DP83TD510E-EVM: Starting TDR gives me TDR_Fail

Hi Dominik,

Looks like you are writing the correct values to the correct registers. Can you please make sure that you have the "Extended Registers" tab set to "Yes" when you run the initial script?

Otherwise, how are you powering the EVM? With the USB power or external?

Thanks,

Lysny

Hello Lysny,
Thank you for your reply. When running the initialization script, the "Extended Registers" tab is set to "Yes". The EVM is powered via USB.

Hi Dominik, has resistor R24 been replaced with a 0 Ohm resistor?

Also, is the device able to link? Are there any issues writing to the registers? Are you able to read back some of the TDR program registers and verify that they are being written correctly?

Thanks,

Lysny

Hi Lysny, yes R24 is shorted and the device is able to connect. Reading back the TDR registers gives me the correct values.

Hi Dominik,

I'll look into this and get back to you in a couple of a days. In the meantime, if you have an external MSP430, you could try to connect directly to the MDIO and MDC pins on the EVM  and communicate with USB-2-MDIO and try the same process to see if there are any different results.

Thanks,

Lysny

Hi Dominik,

I have a few follow up questions. For the "not actively connected to a link partner," I wanted to make sure it is not physically connected to the partner device. If it is and that device is powered and not set to halt idle transmissions, this could cause TDR failure.

Second, could you please provide what type of cable you are using and the length?

Also, could you verify the version of silicon you have on board? This can be checked by sharing register 0x3.

Here is an alternative TDR script that has proved successful. Please try this script out as well. :)

For PHY addr = 00 & Extended Registers = "Yes"

001F 8000 // hard reset

Extended Registers = "No"

begin

000D 0007 //set mmd

000E 0200 // set addr

000D 4007

000E 0000 //data -> disable ANEG

000D 0001 //set mmd

000E 0834 // set addr

000D 4001 

000E 4000 //data -> config Host mode

end

Extended Registers = "Yes"

0301 2403 //tdr

0303 043E //tdr

030E 2520 //tdr

001F 4000 //soft reset

Then complete the same procedure above with PHY addr = "01"

And, enable the TDR like you did previously.

Thanks,

Lysny

Hi Lysny,
the device is not physically connected to a partner. For evaluation purposes i am using a 5m unshielded twisted pair cable, with an open end.
Register 0x3 contains 0x0181.

Thank you for the provided script. After running it, register 0x001E gives me 0x0002, which corresponds to TDR_Fail =1b. The TDR_Done flag is set false. Strangely enough, after executing your script and the TDR start, the register 0x030C contained a non-zero value. For a 5m long cable 0x030C equals 0x0C19. This probably means that a measurement has taken place although TDR_Done is false. How is that possible?

Lysny Woodahl said:

Then complete the same procedure above with PHY addr = "01"

if I am not mistaken, the DP83TD510E-EVM does only have a PHY at addres "00" and "05" so i did the same procedure with PHY addr= "05".

Maybe you can help me clarify.

Thanks,
Dominik

Hi Dominik,

You're right. The 822 is address 05! 

Could you try a few different things, and I'll keep looking into this?

Could you try running the TDR without a cable connected and let me know the results in 0x001e and 0x30c?

If you have another length of cable or type of cable, try the TDR process and report the same registers?

Also, how long are you approximately waiting after starting TDR to read the TDR result values? If the reads are in the same script as the writes the device might not be actually completing the TDR process. 

Also, before running any scripts could you share the value of 0x1e? And, could you also share the value after running the scripts, but before writing TDR_START?

Thank you!

Lysny

Hello Lysny,

the register values, before and after running the scripts you provided, in 0x001E and 0x030C are both 0x0000. I have tried TDR_Start without cable or with a 5m and a 10m cable. After every cable change, i did a full reset. No matter what cable length, 0x001E returns me 0x0002 and 0x030C is always equal to 0x0C19. I don't use a script to read the values from the registers, so I wait a few seconds before reading the registers.

Thank you,

Dominik

Hello Lysny,

Thanks for your answer, now the register values make sense to me!

Lysny Woodahl said:

This is due to the accuracy specs of TDR. 

Where can I find the accuracy specification of TDR, in terms of minimum cable length. I have not found anything in the data sheet. Also, can you let me know when the revised version of the application note has been released, that would be great?

Greetings,
Dominik

Hi Dominik,

The latest app note has been submitted and will be live with the next TI.com update.

The TDR accuracy varies slightly from cable to cable and has different accuracy specs at various lengths. This is the primary reason this info is not listed in the datasheet. Testing is required to properly calibrate TDR, so we encourage users to determine the accuracy based off their measurements.

At distances below ~25-30 m, TDR usually displays that distance. If you have any further questions, please open a new thread. :) This one is marked resolved already.

Thanks,

Lysny