DP83822HF: Transformerless operation

Hello Mr. Aniruddha,

I'm the customer for which Mr. Gupta addressed this question.

In the mean time we received the first modules with the new "DP83822" PHY and we are a realy surprised that the PHY (Tx) works in a reduced MLT-3 mode which actually prevents a transformer-less operation because we havend't a original differential MTL-3 signal here.

Is there a bootstrap option to change the MLT-3 mode? Is there a PHY from TI available which supports standart MLT-3 entcoding and sth fits in form, fit and function with the "DB83822"?

We are looking foreward to hear from you soon because we are talking about an replacment which means limited time.

Best regards
Christian

Hi Christian,

The 822 does not have this so called 'reduced MLT3' signaling.
Please explain what you mean by this.
The 822 is fully compliant to the 100BASE-TX standard.

Kind regards,
Ross

Hello Ross,

With reduced MLT-3 signal I mean the so called minimal current MLT-3 signal which is not a original differential signal with two complementary signals.

Best regrads and thank you in advance

Christian

Hi Christian,

I am sorry but i still do not understand what you are saying here.
MTL3 is a differential signal. The voltage swing of this signal should be between 950mV to 1050mV for positive peak and same for negative peak.
Most PHYs have options to change the gain of the driver, which can modify the voltage swing.
Can you explain what the issue is and what you are trying to do?

Thank you,
Ross

Hi Ross,

What I try to explain is that a original MLT-3 signal look like the picture below, Tx+ and Tx- are bothe MLT-3 encoded and just contrary negated:

And the PHY from TI is using two different encoded signal which in sum results in a MLT-3 signal but not both signals are particular MLT-3 encoded.

Best regards and thank you in advance

Christian S.

Hi Christian,

Thank you for the explanation. Now I understand what you are asking.
By setting register 0x404h to 0x4h instead of 0x20h you will see MLT-3 on both pins.

Kind regards,
Ross

Hi Ross,

perfect, so we are able to adjust the output coding:

Would you please be so friendly and forward me more information regarding this Register we changed because it isn't mentioned in the datasheet.

Moreover ... as I explaint you we need this PHY as a replacment on a existing design, this means we are not able to change the firmware of the target module. Would it be possible to change the default value of this register (0x404) at TI?

Best regards and thank you in advance

Christian Sielaff

Hello Ross,

could you please be so friendly and list us the TI PHYs wich works by default in the original MLT-3 mode?

In the mean time we also tested the "TLK106" PHY but also unsuccessfully even if it is recommended for transformer-less operation.

We are really sad about this because the datasheets are not pointing to this behavior and the description of the MLT-3 encoding are interpretable in both directions.

This could also be interpreted as a deliberate delusion.

But if also the application notes to be wrong, the reason could be also an unterstanding problem.

Best regards and thank you in advance

Christian Sielaff