DP83TC811R-Q1: USB to BRR adapter, network cable not connected

Hi Michael,

Are you able to read/write registers? Could you also kindly provide the schematic?

Thank you,

Nikhil

Hi Nikhil,

unfortunately i can only provide you extracts of the circuit diagram.

Since the IC should only be adjusted by configured pins, we have not read/written any registers yet. I will take care to do this.

Best regards,

Michael

BRR_schematic.pdf

Hi Michael,

Based on the schematic it looks like the AC coupling cap is DNP'd and the magnetics network is placed. Also the common mode termination network is DNP'd, can you confirm? The datasheet recommends the use of a CMC with a common mode termination network as shown in Section 9 of the datasheet. 

Thank you,

Nikhil

Hi Nikhil,

Yes, the AC coupling and termination aren't mounted, because we wanted to get the isolation according to the IEC 62368.

We already tested with the common placement of the AC coupling with termination, but we will test it again with the "bad" devices.

Thank you for this advice. I will give feedback, if this will change anything.

Best regards,

Michael

Hi Nikhil,

No difference between these different ways of termination in this case.

But  noticed the following behaviour:

PC detects USB to Ethernet adapter, no cable connected.

I held the DP83 in reset and the USB to MII MAC signalled the PC, that a cable is detected. After releasing the DP83 from reset, the cable was again "unplugged". The behaviour is reproducible.

Best regards,

Michael

Hi Michael,

For the remainder of the debug, can I ask to keep the AC capacitor and termination circuit setup? This is what is known by the datasheet.

Does the link partner have any internal pulls that could be affecting the strap settings of the device? 

Can I ask again if you are able to read/write registers? If so could you provide a read of registers 0, 1, and extended register 467?

Thank you,

Nikhil

Hello Nikhil,

Here the registers from the working devices:

Master:

Slave:

I can't read out the registers of the not working devices.

Best regards,

Michael

Hi Michael,

For the devices that aren't working, are you able to observe the clock on RX_CLK or the PAM3 waveforms?

Thank you,

Nikhil

Hello Nikhil,

Attached, a screenshot from the RX_CLK:

Best regards,

Michael

Hi Michael,

Thank you for your patience. The fact that the clock is running but the MDIO/MDC is not accessible points to the possibility that the device is in a sleep state. It's possible the 4.7k pull-up on the Wake pin is too weak a pull-up and the PHY is not awake. Can you replace this with a 2.49k resistor?

Thank you,

Nikhil

Hi Nikhil,

I changed the pull up resistors on Wake pin to 2.4k resistors, but it has no effect.

Best regards,

Michael

Hi Michael,

Of the not-working devices, are they programmed for master or slave? If they are set for master, are you able to observe FLPs upon power-up?

Thank you,

Nikhil

Hello Nikhil,

no, i can't see FLP Pulses in the power-up sequence at the not-working devices.

I can change the Master/Slave Mode with a slider. When a device isn't working in Slave mode, it also doesn't work in master mode.

Best reagrds,

Michael

Hi Michael,

After further review, it looks like in the schematic, the LED strap resistors do not have parallel pulls as shown in section 8.5.2 of the datasheet. This may cause an incorrect voltage strapping for the LED strap modes. As a debug step, can we remove the 0 ohm resistors on the LED pins such that the LED pins are left floating? It's possible that LED_1 has been incorrectly strapped into managed mode. 

Thank you,

Nikhil

Hello Nikhil,

I tried it, but i couldn't see a differnet behaviour.

Best regards,

Michael

Hi Michael,

To summarize what we're seeing: Of the "not-working" devices

1. They do not work in either master or slave

2. There are no FLPs when in master mode

3. There is no register access

4. RX_CLK is still active 

Can you kindly confirm that this is true when the PHY is isolated from the MAC and link partner?

Thank you,

Nikhil

Hi Nikhil,

Yes, i tested all these points 1-4 again, same result.

I disconnected the Rx and Tx signals between PHY and MAC, but i can't see any difference.

No FLPs in master mode, no register access, Rx_CLK is active, no Link to partner is detected.

best regards,

Michael

Hi Michael,

Are you able to probe for FLPs closer to the PHY, such as on the AC coupling caps?

What is the clock source? Can you provide a part number?

Can you probe LED0, LED1, and RX_D3 to confirm the voltage on these pins?

Thank you,

Nikhil

Hi Nikhil,

No, i couldn't see FLPs there.

Clock source is: O_25,0-JO32-B-1V3-1-T1-LF

LED0, LED1 and RX_D3 are on 0V(GND).

Best regards,

Michael

Hi Michael,

We are looking into this issue and will provide feedback by Tuesday next week.

Thank you,

Nikhil

Hi Michael,

Can you please confirm/try the following?

1. Probe the inhibit pin, is this high or low?

2. Is MDIO pull-up 2.2k?

3. Replace RX_D0 pull-up with 2.49k

Thank you,

Nikhil