WARNING: internal 10|100 PHY requires magjack hardware!

[t4_4t]

Below is my recognition of terminating resistance.

The resistance of this part is the part which must strictly follow the standard of "49.9Ω +- 1%".
The value of the resistance directly connects to the quality of the communication signal.
Originally, these parts is a part where the value should not be changed for convenience of the designer.

If do not comply with this rule, please see below for what disadvantages occur.
Impedance matching reflection:
https://www.google.com/search?biw=1272&b...reflection

For example:
https://electronics.stackexchange.com/qu...abit-ether
With such a communication waveform, the distortion further increases.

You need to be carefully that the part with such characteristics has been changed to "62Ω".

---

Parts I used "UT7500L-18-T92-B-K" do not seem to be selling for overseas.

There is no problem substituting with "MCP1702-1802E/TO".

For capacitors, first add a larger value, it is intention to get a margin.
For example: 10uF + 1uF(ceramic).

After that,
I think that it is a good approach to reduce while confirming the possibility of network operation.
First of all, it is important to confirm that it operates correctly.
If it does not work, everything you do is wasteful.

Perhaps, ultimately it will be okay only with 1uF(ceramic) on the output side.


---

> I have zero experience in soldering on SMD circuits...

The tools I use are not special.
It is a standard soldering iron. And I do not have "soldering station".
https://www.amazon.com/Hakko-temperature...B006MQD7M4

The solder material of the board is "lead-free", so it would be better to have flux.
For example:
https://www.amazon.com/HOZAN-Flux-H-722/...older+flux


If you are inexperienced, you can not go on suddenly, practice is necessary.
However, its difficulty is:
 Anyone can do it by practicing about 1 hour using the junk board.
It's the level of difficulty above.

I will present a photo of the relevant part later.
Please wait for a while.

[t4_4t]

The part surrounded by the red line is "1.8 v + [L (coil) x 4]".
From green line ("VCC_8 Line"), find the place where you can easily to attach.
Then draw out with wiring and connect to the four terminating resistors.

The wiring I actually did is the same as the photo I posted.

The original photo is as follows:
https://www.cnx-software.com/2017/06/08/...d-usb-3-0/
Since the "Version (1.1)" of the board is different, there are slightly different parts in the periphery
However, the layout of the points to be noted is the same.

There is a possibility that I will make a mistake.
Therefore, you need to confirm that it is correct on your own before you do the work.

---

The signal we are handling with is 125MHz.
As you can see from my example:
 Just, Only moved the terminating resistor from "RJ45" position to "ROCK64 Board".
 But it turned into can not work.

The signal we are handling with is of such kind.
Therefore, Even if you can not make it work,
I will not be able to give you more useful advice.

Please do not forget that.


Good luck.

[WrongWorld]

Thanks a lot for all the info, I think they will be very useful to anybody wanting to add the second Ethernet port to Rock64.

I have followed another approach: since I was not able to find locally any 1.8V voltage regulator, I resorted to an LM317L for which I calculated the needed resistors for outputting the needed 1.8V: once I will have the circuit properly working, I will be back to this at a later time, publishing also the relevant schema.

My problem is now in enabling the eth1 device. I am currently using DietPi 6.19.7 with ayufan kernel 4.4.132-1075-rockchip-ayufan-ga83beded8524. I have tried the command

Code:

sudo enable_dtoverlay eth1 ethernet@ff550000 okay

but I obtain invariably the following output, and the eth1 port is not made available:

Code:

Applying...
/dts-v1/;
/ {
       fragment@0 {
               target-path = "/ethernet@ff550000";
               __overlay__ {
                       status="okay";
               };
       };
};
Warning (unit_address_vs_reg): Node /fragment@0 has a unit name, but no reg property
unapplied <<---- !!!

I tried to do the same with another device, for instance i2c0, and it works perfectly. Am I using the wrong kernel??? What am I missing?