USB and ethernet ports are not working - D14 led not powered

206 views
Skip to first unread message

Sven Bock

unread,
Oct 5, 2015, 11:50:16 AM10/5/15
to BeagleBoard
Hello, 

I do not manage to enable my USB or ethernet ports on my beagleboard xm rev C. The D14 Led is off. I use an external power supply of 5V 4A.

I installed Ubuntu 14.04 according to:
This one boots into a command line.

I also tried it with the production image:
This one correctly boots the graphical interface

All usb ports, as well as the ethernet are dead.
I successfully managed connecting with a USB to serial adapter to the beagleboard. So I actually have a terminal and be able to type!
According to the linux configuration parameters below for USB, the ports are enabled.
/sys/bus/usb/devices/usb1/power/autosuspend_delay_m
/sys/bus/usb/devices/usb1/power/control 

Is there a beagleboard specific configuration I have missed? 
Do I need a uEnv.txt in the boot partition to power up the usb and ethernet during boot? If yes, what is the essential code for enabling the usb and ethernet?

Should they be working by default after flashing the image? 
Could that be just a configuration problem or more a hardware issue?

Any hint, what is going on with my board is greatly appreciated.

Thanks!
Sven

Roger Quadros

unread,
Oct 6, 2015, 3:25:10 AM10/6/15
to beagl...@googlegroups.com
Hi Sven,

On 05/10/15 18:50, Sven Bock wrote:
> Hello,
>
> I do not manage to enable my USB or ethernet ports on my beagleboard xm rev C. The D14 Led is off. I use an external power supply of 5V 4A.
>
> I installed Ubuntu 14.04 according to:
> http://elinux.org/BeagleBoardUbuntu
> This one boots into a command line.
>
> I also tried it with the production image:
> http://elinux.org/Beagleboard:BeagleBoard-xM
> This one correctly boots the graphical interface
>
> All usb ports, as well as the ethernet are dead.
> I successfully managed connecting with a USB to serial adapter to the beagleboard. So I actually have a terminal and be able to type!
> According to the linux configuration parameters below for USB, the ports are enabled.
> /sys/bus/usb/devices/usb1/power/autosuspend_delay_m
> /sys/bus/usb/devices/usb1/power/control
>
> Is there a beagleboard specific configuration I have missed?
> Do I need a uEnv.txt in the boot partition to power up the usb and ethernet during boot? If yes, what is the essential code for enabling the usb and ethernet?

You don't need any uEnv. All you need in the kernel is the USB EHCI drivers.

CONFIG_USB_EHCI_HCD=y
CONFIG_USB_EHCI_HCD_OMAP=y

You will also need the TWL PMIC GPIO driver
CONFIG_GPIO_TWL4030=y
>
> Should they be working by default after flashing the image?
> Could that be just a configuration problem or more a hardware issue?

led D14 must turn ON when USB hub is powered. Looks like the software
isn't turning on power for the USB hub so I wouldn't suspect hardware yet.

>
> Any hint, what is going on with my board is greatly appreciated.

Can you please paste your kernel boot log in pastebin and the output of
cat /sys/kernel/debug/gpio

cheers,
-roger

Sven Bock

unread,
Oct 7, 2015, 9:13:58 AM10/7/15
to BeagleBoard, rog...@ti.com
Hello Roger,

thanks for your reply.



You don't need any uEnv. All you need in the kernel is the USB EHCI drivers.

CONFIG_USB_EHCI_HCD=y
CONFIG_USB_EHCI_HCD_OMAP=y

You will also need the TWL PMIC GPIO driver
CONFIG_GPIO_TWL4030=y

Where should I find this config? Also in "/sys/kernel"?

led D14 must turn ON when USB hub is powered. Looks like the software
isn't turning on power for the USB hub so I wouldn't suspect hardware yet.

It is not turning on. 

Can you please paste your kernel boot log in pastebin and the output of
        cat /sys/kernel/debug/gpio
 
I attached the boot.log and the dmesg
Here is the output of gpio:
 
root@arm:/sys/kernel/debug# cat gpio
GPIOs 0-31, platform/49050000.gpio, gpio:

GPIOs 32-63, platform/49052000.gpio, gpio:

GPIOs 64-95, platform/49054000.gpio, gpio:

GPIOs 96-127, platform/49056000.gpio, gpio:
 gpio-115 (reset               ) out hi   
 gpio-117 (?                   ) out lo   
 gpio-118 (?                   ) out lo   

GPIOs 128-159, platform/49058000.gpio, gpio:

GPIOs 160-191, platform/48310000.gpio, gpio:
 gpio-164 (user                ) in  lo IRQ

GPIOs 1004-1023, platform/twl4030-gpio, twl4030, can sleep:
 gpio-1006 (tfp410 PD           ) out lo   
 gpio-1022 (hsusb2_vbus         ) out hi   

Do you see anything wrong in this configuration?

best,
sven
 
boot.log
dmesg

Roger Quadros

unread,
Oct 7, 2015, 9:54:48 AM10/7/15
to Sven Bock, BeagleBoard
Sven,

On 07/10/15 16:13, Sven Bock wrote:
> Hello Roger,
>
> thanks for your reply.
>
>
> You don't need any uEnv. All you need in the kernel is the USB EHCI drivers.
>
> CONFIG_USB_EHCI_HCD=y
> CONFIG_USB_EHCI_HCD_OMAP=y
>
> You will also need the TWL PMIC GPIO driver
> CONFIG_GPIO_TWL4030=y
>
>
> Where should I find this config? Also in "/sys/kernel"?
>

This is when you build the kernel. But never mind. From your cat gpio it looks like
it is enabled as it shows twl4030-gpio.

>
> led D14 must turn ON when USB hub is powered. Looks like the software
> isn't turning on power for the USB hub so I wouldn't suspect hardware yet.
>
> It is not turning on.
>
>
> Can you please paste your kernel boot log in pastebin and the output of
> cat /sys/kernel/debug/gpio
>
>
> I attached the boot.log and the dmesg
> Here is the output of gpio:
>
> root@arm:/sys/kernel/debug# cat gpio
> GPIOs 0-31, platform/49050000.gpio, gpio:
>
> GPIOs 32-63, platform/49052000.gpio, gpio:
>
> GPIOs 64-95, platform/49054000.gpio, gpio:
>
> GPIOs 96-127, platform/49056000.gpio, gpio:
> gpio-115 (reset ) out hi
> gpio-117 (? ) out lo
> gpio-118 (? ) out lo
>
> GPIOs 128-159, platform/49058000.gpio, gpio:
>
> GPIOs 160-191, platform/48310000.gpio, gpio:
> gpio-164 (user ) in lo IRQ
>
> GPIOs 1004-1023, platform/twl4030-gpio, twl4030, can sleep:
> gpio-1006 (tfp410 PD ) out lo
> gpio-1022 (hsusb2_vbus ) out hi
>
> Do you see anything wrong in this configuration?

This is fine. From your kernel log I see this

[ 3.789886] ehci-omap: OMAP-EHCI Host Controller driver
[ 3.790405] ehci-omap 48064800.ehci: Can't get PHY device for port 1: -517

This means that the EHCI USB controller couldn't get the PHY device
and has bailed out. It seems to occur in mainline as well.

I'll investigate more and keep you updated.

cheers,
-roger

Roger Quadros

unread,
Oct 7, 2015, 9:58:48 AM10/7/15
to Sven Bock, BeagleBoard
Well I'm wrong. A bit later it seems to have been registered fine.

So hsusb2_vbus line should have gone low to enable the USB HUB power
and hence D14. That doesn't seem to be happening.

cheers,
-roger

Sven Bock

unread,
Oct 7, 2015, 10:15:49 AM10/7/15
to beagl...@googlegroups.com
Thank you so much for your support Roger.
I assume with the stock rom the D14 should turn on by default. 
Could there be something set on the board, independent of the sd card?
--
Cordialement / Best Regards / Mit freundlichen Grüßen /
Sven Bock M.Sc.
Research Engineer
Génération Robots | HumaRobotics

Roger Quadros

unread,
Oct 7, 2015, 10:20:51 AM10/7/15
to beagl...@googlegroups.com
On 07/10/15 17:15, Sven Bock wrote:
> Thank you so much for your support Roger.
> I assume with the stock rom the D14 should turn on by default.
> Could there be something set on the board, independent of the sd card?

I can't think of anything like that. The kernel should drive the D14 line
to the right state when the USB driver starts.

I just tested mainline kernel on my beagleboard-xm rev.B and it worked fine.

Which device tree blob are you loading with the kernel? For rev.C you need to
load omap3-beagle-xm.dtb and not omap3-beagle-xm-ab.dtb

cheers,
-roger
> **Cordialement */ **Best Regards / Mit freundlichen Grüßen / *
> Sven Bock M.Sc.
> *Research Engineer
> *
> *Tel: +33 556 393 705*
> *Génération Robots <http://www.generationrobots.com/> | HumaRobotics <http://www.humarobotics.com/>*
>
> --
> For more options, visit http://beagleboard.org/discuss
> ---
> You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com <mailto:beagleboard...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.

Sven Bock

unread,
Oct 7, 2015, 10:35:26 AM10/7/15
to beagl...@googlegroups.com
On Wed, Oct 7, 2015 at 4:20 PM, Roger Quadros <rog...@ti.com> wrote:
On 07/10/15 17:15, Sven Bock wrote:
> Thank you so much for your support Roger.
> I assume with the stock rom the D14 should turn on by default.
> Could there be something set on the board, independent of the sd card?

I can't think of anything like that. The kernel should drive the D14 line
to the right state when the USB driver starts.

I just tested mainline kernel on my beagleboard-xm rev.B and it worked fine.

Which device tree blob are you loading with the kernel? For rev.C you need to
load omap3-beagle-xm.dtb and not omap3-beagle-xm-ab.dtb

I used the command below for settting up the sd card as described on the wiki
sudo ./setup_sdcard.sh --mmc /dev/sdb --dtb omap3-beagle-xm 


You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/wWLfOolkR4E/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Robert Nelson

unread,
Oct 7, 2015, 10:42:20 AM10/7/15
to Beagle Board
On Wed, Oct 7, 2015 at 9:35 AM, Sven Bock <s...@generationrobots.com> wrote:
>
>
> On Wed, Oct 7, 2015 at 4:20 PM, Roger Quadros <rog...@ti.com> wrote:
>>
>> On 07/10/15 17:15, Sven Bock wrote:
>> > Thank you so much for your support Roger.
>> > I assume with the stock rom the D14 should turn on by default.
>> > Could there be something set on the board, independent of the sd card?
>>
>> I can't think of anything like that. The kernel should drive the D14 line
>> to the right state when the USB driver starts.
>>
>> I just tested mainline kernel on my beagleboard-xm rev.B and it worked
>> fine.
>>
>> Which device tree blob are you loading with the kernel? For rev.C you need
>> to
>> load omap3-beagle-xm.dtb and not omap3-beagle-xm-ab.dtb
>
>
> I used the command below for settting up the sd card as described on the
> wiki
> sudo ./setup_sdcard.sh --mmc /dev/sdb --dtb omap3-beagle-xm

This install's a version of u-boot that correctly detects the xmab vs xmc..

I use to carry a patch to fix ehci on v4.2.x, but this wasn't needed
anymore on my xmc..

What board do you have? and can you add the serial log with the u-boot version?

Regards,

--
Robert Nelson
https://rcn-ee.com/

Sven Bock

unread,
Oct 7, 2015, 10:54:11 AM10/7/15
to beagl...@googlegroups.com
I have the beagleboard xm rev C. I also attached you a picture of it. 

and can you add the serial log with the u-boot version?
where do i find this log? 

Regards,

--
Robert Nelson
https://rcn-ee.com/

IMG_20151007_164558.jpg

Robert Nelson

unread,
Oct 7, 2015, 11:02:19 AM10/7/15
to Beagle Board
> and can you add the serial log with the u-boot version?
> where do i find this log?

via the serial port.. But you have a C, same version i have here.. The
"early" xMA's had nand so some of the environmental variables can mess
with things..

Sven Bock

unread,
Oct 7, 2015, 11:13:22 AM10/7/15
to beagl...@googlegroups.com
Last hope: I use a 32 GB San disk ultra class 10 for the beagle. I don't see why, but could that be  a reason why the usb +ethernet ports are not working?

--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/wWLfOolkR4E/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Roger Quadros

unread,
Oct 8, 2015, 4:03:21 AM10/8/15
to beagl...@googlegroups.com
On 07/10/15 18:13, Sven Bock wrote:
> Last hope: I use a 32 GB San disk ultra class 10 for the beagle. I don't see why, but could that be a reason why the usb +ethernet ports are not working?

No, SD card type shouldn't affect USB.
Can you provide the u-boot log that Robert requested. You can take a capture of the serial port
right from power up via the serial terminal application.

We just need to ensure that u-boot is loading the right DTB.

cheers,
-roger

>
> On Wed, Oct 7, 2015 at 5:01 PM, Robert Nelson <robert...@gmail.com <mailto:robert...@gmail.com>> wrote:
>
> > and can you add the serial log with the u-boot version?
> > where do i find this log?
>
> via the serial port.. But you have a C, same version i have here.. The
> "early" xMA's had nand so some of the environmental variables can mess
> with things..
>
> Regards,
>
> --
> Robert Nelson
> https://rcn-ee.com/
>
> --
> For more options, visit http://beagleboard.org/discuss
> ---
> You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/wWLfOolkR4E/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com <mailto:beagleboard%2Bunsu...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.
>
>
>
>
> --
> **Cordialement */ **Best Regards / Mit freundlichen Grüßen / *
> Sven Bock M.Sc.
> *Research Engineer
> *
> *Tel: +33 556 393 705*
> *Génération Robots <http://www.generationrobots.com/> | HumaRobotics <http://www.humarobotics.com/>*
>
> --
> For more options, visit http://beagleboard.org/discuss
> ---
> You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com <mailto:beagleboard...@googlegroups.com>.

hawk...@gmail.com

unread,
Oct 11, 2015, 12:30:02 PM10/11/15
to BeagleBoard
I don want to hijack this string, just commenting that my BeagleBone Black is having what appears to be the exact same issue, or at least similar enough to be put in the same category.
I can access just fine using the mini USB port.
I cannot access via ethernet, it appears no IP address was set from DHCP server because the port seems active (lights all blinky and stuff).
I connected an FTDI serial cable and the connection is spotty but when i do (seem) to connect it doesnt recognize "root" use - all password attempts fail.

No LEDs lit except nearest power connection.

No SD card mounted (at the moment).

My MacBook Pro lives on my home wifi while the BBB is connected to a switch, connected directly to the router. I know the IPs are same subnet because my Drobo is also on that switch and I access that from my laptop continually.

I'm reading through your previous entries and will look for obvious similarities and address as suggested.

Thanks.

John

Robert Nelson

unread,
Oct 11, 2015, 4:57:36 PM10/11/15
to Beagle Board, hawk...@gmail.com
Hi John,

On Sun, Oct 11, 2015 at 8:21 AM, <hawk...@gmail.com> wrote:
> I don want to hijack this string, just commenting that my BeagleBone Black
> is having what appears to be the exact same issue, or at least similar
> enough to be put in the same category.

Nope sorry, completely different hardware, and a completely different
usb "PHY" ip...

> I can access just fine using the mini USB port.
> I cannot access via ethernet, it appears no IP address was set from DHCP
> server because the port seems active (lights all blinky and stuff).
> I connected an FTDI serial cable and the connection is spotty but when i do
> (seem) to connect it doesnt recognize "root" use - all password attempts
> fail.

The default images printout the user/password combination.. If it
doesn't contact who you go the image from..

>
> No LEDs lit except nearest power connection.
>
> No SD card mounted (at the moment).
>
> My MacBook Pro lives on my home wifi while the BBB is connected to a switch,
> connected directly to the router. I know the IPs are same subnet because my
> Drobo is also on that switch and I access that from my laptop continually.
>
> I'm reading through your previous entries and will look for obvious
> similarities and address as suggested.

Roger Quadros

unread,
Oct 12, 2015, 3:41:14 AM10/12/15
to beagl...@googlegroups.com, hawk...@gmail.com
Hi,

On 11/10/15 23:57, Robert Nelson wrote:
> Hi John,
>
> On Sun, Oct 11, 2015 at 8:21 AM, <hawk...@gmail.com> wrote:
>> I don want to hijack this string, just commenting that my BeagleBone Black
>> is having what appears to be the exact same issue, or at least similar
>> enough to be put in the same category.
>
> Nope sorry, completely different hardware, and a completely different
> usb "PHY" ip...

Actually, on bone black, ethernet is not over USB. It is directly over MII.

>
>> I can access just fine using the mini USB port.
>> I cannot access via ethernet, it appears no IP address was set from DHCP
>> server because the port seems active (lights all blinky and stuff).
>> I connected an FTDI serial cable and the connection is spotty but when i do
>> (seem) to connect it doesnt recognize "root" use - all password attempts
>> fail.
>
> The default images printout the user/password combination.. If it
> doesn't contact who you go the image from..
>
>>
>> No LEDs lit except nearest power connection.
>>
>> No SD card mounted (at the moment).
>>
>> My MacBook Pro lives on my home wifi while the BBB is connected to a switch,
>> connected directly to the router. I know the IPs are same subnet because my
>> Drobo is also on that switch and I access that from my laptop continually.
>>
>> I'm reading through your previous entries and will look for obvious
>> similarities and address as suggested.
>

cheers,
-roger

MHE 何海宇

unread,
Nov 17, 2017, 7:23:24 PM11/17/17
to BeagleBoard
Hi,
I experience exact same issue on the 3.14.20 linux kernel, u-boot detect ethernet port, load the omap3-beagle-xm.dtb after linux kernel loaded, the led is off, nothing never happens again nor usb, nor ethernet, seems the part is incompatible with the kernel.. what matters is the fb device not working too.. I suspect these things linked to same root cause.
Reply all
Reply to author
Forward
0 new messages