AM4377: USB2.0 Compliance Test

Hello Kenley,

What Processor SDK version do you use on AM4377?

Is the USB port under test in host or device mode?

Best Regards,

Borislav Lazarkov

Hello Kenley,

Please review this thread:
AM4378: USB Compliance Test Tools for AM4378 with Linux

Best Regards,

Borislav Lazarkov

Hi Borislav

Here is the SDK version. Sorry for the late reply.

PROCESSOR-SDK-LINUX-RT-AM437X

　Version: 09.03.05.02

　Version: 06.00.00.07

Regards,

Kenley

Hello Kenley,

Did you review the thread that I have sent to you and if yes did it help?

Best Regards,

Borislav Lazarkov

Hi Lazarkov,

Yes thank you.

Customer has additional questions.

Customer understand the materials for USB compliance testing.
If customer wants to perform a USB SI simulation, Is ibis available on ti.com compatible?
When customer looked at the contents, DM/DP was NC, and there was no model included.

If you have a dedicated ibis, can you provide it?

Regards,

Kenley

Hello Kenley,

Since the USB DP/DM pins use an integrated analog PHY cell (IOZ type), they are outside the scope of IBIS modeling. IBIS is designed for standard digital buffers, not analog/mixed-signal USB PHY I/O cells. This is why no IBIS model exists for the AM4377 USB DP/DM pins.

Best Regards,

Borislav Lazarkov

Hi Borislav 

Thank you for the reply.

For USB test mode, Customer changed the register in USB Test Mode on DWC3
# Devmem2 0x483D0420 w 0xA0
# Devmem2 0x483D0020 w 0x04
# Devmem2 0x483D0424 w 0x40000000
With the MW(memory write) command.
Cannot enter test mode because last register change is not possible.
The target port is USB1 in host mode, so it is the same as described in USB Test Mode on DWC3.
Unlike the AM437x EVM, to change the register at address 0x483D0424: Does this mean that other registers that are not written need to be changed too?

Best regards,

Kenley

Hello Kenley,

I have written to my colleague to support you on this topic.

Please expect a delay from his response for 1-2 days.

Thank you for your patience!

Best Regards,

Borislav Lazarkov

Hi Kenley,

It seems like the customer is attempting to run these register writes in U-boot. Can the customer try the same register writes but in Linux kernel with devmem2 instead of mw/md? Also, did they confirm the USB1 port is indeed in host mode? I am not sure if they have checked on the oscilloscope itself, but below is what the compliance packets should look like.

Best regards,

Luis Parga