9 - USB3.0 Gigabit LAN with RTL8153

This project has been completed, and the latest version has been submitted to the Hard Creation Society
. Friends with conditions can support it.
https://x.jlc.com/platform/detail/c80f0b348edc45d4872dcdcd6f31eca0
The new version uses EA3036 to provide 1.2V and 3.3V, so the temperature problem is solved.
 
First of all, this project is closely related to Link's RTL8153.
The RTL8153-VB under his project is 40PIN. I went around and found it expensive and difficult to buy.
I bought a few wrong ones on Xianyu, which are 48PIN (RTL8153-CG), I was too cheap to read it carefully at that time
, then I combined Link's engineering and official data books, and tried to verify the 48PIN version of the PCB.
 
At first, after soldering, USB3 was not recognized.
I thought the capacitor (CTX) was placed in the wrong line, and then I accidentally connected a USB2 and found that it worked. Then I tried to change CTX and found that USB3 was recognized. It seems that the soldering is the problem.
In addition, there is a flying wire on the cover because I reversed the setting of the chip about the external clock source frequency. It should be pulled up to match the 24m capacitor externally, and pulled down to input the 48m clock externally. I pulled it down
by default. I originally wanted to use 24m. This error has been fixed in the published schematics and PCB, and other layouts have not been changed.
 
The 1V of RTL8153 itself has two output options, one is to choose the DC-DC that comes with the chip,
which requires an external inductor, and the other is the LDO mode output. The two cannot operate at the same time. Then I chose LDO, and then according to the measurement, I found that the chip power consumption is as high as 1W (in USB2 mode, USB3 mode may be higher).
This means that, and I actually felt it, the chip started to get hot after a while. (Of course, I don’t rule out my design problem. After all, Link’s RTL8153 is also in LDO mode.)
So I regretted using the LDO mode. The DC-DC noise is a little louder, anyway, it’s the only solution on the board.
The main reason is that the data book also explained it very clearly at that time. DC-DC generates less heat and LDO has less noise.
 
Currently, this page is only open, and I only drew the LDO version of RTL8153.
In addition to some minor shortcomings of LDO mentioned above, there is also an actLED that I did not correctly connect to the LED of the RJ45 socket, but placed it on an external LED. However, this does not seem to be a big deal.
 
 
The next step is the actual measurement. USB2 can run at 35M per second, and USB3 can run at 100M per second. The standards for gigabit speeds
 
are USB2 and USB3 respectively .
 
The actual measurement part is mainly in the video.
That's about all I can say. See the attachment for the data book.