USB1 HOST VBUS error workaround (delay coming) issue

zhirong wu

Other Parts Discussed in Thread: AM3354, AM3359

In my board with am3354, I found that the USB1 HOST is not work; The USB1_VBUS pin is fixed to connect with 5V , and USB1_ID is connected to ground directly for USB1 HOST configure;

when power on the board, the logs show USB1 work at HOST mode, and USB1_DRVVBUS will have wave from Low to high, about several millisecond, the USB1_DRVVBUS be Low, then show a log message "VBUS error workaround (delay coming)";

Because USB1_VBUS pin is power by 5v directly, so I think regardless the USB1_DRVVBUS status, VBUS pin should detect a high status(>4.4V); So where maybe I mistake and need to check ?

below is the log of usb1:

root@am335x-evm:~# cat /proc/driver/musb_hdrc.1
Status: MHDRC, Mode=Host (Power=e0, DevCtl=88)
OTG state: a_wait_vfall; inactive
Options: pio, otg (peripheral+host), [eps=16]
Peripheral address: 00
Root port status: 00000100

over 11 years ago

0 Biser Gatchev-XID over 11 years ago

TI__Guru**** 393215 points

Hi Wu,

I think the problem is that the USB module expects to detect voltage on USB1_VBUS pin after the USB1_DRVVBUS pin drives from Low to High level. If you have a VBUS power switch driven by USB1_DRVVBUS try taking the signal for USB1_VBUS from the connector side of the switch.

0 zhirong wu over 11 years ago in reply to Biser Gatchev-XID

Intellectual 270 points

HI Biser

The board have a VBUS switch chip driven by USB1_DRVVBUS, I disconnect the VBUS pin from the fixed 5v power and change it connect to this chip output, then power on board, the USB1_DRVVBUS will have a status changed from Low to High to enable the switch chip output 5V for VBUS, but the VBUS pin of USB1 module can not detect High status yet, the result same as above. This is my confusion.

0 Biser Gatchev-XID over 11 years ago in reply to zhirong wu

TI__Guru**** 393215 points

Strange, the USB1_DRVVBUS should turn on the VBUS switch after a device has been plugged in the USB connector. Maybe there is a software problem, like wrong driver configured? Sorry I can't help about software.

0 zhirong wu over 11 years ago in reply to Biser Gatchev-XID

Intellectual 270 points

Hi Biser

Code build from SDK release, and the same code run in the evm reference board, the usb1 host work fine; the difference between evm and my board is that evm board use am3359, so I think the code should be no problem.

0 zhirong wu over 11 years ago in reply to zhirong wu

Intellectual 270 points

I measure the DRVVBUS signal that will from Low to High, then to Low after 100ms;

The DRVVBUS must keep High, then the switch chip will output 5V; the VBUS pin of USB1 module can not detect the 5v input in 100ms now.

0 Biser Gatchev-XID over 11 years ago in reply to zhirong wu

TI__Guru**** 393215 points

Hi Wu,

The Support Team in Dallas have looked into this issue. Here is a quote from their e-mail:

"I think the 2-sec timer to start the session is in there, since the wave shows USB1_DRVVBUS goes from low to high, which means session started. Then the message ‘VBUS error workaround (delay coming)’ is telling there was VBUS error happened and then delay to retry the session. I have not seen this message happen so far in any case so not sure if the workaround (re-start session) does really work. It still looks like h/w issue because of the VBUS error."

Can you please check what is the voltage on VBUS, and possibly if the VBUS pin of USB1 is not soldered properly? By the way, on the EVM boards there is a 1.2kOhm serial reistor on VBUS going to the processor VBUS pin, but I don't think this is significant.

0 zhirong wu over 11 years ago in reply to zhirong wu

Intellectual 270 points

Maybe I have found the reason.

The 5V power is connected a resistance 1.2kohm and a voltage regulator diode to VBUS pin of USB1, the voltage regulator diode(5.1V) is between the VBUS pin and ground;when a 5V power via the 1.2kohm and a voltage regulator diode to VBUS pin, then I measure the VBUS pin , the voltage is only 0.7V! so the USB1 module DEVCTL register bit[4:3] is 0! Remove the diode, it is working fine!

And I rework the VBUS pin to fixed 5V power then boot up the board, it's working too!

0 Biser Gatchev-XID over 11 years ago in reply to zhirong wu

TI__Guru**** 393215 points

Great! Glad you solved this!

0 Biser Gatchev-XID over 11 years ago in reply to Biser Gatchev-XID

TI__Guru**** 393215 points

Hi Wu,

Did you check that the zener diode is not soldered backwards? This 0.7V level sounds like it was.

0 zhirong wu over 11 years ago in reply to Biser Gatchev-XID

Intellectual 270 points

Hi Biser:

My God! The zener diode chip is connected in reverse by SMT. After correct the diode, I measure the VBUS pin of USB1 is 4.61V, input voltage is 4.99V.So this is meet VBUS pin request.

Thank you for reminding me!

0 Shamrel Alexander over 11 years ago in reply to zhirong wu

Prodigy 170 points

Hello!
We use the VAR-SOM-AM33, when connecting the USB flash drive in USB0, a message appears:

"# [ 61.617431] ti81xx_interrupt 1145: VBUS error workaround (delay coming)"

The USB00_VBUS pin is fixed to connect with 5V. Lines USB0_ID and USB0_DRVVBUS on the module is not available. What to do?

0 Biser Gatchev-XID over 11 years ago in reply to Shamrel Alexander

TI__Guru**** 393215 points

This is a third party product. I don't have any documentation. Afraid I can't help.

Processors

Processors forum

USB1 HOST VBUS error workaround (delay coming) issue