DP83869HM: PHY & SFP Link Not Detecting

HI Team,

If possible, could customer read the following registers to confirm on the strapping and mode of operation?

• Register  0x0000 to 0x001F
• Register 0x0044
• Register 0x006E

If possible, could customer try to write following script and see if that help with the performance?

If customer remove the fiber module, could customer see a link up?

--

Regards,

Hillman Lin

Hi Lin,

We can try reading these registers, check and configure the registers.

Since we have designed it as an unmanaged switch so configuration of registers everytime won't be a feasible option for us.Do you see any other alternatives?

We have plugged 1g module on both sfp & Ethernet ports simultaneously and at a time but couldnot see any link detection at any moment of time.

Hi Saha,

It seems like a Cooper link issue.

I want to check if the PHY is strap correctly. Could I get more information on the strapping?

If we don't have the ability to access the register, We can take a look on schematic for further debug.

--

Regards,

Hillman Lin

Hi Lin,

I have shared schematics in the below link for strapping configuration. Please have a look (first message)

https://e2e.ti.com/support/interface-group/interface/f/interface-forum/1368009/dp83869hm-schematics-review/5231785#5231785

The only change we have used integrated magnetics(7499110124) instead of discrete magnetics

We have measured the strapping voltages, I will share that data tomorrow.

OPMODE lines are stables after all the Vrails are stable till the POR reset release.

SMI interface debugging as you said, we will try tomorrow and update you.

Hi Saha,

Thank you for the feedback, I will wait for the SMI interface information for further debug.

--

Regards,

Hillman Lin

Hi Lin, 

We will take time to bring up SMI interface as we are missing Msp430/432 bare metal driver codes. We are developing code for that. (If you have any thing on this pls share).

Meanwhile if you could help with debug steps with hardware bootstraps, we can check on that 

Hi Saha,

Here is the firmware we use for register access:

• 8105.msp430-usb-to-mdio-universal-v3.04.txt

Regarding to the strapping, I would like to see the voltage of LED during power up. If possible, could you take a probe measurement on the LED pin when it power up?

--

Thank you,

Hillman Lin

Hi Lin,

Here are the measured voltages with the configuration set:

OPMODE:100

I am having a doubt, LED straps are configured in Mode 0. So, should I read these pins as logic 1 or 0? The datasheet states "If a given strap input is resistively pulled low then the corresponding output is configured as an active high driver." 

Hi Saha,

Thank you for sharing the information

The initial power up for 50ms you should always read 0V on the LED pin. That is the indication on which mode is your LED configure in.

I will wait for your registers information.

--

Regards,

Hillman Lin

Hi Lin ,

Do you see any discrepancy in strappings read values?

Hi Saha,

The strapping looks fine to me if initial power up for 50ms are always read 0V on the LED pin.

Did you have the scope capture on three LED pins during power up?

--

Regards,

Hillman Lin

Hi Lin,

I will check and share your tomorrow.

In datasheet for LED mode 0 we don't see any parallel PD of 2.49k. whereas in evm there is a 2.49k in parallel to led mode 0 config. is this 2.49k really needed?

Reg. SMI captures our SW team is working on it, I will update you soon.

Hi Saha,

You are correct. LED 0 is not required to have 2.49k ohms pull down. It has internal pull down resistor within the PHY.

--

Regards,

Hillman Lin

Hi Lin,

A quick update.

Our card is up in Media-Converter Mode.

Few changes I want to discuss:
1. For LED Mode 0 strapping the 2.49k is needed as the default config in Media converter mode for the LED pins is Hi-Z.
2. Mirror Configuration needs to be disabled if port A of PHY is connected to port A of the RJ45 connector.
3. Pass-Thru Mode can be enabled or disabled as per application. We tried both it is working fine.

We have tried with SFP-Copper Modules for our testing. We will try with optical modules. 

Hi Saha,

Any update on the register access?

--
Regards,

Hillman Lin

Hi Lin,

We are using Msp432 evb for detecting the phy using the usb,-mdio.universal GUi, but we are not able to.detect phy

Hi Saha,

Are you using the USB2MDIO GUI?

--

Regards,

Hillman Lin

Hi Lin, 
As this is an unmanaged switch & it is working fine with HW strapped mode. We have dropped the configuration using MSP430.

Hi Saha,

So you are able to access register now?

--

Regards,

Hillman Lin

No Lin,
We are not accessing registers anymore. We proceeded with HW strapped for this design.

Hi Saha,

We have limited option to debug since we don't have any register to understanding the status of the PHY currently is in.

The main hypothesis that we have on our side is PHY not strap to correct mode which result in 1000Base-X not working properly. With limited resource to debug, please try to following test:

• Make sure there are no SoC signal sending to the PHY during power up
• Measure the voltage of of Pin 22, Pin 35 and Pin 36 during power up. 

--

Regards,

Hillman Lin

Sure, Lin. we will take care of this. The card is up now