Question Can USB devices interfere with ethernet on AMD motherboard ?

[FAhentai]

Hi guys, on sep. last year, I post a thread here, on that thread, I mentioned that with my Asrock x670 RS PRO, I had problem with on board 2.5G NIC, once I plug in the usb sockets that next to RJ45 socket, the NIC will either down to 100M or not working and can not be found on the device manager. When I unplug those USB sockets, everything back to normal.
At the beginning, I thought it was just Asrock's problem. It might be just a Q.C. issue. Yet now I found that might not be the case.

3 days ago, one of my friend in China asked me why her NIC can not work. I asked her to check everything, from the ethernet cable to the router, eventually I found out that her 2.5G NIC can not be found on the device manager. At very first, I thought her on board NIC is broken down, yet based on my expirence, this kind of problem is fairly rare so I still had my doubts, until she showed me the back of her computer,

QQ-20240215055550 hosted at ImgBB

Image QQ-20240215055550 hosted in ImgBB

ibb.co

As you can see here, the USB socket next to RJ45 was plugged in, and a USB hub was connected on to it with another side keyboard, wireless mouse adopter and a usb flash disk plugged in. This straight away triggered my alert. I asked her to unplug the usb socket next to RJ45 and plug into some where else, guess what? The problem is gone! I was suprised, since she uses MSI MAG B550 Mortar, while ASrock X670E RS PRO is X670. 2 different brand, 2 different platform(am4 vs am5), same problem? It just does not make any sense.

Lastly, I start to think, as far as I and many of you know, in fact, B550 and X670/B650 are almost the same, they all use PROM21 as the motherboard chip (okay, I was about to say the nth/sth bridge yet there is no such thing nowadays isn't it ). Thus, I guess I found the answer. However, I don't know any of you have the same problem or not and till now I still fail to understand why my friend and I would have such strange problem. If any one have the answer, please let me know.

 

[JohnBonhamsGhost]

If you had a cheap enough manufactured board that those type of sockets were sharing power outputs or communication relays then yes, they could degrade each other's max output range.

If that was the case, you'd just have to upgrade to a better option.

 

[FAhentai]

If you had a cheap enough manufactured board that those type of sockets were sharing power outputs or communication relays then yes, they could degrade each other's max output range.

If that was the case, you'd just have to upgrade to a better option.

In fact, neither of them are cheap, far from cheap, Asrock X670E RS PRO is something high end one, MSI MAG B550 Mortar isn't cheap either.
Now I doubt this problem could be commmon for B550, x570, B650 and X670

 

[JohnBonhamsGhost]

neither of them are cheap

the price you paid has nothing to do with the quality of the manufacturing.

cheap enough manufactured

 

[FAhentai]

the price you paid has nothing to do with the quality of the manufacturing.

Humm, still, neither of them should be "cheaply manufactured".
In fact, both of them are some how popular, Asrock X670 RS PRO is almost the only pratical AM5 motherboard that come with 5 M.2 socket, while MSI MAG B550 MORTAR is also very popular back to AM4 time.
I don't know if it is just coincidence, that both of my friend's MB and mine had the same QC issue that leads to the same problem, or it is very common for both of those 2 MB

 

[LinuxDevice]

If the issue is due to power consumed (USB3 can produce significant power for devices, more than USB2), then she can use a powered USB HUB to that socket and the issue would go away (the HUB would provide device power instead of the USB socket from the motherboard). It might also be inadequate design of the motherboard whereby there is cross-talk noise appearing on the 2.5Gb/s network socket when the USB3 is running (digital signals make great noise generators). Have her try a USB3 HUB which has its own power supply to start with.

 

[FAhentai]

If the issue is due to power consumed (USB3 can produce significant power for devices, more than USB2), then she can use a powered USB HUB to that socket and the issue would go away (the HUB would provide device power instead of the USB socket from the motherboard). It might also be inadequate design of the motherboard whereby there is cross-talk noise appearing on the 2.5Gb/s network socket when the USB3 is running (digital signals make great noise generators). Have her try a USB3 HUB which has its own power supply to start with.

The hub she was using at the time is powered and when I was using my old Asrock X670E Rs pro, I plugged in one mouse and one keyboard to the usb port next to RJ45. Same problem.

 

[LinuxDevice]

The hub she was using at the time is powered and when I was using my old Asrock X670E Rs pro, I plugged in one mouse and one keyboard to the usb port next to RJ45. Same problem.

It seems like there is a defect in the motherboard in that case. It could be an individual board defect, or a design defect, or incorrect firmware (changing firmware essentially changes what the hardware is). I would bet against a design defect since so many of those mobos all being defective is unlikely. Is there a BIOS update available? If that doesn't fix it, then I suspect the hardware is bad for some unknown reason that won't be traceable.

 

[FAhentai]

It seems like there is a defect in the motherboard in that case. It could be an individual board defect, or a design defect, or incorrect firmware (changing firmware essentially changes what the hardware is). I would bet against a design defect since so many of those mobos all being defective is unlikely. Is there a BIOS update available? If that doesn't fix it, then I suspect the hardware is bad for some unknown reason that won't be traceable.

Sorry for late replying. I am very sure the BIOS is updated.
Now here I guess I found another noticeble detail, I have confirmed with her that the hub she was using is USB3.0, while the deveice connected onto this including one usb 3.0 flash drive, wireless mouse and keyboard , while back then I also connect mouse and key board to those ports, need less to say they are all usb2.0.
So, it seems like the USB2.0 device can cause that problem?

 

[LinuxDevice]

Sorry for late replying. I am very sure the BIOS is updated.
Now here I guess I found another noticeble detail, I have confirmed with her that the hub she was using is USB3.0, while the deveice connected onto this including one usb 3.0 flash drive, wireless mouse and keyboard , while back then I also connect mouse and key board to those ports, need less to say they are all usb2.0.
So, it seems like the USB2.0 device can cause that problem?

I just want to verify: Mouse and keyboard fail, but USB3 devices work?

This would most likely be a firmware error. However, it could be broken or bent pins in the socket. I will explain some of how this works.

Before USB3 came along each new USB standard (higher speed) got integrated into a single chip. There was initially USB 1.0. Then, when 1.1 came out, a single chip could switch between those modes. Same wiring. When USB2 came out it was the same story: One chip which could control all three, using the same wiring.

USB3 is a separate controller from legacy USB2 and older. USB3 also adds more wires in the socket. When USB2 is plugged into the socket different wires connect, and firmware setup will determine routing of wires to controllers (USB3 is routed to one controller, USB2 is routed to a different controller). So what you are finding suggests that either the wiring, or the programming for switching routes to legacy controllers (which is firmware controlled), has failed. Bent pins do this.

Lane routing for USB3 has succeeded, lane routing for USB2 (or more likely 1.1, but it is the same wiring as 2.0) has failed. Admittedly, this could be broken hardware, e.g., a scratch might have broken a trace, or the legacy controller might have failed, but this is far more unusual than software failure or bent pins.

You might want to examine USB 2.0 pictures to see which pins are used. USB 3.0 includes those same pins, plus others. It is the original USB 2.0 pinout that matters. Examples seen here:
https://en.wikipedia.org/wiki/USB