AM4377: Debug UART issue

Madhusoodana Bairy6

Part Number: AM4377
Other Parts Discussed in Thread: AM4379, TXS0102

Hi Team,

We are using AM4377 processor and FT232RQ USB-UART converter in one of our projects. We are facing some issue with this device and issue description is as below:
We are finding 0.6V to 0.9V on board's 3.3V power rail without powering on the board when we connect it to PC/Laptop. We have seen this in few boards (AM4377, AM4379 etc). So, some interfaces are getting affected by this as power ramp is not from 0 to 3.3V but from 0.9V to 3.3V. This is because there is 1V-3V on UART0 TX & RX lines (different in different boards). How voltage on UART0 TX RX is affecting 3.3V power rail? Is it connected through some channels internally ?

Best Regards,
Madhusoodana Bairy

over 7 years ago

0 Biser Gatchev-XID over 7 years ago

TI__Guru**** 393215 points

Hi,

Is your FT232RQ USB-UART converter powered from the USB port? If this is true, the most likely reason for your issue is that the UART_RX line is drive high from the FTDI while the AM437x is unpowered. This is a very dangerous situation, as power leaks through the device I/O pin to the power rail. This is strictly prohibited by specification. No voltage should be applied to AM437x I/O pins before all power supplies have fully ramped up and stabilized.

0 Madhusoodana Bairy6 over 7 years ago in reply to Biser Gatchev-XID

Intellectual 680 points

Hi Biser,

Thanks for your response.

What is meant by "power leaks through the device I/O pin to the power rail" ? Is there any internal pull ups to UART0 IOs and power is back-driven to these IOs ? (as IO lines are powered off)

Best Regards,

Madhusoodana Bairy

0 Biser Gatchev-XID over 7 years ago in reply to Madhusoodana Bairy6

TI__Guru**** 393215 points

It's not a case of pullups. These leaks occur through the internal reverse diodes of the CMOS output driver. These diodes are inherent to MOSFET transistors.

0 Madhusoodana Bairy6 over 7 years ago in reply to Biser Gatchev-XID

Intellectual 680 points

Hi Biser,

Thanks for your response. Anyway, we are planning to add one buffer (which will not be powered from PC/Laptop) between processor and FTDI chip, which should solve the issue.

Regards,
Madhusoodana Bairy

0 Biser Gatchev-XID over 7 years ago in reply to Madhusoodana Bairy6

TI__Guru**** 393215 points

That will be the best solution. TXS0102 has worked well for boards I have designed.

Processors

Processors forum

AM4377: Debug UART issue