Question Crazy theory about NAS, fans, cables, switches, and NICs, and of course, our favorite, "static charge".

[VirtualLarry]

How prevalent are static charge issues with networking gear? I noticed that my "pro" managed D-Link switch (which is also rack-mountable), has a grounding screw. Should I connect that to anything? The power plug is a 3-prong, and should be grounded already.

Also, I've got my NAS units on a shelf, with 80-92-120mm exhaust fans. They are blowing directly on the cables hanging down from the switch, sitting on top of them.

I have problems with my NICs on my main PC, which is not too far away. I've lost the onboard NIC, a 10GbE-T Asus PCI-E x4 NIC (both not detected at all in Device Manager), and now I'm having some dropout/detection issues with a new fresh CableMatters RealTek-chipset USB3.1 to 2.5GbE-T, which is wired by a 30' Cat6 cable, to a 2.5GbE-T port on my D-Link switch.

Here's my crazy theory. They say not to use a vacuum cleaner (which moves air) to clean PCs. Why? Because, Static Charge.

If moving air can generate a static charge in things, especially around plastics, am I somewhat un-intentionally generating a static charge on my networking cables, going from my NAS units, to my switches, because the cables are plastic-sheathed, and in the path of the moving exhaust air from the NAS units?

How would I tell?

Should I switch to all shielded twisted-pair Cat7 where possible?

Edit: I've heard tell, from Cable Techs, that sometimes cable lines build up static, and screw up modems, and you need to disconnect the coax cable, and touch the end (center conductor), possibly to the shield conductor, using your finger (high-impedance). And that action drains the static charge that accumulates.

Edit: Is it possible that this could be a "real issue"?

I've had to reboot my (managed) D-Link switch a few times, as a newly-connected cable wouldn't pull an IP. I've got a Netgear 8x 1GbE-T and 2x 10GbE-T (unmanaged), and a Netgear 8x 1GbE-T (cheap consumer switch), and the connection back to my router (and thus DHCP server) is through that consumer Netgear switch.

 

[SamirD]

I think the worries are unfounded unless you have different grounds for the devices, in which in case in theory there could be voltage across the ground that can damage things. But this usually happens between buildings that have different grounds that are connected with copper of some sort--not usually a problem at all in the home since it's on a single ground.

Now, as far as your pc eating nics, there is something suspect going on there as nics shouldn't die like that, especially as many types as you've gone through--do you have a different way to wire it into place? Like a different switch?

The other experiment I would try is to connect a wired nic in the system to a wired/wireless 'gaming adapter' or something that allows your wired nic to connect to a wifi network. Technically it wouldn't be wired anymore except to the adapter and if it didn't burn out, I think it would help reinforce the theory that something is going on when it is directly wired.

 

[VirtualLarry]

Well, my shiny new CableMatters RealTek 2.5GbE 8156-chipset NIC is now flaking out. It was on the old USB powered hub, so I disconnected that one from the USB extension cable, and connected a small 4-port passive hub (Wavlink) with power switches / buttons for each port. I then plugged in the CM 2.5GbE USB3 dongle, and a Comfast AC+BT USB2.0 wifi dongle.

It was good for a little while, then, Bong, bing, bong, bing, just keeps disconnecting, reconnecting, some sort of USB error device shows up.

Either my PC is killing USB devices, or something is killing the NICs, or at least damaging them.

Either that, or every single RealTek-chipset NIC that I've installed thus far, had Chinese malware in the drivers, and when it was repeatedly disconnecting and reconnection, it was masquerading as a keyboard device, directing a browser to download more malware in the background, etc. Not outside the realm of possibility here.

I just un-installed NVidia's "USB Type-C driver" (Their newest Turing cards include a Type-C port), in the hopes that perhaps that was installing filter drivers that conflicted with the AMD and Microsoft USB3.0/3.1 drivers for my AM4 board and Ryzen SoC. Then I rebooted.

So far, so good, but maybe it will start to go again.

Maybe the PSU is heating up too, and the voltages are getting unstable.

As of the most recent 2901 BIOS update for my B450-F ROG STRIX, HWMonitor is showing descriptions for the voltage, and it shows +5V at 5.05, but +12V is at 10.0, and drops to 9.8 with both GPUs mining.

Edit: And I forgot to tweak a setting, last time I updated the BIOS. Under the Onboard Devices, way at the bottom, is an option for "Supply Power to USB Ports in Soft Off (S5) state". Which I believe, by default, powers the USB +5V off of the PSU's +5VSB, which has limited current capacity.

I disable this option, in an effort to have the USB +5V supplied by the actual PSU (powered on only) +5V line, which should be more stable.

 

[SamirD]

I'm not sure if that +12v should be that low. That could be the culprit right there.

 

[VirtualLarry]

Well, I'm not going to panic, I'm going to wait until I can dig out my meter. They say software readings are unreliable. I can't count the number of systems that claimed that some voltage or another was at 0.1V, and the system was running fine.

Edit: Checking the BIOS PC Health section, the +3.3V and +5V were slightly over, and the +12V (no GPU AFAIK in BIOS) was 11.968. A touch low, but within spec.

 

[mxnerd]

I also started to experience Realtek USB 3.0 GBe adapater connect/disconnect problem like yours yet again. So I'm giving up this thing.

It seems like a Windows 10 issue though. Since I never experience something like this before with any other OS.

 

[SamirD]

Well, I'm not going to panic, I'm going to wait until I can dig out my meter. They say software readings are unreliable. I can't count the number of systems that claimed that some voltage or another was at 0.1V, and the system was running fine.

Edit: Checking the BIOS PC Health section, the +3.3V and +5V were slightly over, and the +12V (no GPU AFAIK in BIOS) was 11.968. A touch low, but within spec.

True. Nothing beats an actual multi-meter or tester.

 

[SamirD]

I also started to experience Realtek USB 3.0 GBe adapater connect/disconnect problem like yours yet again. So I'm giving up this thing.

It seems like a Windows 10 issue though. Since I never experience something like this before with any other OS.

One way to verify this is to check it on linux. If it's having issues there--device. If it's not--windows error.

 

[killster1]

few questions, does your vacuum pass electricity through the plastic wand? what psu do you have? what ups do you use? what pc case?