DP83848K: Reading Register "BMSR" of DP83848K -basic Mode Status Register

Hi Frank,

Have you confirmed any other registers are working properly? Are you using the correct PHY ADDRESS? 

Thanks,

Cecilia

Hi Cecilia,

Thanks very much for your assistance, and we are using the proper address.  We have 2 devices at this address:

Address 0201h and 0221h

I wonder if this device requires a double reading as indicated on this link for a different part?  https://e2e.ti.com/support/interface/f/138/t/550756?Problems-reading-DP83822-register-0x0001-Basic-Mode-Status-Register

"Hi Antonio,

There is no issue with the DP83822.
The reason why you are having issue is because register 0x01 always requires two reads.
This bit in the register is a LL (latch-low). Thus two reads are required to get the correct real-time status.

Typically double reads are inate to the firmware. Please ensure your firmware does a double read.

Kind regards, Ross"

Thanks very much,

Frank

Hi Frank,

That is correct. The DP83848 also requires a double read since it is latch low.

More details on this e2e thread: https://e2e.ti.com/support/interface/f/138/t/654758

Thanks,

Cecilia

Hi Cecilia,

Thanks for your info, but still having issues.  I will be troubleshooting the hardware ckt and advise progress or issues later on.

Thanks again,

Frank

Frank,

Would you have access to an MSP launchpad or use our USB2MDIO tool? I am wondering if you'd be able to read and write to registers directly from MDIO/MDC pin to confirm it is not a PHY issue. 

Thanks,

Cecilia

Hi Cecilia,

Never used those tools and need to review them.  Thanks very much,

Frank

Cecilia,

Are there any pins on this IC that I need to preset (high/low) before reading the register ("BMSR"), other than setting the address pin?

Thanks,

Frank

Hi Frank, there are no other presets needed, you just need to be able to access the MDIO and MDC pin of the PHY and verify the correct address of your device based on the straps. 

Hi Cecilia,

As per my previous comment, we are using two devices on our product, and one of them is working fine as I am able to pull the right data.

The second device always gives "FFFF" even with the a double read, so I wonder if may be an interference between both of them or I may need to reset the second device before reading the register.

Device's addresses:

First device -Address 16:   PHY and REG Address: 00000201h - 00000000.00000000.00000010.00000001

Second device  -Address 17:   PHY and REG Address: 00000221h - 00000000.00000000.00000010.00100001

I am not able to follow the schematic. Can you show me exactly how you are connecting both devices to the MDIO? Is there a way you can isolate the second device and try just reading the registers of that device alone?

Thanks,

Cecilia

Hi Cecilia,

Please refer to attached pictures with more details for each device:

Thanks,

Frank

Hi Cecilia,

Just wanted to confirm if the "Clock IN (X1)" is required in order to read any of the registers, as shown below:

If is required, I am not sure why I am able to read the 1st device, but not the second??? As I don't have any clock signal available during boundary scan test.

During normal operation a FPGA device provides the reference clock (25MHz) to the DP83848K, but not during boundary scan test.

Please advise.  Thanks,

Frank

Hi Frank,

The clock is necessary for both PHY's for proper functionality, not specifically for MDIO/MDC access but for general use. Can you share the portion of the schematic where both MDC and MDIO meet together? Also, were you able to try out my theory of isolating one of the devices to only read one PHY?

Thanks,

Cecilia

Hi Cecilia,

A FPGA device drives both "MDC and MDIO" interface, as shown below:

The schematics provided previously shows the rest of the connections.

Thanks,

Frank

Frank - were you able to try out my theory of isolating one of the devices to only read one PHY?

Thanks,

Cecilia

Hi Cecilia,

No need to isolate, as I changed the code slightly and  is working now.  However, reading the resisters is not very repeatable.  Sometimes it reads "FFFF" and other time provides the real data.

Not sure why, but I am able to read the data by having a loop (reads the registers up to 10 times), until the register reads back any data other than "FFFF".

Thanks,

Frank

Hi Frank,

Can you probe the MDIO data for both lines as well as the MDC? I am wondering if there is a signal integrity issue. What was the code you changed that was able to make it work?

Thanks,

Cecilia