Phy 2 Phy Connection From Orin Nano?

Robotics & Edge Computing Jetson Systems Jetson Orin Nano
, ,
1

We are currently developing a Ethernet-switch expansion board for the NVIDIA Jetson Orin Nano.
The device is intended to work by connecting an external base100t switch IC with internal PHYs (KSZ8895MLIX) to the ethernet PHY on the Jetson SoM.

As the Jetson has a gigabit Ethernet PHY inside, two of the pairs are left unconnected, so that both ethernet systems can auto-negotiate to communicate at base100t. Additionally, because the connection between the Switch IC and the Jetson is made on the same board, it was thought that capacitive coupling instead of magnetics should be used, to save cost and board space.

In regards to this we have several questions, we hope you can help answer.

1. Does the Nvidia Jetson Nano have internal termination on the ethernet PHY.

2. For the capacitive coupling, is the device able to be configured in capacitive coupling mode (series capacitors instead of magnetics).

3. Is it correctly understood that, for connecting a base100t PHY to the base1000t PHY on the Jetson, we should leave the unused ethernet diff pairs unconnected.

Ethernet-base100-base1000.drawio
Ethernet-base100-base1000.drawio1042×422 25.7 KB

2

*** Please note that this reply is generated by LLM automatically ***

I’ll do my best to provide a comprehensive answer to your questions.

1. Does the Nvidia Jetson Nano have internal termination on the ethernet PHY?

Unfortunately, the provided documentation does not explicitly mention whether the Nvidia Jetson Nano has internal termination on the ethernet PHY. However, it is common for many modern SoCs, including the Nvidia Jetson series, to have internal termination resistors on their PHYs. These resistors help to match the impedance of the PHY to the impedance of the external circuit, which can improve signal integrity and reduce electromagnetic interference (EMI).

To confirm whether the Nvidia Jetson Nano has internal termination on its ethernet PHY, I would recommend consulting the Nvidia Jetson Nano datasheet or technical reference manual. If you don’t have access to these documents, you can also try contacting Nvidia support or searching for similar questions on the Nvidia Jetson forums.

2. For the capacitive coupling, is the device able to be configured in capacitive coupling mode (series capacitors instead of magnetics)?

The documentation does not explicitly mention whether the Nvidia Jetson Nano supports capacitive coupling mode for its ethernet PHY. However, it is possible to use capacitive coupling instead of magnetics for ethernet connections, as you mentioned.

In general, capacitive coupling is a common technique used in high-speed digital circuits to couple signals between two circuits while blocking DC voltages. In the context of ethernet, capacitive coupling can be used to couple the differential signals between the PHY and the external circuit.

To determine whether the Nvidia Jetson Nano supports capacitive coupling mode, I would recommend consulting the Nvidia Jetson Nano datasheet or technical reference manual. You can also try searching for similar questions on the Nvidia Jetson forums or contacting Nvidia support.

3. Is it correctly understood that for connecting a base100t PHY to the base1000t PHY on the Jetson, we should leave the unused ethernet diff pairs unconnected?

Yes, that is correct. When connecting a base100t PHY to a base1000t PHY, it is common to leave the unused ethernet diff pairs unconnected. This is because the base100t PHY only uses two pairs of wires (pins 1-2 and 3-6) to transmit and receive data, while the base1000t PHY uses all four pairs of wires (pins 1-2, 3-6, 4-5, and 7-8).

By leaving the unused pairs unconnected, you can prevent any potential signal reflections or crosstalk between the two PHYs. This can help to improve signal integrity and reduce errors during data transmission.

In your case, since you are connecting a base100t switch IC to the base1000t PHY on the Jetson, it is correct to leave the unused ethernet diff pairs unconnected. This will allow the two PHYs to autonegotiate and communicate at the base100t speed.

*** Please give the thumbs up if you get answers you like. Or provide feedback to help us improve the answer. ***

3

This still leaves question 1 and 2 unanswered

6

1. Does the Nvidia Jetson Nano have internal termination on the ethernet PHY.

No, there is no termination between the PHY on the Jetson Orin Nano and the SO-DIMM pins.

2. For the capacitive coupling, is the device able to be configured in capacitive coupling mode (series capacitors instead of magnetics).

I am checking with the team on this one.

8

As per Jetson Orin NX Series and Jetson Orin Nano Series Tuning and Compliance Guide Application Note, the PHY on the Jetson Orin Nano is Realtek RTL8111HNI. Please contact Realtek for support on this particular question and if another driver would have to be used to be compatible with your setup. NVIDIA only directly supports the usecase with magnetics shown in the Jetson Orin NX Series and Orin Nano Series Design Guide.