Unrecognized USB Device Error

[sudheerg_]

Hi,

I am using Windows CE RTOS with DM3730 Cortex-A8 Processor.

It has a OTG Port. When we connect the DM3730 based hardware to desktop machine we are getting unrecognized USB Device error.

If we insert a catalyst protocol analyzer between the customzied hardware and PC to capture the issue the issue itself is not reproducible.

Can you let me know some debugging tips on how to proceed on this.

Thank you for your time.

Best Regards,
GSR

[Jan Axelson]

With the analyzer connected, the device works? Without it, the device isn't recognized every time?

See if the setupapilog file has a clue:

http://support.microsoft.com/kb/927521

[sudheerg_]

Hi Jan,

Thank You for the reply. Please do let me know if there are any links which talks for Windows XP host machine.

I want to clarify that, the device always works with protocol analyzer connected. But without protocol analyzer on some host machines this error is seen on some other machines the enumeration is success.

Thank You & Regards,
GSR

[Tsuneo]

DM3730 has just a ULPI interface for USB connection. Your board should mount a ULPI PHY chip.
a) Check the supply voltage, especially 1.8V, to this PHY chip. Isn't the voltage much higher than 1.8V?
or
b) check the value of RBIAS resistor (usually, 8k06, +/-1%).

A hardware protocol analyzer apply extra load to the D+/D- output buffer of the ULPI transceiver. It may reduce excess voltage (current) swing of the output buffers, caused by irregular voltage supply.

Tsuneo

[sudheerg_]

Hi Tsuneo,

Thank You for the reply and your time.

We will look into the suggestions given by you.

Some more information that we have as of now is, we are using a external chip which takes ULPI and provides USB Signals.

When we connect to host PC the device detects a High Speed Bus Reset Signal. It is able to read the first Setup packet (Get Device Descriptor: 0x80 0x06 0x0100 0x00 0x40). In response to this the DM3730 firmware writes a 18-Byte Device Descriptor to EP0 FIFO and sets appropriate bits to transfer the data.

But after 674mS the host again sends the Setup packet which is given above. The device firmware again reads it writes into FIFO the 18-bytes response and sets the required register bits to transmit it.

Again after 675mS the sequence of events happens which are given above.

After this no activity is observed from device firmware.

Do let me know if you had observed this behavior and if you have any more debugging steps to narrow down the issue.

Once again thank you for your valuable time.

Best Regards,
GSR