Hi !
I am working on a board that has the same ethernet configuration as
the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from National.
I simply can't get the ethernet up and running.
I know for sure that the register management interface MDIO/MDC works
out fine as I am capeable of configurering and verifying the settings
in the PHYTER from the LPC2368
Furthermore National has a Integrity Tool from which it is possible
via a parallel port on a PC to read out the registers in the PHYTER.
These have been reviewed by National and everything seems to be fine.
Meaning that the physical magnetic side of the PHYTER seems to work
as well. Good Link (green LED) and activity (yellow LED) blinks and
so on.
I suspect that it has something to do with the RMII interface.
I am using Rowley Crossworks and the GNUEasyWeb to test. It works
just fine on the MCB2300 but not one my board.
The LPC is a revision "-" on both boards.
The PCB is four layer.
Does anyone have experience in this specific setup ?
Any help would be highly appreciated.
Thanks
LPC2368 and DP83848
Started by ●November 23, 2007
Reply by ●November 23, 20072007-11-23
morten_kahr wrote:
> Hi !
>
> I am working on a board that has the same ethernet configuration as
> the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from National.
>
> I simply can't get the ethernet up and running.
I know how that can be. ;-)
> I know for sure that the register management interface MDIO/MDC works
> out fine as I am capeable of configurering and verifying the settings
> in the PHYTER from the LPC2368
> Furthermore National has a Integrity Tool from which it is possible
> via a parallel port on a PC to read out the registers in the PHYTER.
> These have been reviewed by National and everything seems to be fine.
> Meaning that the physical magnetic side of the PHYTER seems to work
> as well. Good Link (green LED) and activity (yellow LED) blinks and
> so on.
> I suspect that it has something to do with the RMII interface.
I had problems like this using the sample code from FreeRTOS on an Olimex
board using a Micrel PHY. I had to add code to reset and configure the
MAC_MCFG register clock divider. Until then the PHY and MAC couldn't really
communicate, but the PHY would power up ready to go and would link to my hub
by itself. Some specification lists the maximum clock at 2.5MHz, but the
datasheet for my Micrel allows up to 5MHz clocking of commands so YMMV.
BTW, I'm using that revision ("-") LPC2378 and it has some serious issues,
be sure to look at the Errata sheet.
> I am using Rowley Crossworks and the GNUEasyWeb to test. It works
> just fine on the MCB2300 but not one my board.
>
> The LPC is a revision "-" on both boards.
> The PCB is four layer.
>
> Does anyone have experience in this specific setup ?
No experience with that PHY, but I did manage to get mine receiving packets.
It was a real hair-puller as everything seems just fine, just no packet
transfers. If you want I can send the code I'm using right now with my
board. The User Manual has a pretty good description of how to initialize
it all, it just took me a while to find it. The FreeRTOS code doesn't
follow the User Manual procedures exactly so there may be something missing
from my changes to the FreeRTOS stuff. I intend to completely rewrite it
per the User Manual, I just haven't gotten to it yet. ;-)
> Hi !
>
> I am working on a board that has the same ethernet configuration as
> the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from National.
>
> I simply can't get the ethernet up and running.
I know how that can be. ;-)
> I know for sure that the register management interface MDIO/MDC works
> out fine as I am capeable of configurering and verifying the settings
> in the PHYTER from the LPC2368
> Furthermore National has a Integrity Tool from which it is possible
> via a parallel port on a PC to read out the registers in the PHYTER.
> These have been reviewed by National and everything seems to be fine.
> Meaning that the physical magnetic side of the PHYTER seems to work
> as well. Good Link (green LED) and activity (yellow LED) blinks and
> so on.
> I suspect that it has something to do with the RMII interface.
I had problems like this using the sample code from FreeRTOS on an Olimex
board using a Micrel PHY. I had to add code to reset and configure the
MAC_MCFG register clock divider. Until then the PHY and MAC couldn't really
communicate, but the PHY would power up ready to go and would link to my hub
by itself. Some specification lists the maximum clock at 2.5MHz, but the
datasheet for my Micrel allows up to 5MHz clocking of commands so YMMV.
BTW, I'm using that revision ("-") LPC2378 and it has some serious issues,
be sure to look at the Errata sheet.
> I am using Rowley Crossworks and the GNUEasyWeb to test. It works
> just fine on the MCB2300 but not one my board.
>
> The LPC is a revision "-" on both boards.
> The PCB is four layer.
>
> Does anyone have experience in this specific setup ?
No experience with that PHY, but I did manage to get mine receiving packets.
It was a real hair-puller as everything seems just fine, just no packet
transfers. If you want I can send the code I'm using right now with my
board. The User Manual has a pretty good description of how to initialize
it all, it just took me a while to find it. The FreeRTOS code doesn't
follow the User Manual procedures exactly so there may be something missing
from my changes to the FreeRTOS stuff. I intend to completely rewrite it
per the User Manual, I just haven't gotten to it yet. ;-)
Reply by ●November 23, 20072007-11-23
Hi!
In my design, bassed on the MCB2300 bord, ive forgetten two issues:
- The 2,2 kohms pull-up for the PIN 39 of the DP83848
- The 1.5 kohms pull-up for the MDIO PIN 30 of the DP83848.
Anyway, I suggest to look in detail all pull-ups pulls-down for config
of the chip at startup.
Im sorry for my english...
Greetings and good luck!
German
morten_kahr escribi
>
> Hi !
>
> I am working on a board that has the same ethernet configuration as
> the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from National.
>
> I simply can't get the ethernet up and running.
> I know for sure that the register management interface MDIO/MDC works
> out fine as I am capeable of configurering and verifying the settings
> in the PHYTER from the LPC2368
> Furthermore National has a Integrity Tool from which it is possible
> via a parallel port on a PC to read out the registers in the PHYTER.
> These have been reviewed by National and everything seems to be fine.
> Meaning that the physical magnetic side of the PHYTER seems to work
> as well. Good Link (green LED) and activity (yellow LED) blinks and
> so on.
> I suspect that it has something to do with the RMII interface.
>
> I am using Rowley Crossworks and the GNUEasyWeb to test. It works
> just fine on the MCB2300 but not one my board.
>
> The LPC is a revision "-" on both boards.
> The PCB is four layer.
>
> Does anyone have experience in this specific setup ?
>
> Any help would be highly appreciated.
> Thanks
>
>
In my design, bassed on the MCB2300 bord, ive forgetten two issues:
- The 2,2 kohms pull-up for the PIN 39 of the DP83848
- The 1.5 kohms pull-up for the MDIO PIN 30 of the DP83848.
Anyway, I suggest to look in detail all pull-ups pulls-down for config
of the chip at startup.
Im sorry for my english...
Greetings and good luck!
German
morten_kahr escribi
>
> Hi !
>
> I am working on a board that has the same ethernet configuration as
> the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from National.
>
> I simply can't get the ethernet up and running.
> I know for sure that the register management interface MDIO/MDC works
> out fine as I am capeable of configurering and verifying the settings
> in the PHYTER from the LPC2368
> Furthermore National has a Integrity Tool from which it is possible
> via a parallel port on a PC to read out the registers in the PHYTER.
> These have been reviewed by National and everything seems to be fine.
> Meaning that the physical magnetic side of the PHYTER seems to work
> as well. Good Link (green LED) and activity (yellow LED) blinks and
> so on.
> I suspect that it has something to do with the RMII interface.
>
> I am using Rowley Crossworks and the GNUEasyWeb to test. It works
> just fine on the MCB2300 but not one my board.
>
> The LPC is a revision "-" on both boards.
> The PCB is four layer.
>
> Does anyone have experience in this specific setup ?
>
> Any help would be highly appreciated.
> Thanks
>
>
Reply by ●November 24, 20072007-11-24
Hi,
I also use a DP83848, but the arm is a LPC2378. Anyway, as far as i know LPC23xx bugs are commom for all devices (only the EMC write bug only hapens in the 2378, as the other devs dont have an EMC), well I have many problems with ethernet runing in 100MBps, but there's a post here of someone telling to configure the PLL autput to be around 196MHz (if my memory doesnt fail), and it owrked out fine, but the frequence (196MHz) is out of the range especified in the user manual. Another problem can be with the ethernet DMA, for the DMA to work properly, the descriptors and all must be in the ethernet RAM (wich is 16Kb), this could also be a problem. My board only uses two layers and after these changes the ethernet is working fine (though I'm not using the PLL output in the forbindden value and so not using 100MBs).
Regards,
Fernando Almeida.
morten_kahr escreveu:
Hi !
I am working on a board that has the same ethernet configuration as
the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from National.
I simply can't get the ethernet up and running.
I know for sure that the register management interface MDIO/MDC works
out fine as I am capeable of configurering and verifying the settings
in the PHYTER from the LPC2368
Furthermore National has a Integrity Tool from which it is possible
via a parallel port on a PC to read out the registers in the PHYTER.
These have been reviewed by National and everything seems to be fine.
Meaning that the physical magnetic side of the PHYTER seems to work
as well. Good Link (green LED) and activity (yellow LED) blinks and
so on.
I suspect that it has something to do with the RMII interface.
I am using Rowley Crossworks and the GNUEasyWeb to test. It works
just fine on the MCB2300 but not one my board.
The LPC is a revision "-" on both boards.
The PCB is four layer.
Does anyone have experience in this specific setup ?
Any help would be highly appreciated.
Thanks
---------------------------------
Abra sua conta no Yahoo! Mail, o ico sem limite de espa para armazenamento!
I also use a DP83848, but the arm is a LPC2378. Anyway, as far as i know LPC23xx bugs are commom for all devices (only the EMC write bug only hapens in the 2378, as the other devs dont have an EMC), well I have many problems with ethernet runing in 100MBps, but there's a post here of someone telling to configure the PLL autput to be around 196MHz (if my memory doesnt fail), and it owrked out fine, but the frequence (196MHz) is out of the range especified in the user manual. Another problem can be with the ethernet DMA, for the DMA to work properly, the descriptors and all must be in the ethernet RAM (wich is 16Kb), this could also be a problem. My board only uses two layers and after these changes the ethernet is working fine (though I'm not using the PLL output in the forbindden value and so not using 100MBs).
Regards,
Fernando Almeida.
morten_kahr escreveu:
Hi !
I am working on a board that has the same ethernet configuration as
the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from National.
I simply can't get the ethernet up and running.
I know for sure that the register management interface MDIO/MDC works
out fine as I am capeable of configurering and verifying the settings
in the PHYTER from the LPC2368
Furthermore National has a Integrity Tool from which it is possible
via a parallel port on a PC to read out the registers in the PHYTER.
These have been reviewed by National and everything seems to be fine.
Meaning that the physical magnetic side of the PHYTER seems to work
as well. Good Link (green LED) and activity (yellow LED) blinks and
so on.
I suspect that it has something to do with the RMII interface.
I am using Rowley Crossworks and the GNUEasyWeb to test. It works
just fine on the MCB2300 but not one my board.
The LPC is a revision "-" on both boards.
The PCB is four layer.
Does anyone have experience in this specific setup ?
Any help would be highly appreciated.
Thanks
---------------------------------
Abra sua conta no Yahoo! Mail, o ico sem limite de espa para armazenamento!
Reply by ●November 26, 20072007-11-26
Hi Michael !
I would really appreciate that as I am out of of ideas at this point
I'll try check up on the MAC_MCFG.
I suspect that the problem lies in the RMII interface and perhaps in
the layout of the TxD, RxD... So forcing the speed down might help.
Application example can be sent to e-mail address.
Thanks in advance
--- In l..., "michael brown" wrote:
>
> morten_kahr wrote:
> > Hi !
> >
> > I am working on a board that has the same ethernet configuration
as
> > the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from
National.
> >
> > I simply can't get the ethernet up and running.
>
> I know how that can be. ;-)
>
> > I know for sure that the register management interface MDIO/MDC
works
> > out fine as I am capeable of configurering and verifying the
settings
> > in the PHYTER from the LPC2368
> > Furthermore National has a Integrity Tool from which it is
possible
> > via a parallel port on a PC to read out the registers in the
PHYTER.
> > These have been reviewed by National and everything seems to be
fine.
> > Meaning that the physical magnetic side of the PHYTER seems to
work
> > as well. Good Link (green LED) and activity (yellow LED) blinks
and
> > so on.
> > I suspect that it has something to do with the RMII interface.
>
> I had problems like this using the sample code from FreeRTOS on an
Olimex
> board using a Micrel PHY. I had to add code to reset and configure
the
> MAC_MCFG register clock divider. Until then the PHY and MAC
couldn't really
> communicate, but the PHY would power up ready to go and would link
to my hub
> by itself. Some specification lists the maximum clock at 2.5MHz,
but the
> datasheet for my Micrel allows up to 5MHz clocking of commands so
YMMV.
> BTW, I'm using that revision ("-") LPC2378 and it has some serious
issues,
> be sure to look at the Errata sheet.
>
> > I am using Rowley Crossworks and the GNUEasyWeb to test. It works
> > just fine on the MCB2300 but not one my board.
> >
> > The LPC is a revision "-" on both boards.
> > The PCB is four layer.
> >
> > Does anyone have experience in this specific setup ?
>
> No experience with that PHY, but I did manage to get mine receiving
packets.
> It was a real hair-puller as everything seems just fine, just no
packet
> transfers. If you want I can send the code I'm using right now
with my
> board. The User Manual has a pretty good description of how to
initialize
> it all, it just took me a while to find it. The FreeRTOS code
doesn't
> follow the User Manual procedures exactly so there may be something
missing
> from my changes to the FreeRTOS stuff. I intend to completely
rewrite it
> per the User Manual, I just haven't gotten to it yet. ;-)
>
I would really appreciate that as I am out of of ideas at this point
I'll try check up on the MAC_MCFG.
I suspect that the problem lies in the RMII interface and perhaps in
the layout of the TxD, RxD... So forcing the speed down might help.
Application example can be sent to e-mail address.
Thanks in advance
--- In l..., "michael brown" wrote:
>
> morten_kahr wrote:
> > Hi !
> >
> > I am working on a board that has the same ethernet configuration
as
> > the KEIL MCB2300 board. A LPC2368 and a DP83848 PHYTER from
National.
> >
> > I simply can't get the ethernet up and running.
>
> I know how that can be. ;-)
>
> > I know for sure that the register management interface MDIO/MDC
works
> > out fine as I am capeable of configurering and verifying the
settings
> > in the PHYTER from the LPC2368
> > Furthermore National has a Integrity Tool from which it is
possible
> > via a parallel port on a PC to read out the registers in the
PHYTER.
> > These have been reviewed by National and everything seems to be
fine.
> > Meaning that the physical magnetic side of the PHYTER seems to
work
> > as well. Good Link (green LED) and activity (yellow LED) blinks
and
> > so on.
> > I suspect that it has something to do with the RMII interface.
>
> I had problems like this using the sample code from FreeRTOS on an
Olimex
> board using a Micrel PHY. I had to add code to reset and configure
the
> MAC_MCFG register clock divider. Until then the PHY and MAC
couldn't really
> communicate, but the PHY would power up ready to go and would link
to my hub
> by itself. Some specification lists the maximum clock at 2.5MHz,
but the
> datasheet for my Micrel allows up to 5MHz clocking of commands so
YMMV.
> BTW, I'm using that revision ("-") LPC2378 and it has some serious
issues,
> be sure to look at the Errata sheet.
>
> > I am using Rowley Crossworks and the GNUEasyWeb to test. It works
> > just fine on the MCB2300 but not one my board.
> >
> > The LPC is a revision "-" on both boards.
> > The PCB is four layer.
> >
> > Does anyone have experience in this specific setup ?
>
> No experience with that PHY, but I did manage to get mine receiving
packets.
> It was a real hair-puller as everything seems just fine, just no
packet
> transfers. If you want I can send the code I'm using right now
with my
> board. The User Manual has a pretty good description of how to
initialize
> it all, it just took me a while to find it. The FreeRTOS code
doesn't
> follow the User Manual procedures exactly so there may be something
missing
> from my changes to the FreeRTOS stuff. I intend to completely
rewrite it
> per the User Manual, I just haven't gotten to it yet. ;-)
>
