTasker.com/forum/
http://demo.uTasker.com
>
> Hello,
>
> Thank you for sharing that Nick, it seems that you have good idea of
> how to interface the PHY chip with the SAM7x. There are differences as
> you say, as I have not been able to simply compile the evaluation
> software and make it work with the 8201, even though I had the idea
> that since the interfaces are universal, it should be easy to do so.
> In short, I do't know to put the time and change the present driver of
> DM9161 to work with the 8201, or just try to look for DM9161. What are
> the differences between the chips that you have used Nick? I mean
> surely the reset functionality is unique in each case.
>
> Thank you and regards.
Reply by ●November 2, 20082008-11-02
Yes, the LED lines vary and in some chips you need to prewire some address lines
as well. Two important items: (1) the RMII style interface didn't work on
the sam7x a while back. I don't know if that was fixed in a later rev or is
still broken. MII works great. I too had trouble locating a DM9161A so I used
the intel LXT972A instead and was careful with the layout. I followed the PHY
vendor's layout recommendations and was sure to keep all the traces between
the PHY and the SAM7X about the same length, and short. On first pass everything
worked ok. I found Atmel's sample code for the EMAC to be helpful. Adding
an RTOS and uIP I quickly had a fully working TCP/IP aware device. Of course,
your application may vary.
Chris
----- Original Message -----
From: Mark Butcher
To: A...
Sent: Sunday, November 02, 2008 3:16 PM
Subject: [AT91SAM] Re:about AT91SAM and ethernet
Hi
The standard registers of all PHY are defined by the Ethernet
specification and so (as long as advanced features of a specific chip)
are not used a single driver should work.
Where this is a big difference is the way that they default to a
standard mode and obtain their address - this is usually over
pullup/down resistors on certain pins (sometimes MII bus lines and
sometimes LED lines).
Often the driver will need to have a little help to override the
default settings.
The following give more details specifically for working with the SAM7X:
http://www.utasker.com/forum/index.php?topic1.0
Regards
Mark
http://www.uTasker.com
http://www.uTasker.com/forum/
http://demo.uTasker.com
>
> Hello,
>
> Thank you for sharing that Nick, it seems that you have good idea of
> how to interface the PHY chip with the SAM7x. There are differences as
> you say, as I have not been able to simply compile the evaluation
> software and make it work with the 8201, even though I had the idea
> that since the interfaces are universal, it should be easy to do so.
> In short, I do't know to put the time and change the present driver of
> DM9161 to work with the 8201, or just try to look for DM9161. What are
> the differences between the chips that you have used Nick? I mean
> surely the reset functionality is unique in each case.
>
> Thank you and regards.
--------
7:56 PM
Chris
----- Original Message -----
From: Mark Butcher
To: A...
Sent: Sunday, November 02, 2008 3:16 PM
Subject: [AT91SAM] Re:about AT91SAM and ethernet
Hi
The standard registers of all PHY are defined by the Ethernet
specification and so (as long as advanced features of a specific chip)
are not used a single driver should work.
Where this is a big difference is the way that they default to a
standard mode and obtain their address - this is usually over
pullup/down resistors on certain pins (sometimes MII bus lines and
sometimes LED lines).
Often the driver will need to have a little help to override the
default settings.
The following give more details specifically for working with the SAM7X:
http://www.utasker.com/forum/index.php?topic1.0
Regards
Mark
http://www.uTasker.com
http://www.uTasker.com/forum/
http://demo.uTasker.com
>
> Hello,
>
> Thank you for sharing that Nick, it seems that you have good idea of
> how to interface the PHY chip with the SAM7x. There are differences as
> you say, as I have not been able to simply compile the evaluation
> software and make it work with the 8201, even though I had the idea
> that since the interfaces are universal, it should be easy to do so.
> In short, I do't know to put the time and change the present driver of
> DM9161 to work with the 8201, or just try to look for DM9161. What are
> the differences between the chips that you have used Nick? I mean
> surely the reset functionality is unique in each case.
>
> Thank you and regards.
--------
7:56 PM
Reply by ●November 2, 20082008-11-02
Hello,
The link mentioned in the reply by Mark does not work for me, maybe it is not accesible from here, but I assume it is something to do with utasker?
Thank you and regards.
--- On Sun, 11/2/08, Mark Butcher wrote:
From: Mark Butcher
Subject: [AT91SAM] Re:about AT91SAM and ethernet
To: A...
Date: Sunday, November 2, 2008, 12:16 PM
Hi
The standard registers of all PHY are defined by the Ethernet
specification and so (as long as advanced features of a specific chip)
are not used a single driver should work.
Where this is a big difference is the way that they default to a
standard mode and obtain their address - this is usually over
pullup/down resistors on certain pins (sometimes MII bus lines and
sometimes LED lines).
Often the driver will need to have a little help to override the
default settings.
The following give more details specifically for working with the SAM7X:
http://www.utasker. com/forum/ index.php? topic1. 0
Regards
Mark
http://www.uTasker. com
http://www.uTasker. com/forum/
http://demo. uTasker.com
>
> Hello,
>
> Thank you for sharing that Nick, it seems that you have good idea of
> how to interface the PHY chip with the SAM7x. There are differences as
> you say, as I have not been able to simply compile the evaluation
> software and make it work with the 8201, even though I had the idea
> that since the interfaces are universal, it should be easy to do so.
> In short, I do't know to put the time and change the present driver of
> DM9161 to work with the 8201, or just try to look for DM9161. What are
> the differences between the chips that you have used Nick? I mean
> surely the reset functionality is unique in each case.
>
> Thank you and regards.
The link mentioned in the reply by Mark does not work for me, maybe it is not accesible from here, but I assume it is something to do with utasker?
Thank you and regards.
--- On Sun, 11/2/08, Mark Butcher wrote:
From: Mark Butcher
Subject: [AT91SAM] Re:about AT91SAM and ethernet
To: A...
Date: Sunday, November 2, 2008, 12:16 PM
Hi
The standard registers of all PHY are defined by the Ethernet
specification and so (as long as advanced features of a specific chip)
are not used a single driver should work.
Where this is a big difference is the way that they default to a
standard mode and obtain their address - this is usually over
pullup/down resistors on certain pins (sometimes MII bus lines and
sometimes LED lines).
Often the driver will need to have a little help to override the
default settings.
The following give more details specifically for working with the SAM7X:
http://www.utasker. com/forum/ index.php? topic1. 0
Regards
Mark
http://www.uTasker. com
http://www.uTasker. com/forum/
http://demo. uTasker.com
>
> Hello,
>
> Thank you for sharing that Nick, it seems that you have good idea of
> how to interface the PHY chip with the SAM7x. There are differences as
> you say, as I have not been able to simply compile the evaluation
> software and make it work with the 8201, even though I had the idea
> that since the interfaces are universal, it should be easy to do so.
> In short, I do't know to put the time and change the present driver of
> DM9161 to work with the 8201, or just try to look for DM9161. What are
> the differences between the chips that you have used Nick? I mean
> surely the reset functionality is unique in each case.
>
> Thank you and regards.
