LGA1155 motherboard - no USB ports work

Submitted ⁨⁨2⁩ ⁨days⁩ ago⁩ by ⁨glitching@lemmy.ml⁩ to ⁨askelectronics@discuss.tchncs.de⁩

https://lemmy.ml/pictrs/image/d2a63eac-4e7e-4bdb-bd5c-f0f3003c3ded.jpeg

trying to save an old motherboard, it’s a Gigabyte GA-B75M-D2V. tried searching, got nothing usable back.

problem: all USB ports don’t work - all the ones on the back, both USB 2.0 and 3.0, and all front ports. USB ports not working is common accross boards and manufacturers, but it’s usually a subset, some work some don’t and you can still use the thing. it’s not a software issue. got a PCI USB 3.0 card, but that thing doesn’t work for booting off of it or in UEFI setup, etc.

otherwise, everything works - LAN, sound, audio, SATA, etc. I feel the platform still has a lot of life left, especially for my use cases.

no visible burns or busted caps and such. hoping it’s something simple, any ideas what to look at?

source

Comments

Sort:hotnew top

Romkslrqusz@lemmy.zip ⁨1⁩ ⁨day⁩ ago
You’ve confirmed that you have 5V / peripheral devices are being powered

Measure resistance between USB’s D+ and D- as well as resistance between each of them and GND

source
tychosmoose@lemmy.world ⁨2⁩ ⁨days⁩ ago
Did this happen over time or suddenly all ports dead? Or maybe this is new to you and unknown.

I would try clearing the CMOS is you haven’t already. Unplug power, remove the CMOS battery and wait a minute or more (or short the + and - contacts for a few seconds).

If that doesn’t change things, maybe inspect the CPU. Any problems visible on its contacts or the socket?

source
- glitching@lemmy.ml ⁨2⁩ ⁨days⁩ ago
  got the board for free, it had those issues with the dude’s CPU, it has it with mine too, socket looked OK, no bent pins I can see. isn’t a settings or software issue, tried all that.
  
  source
  - tychosmoose@lemmy.world ⁨2⁩ ⁨days⁩ ago
    Just to make sure - you physically drained the CMOS to reset it, right? That can have different results vs resetting to defaults in the BIOS/UEFI interface.
    
    source
    -> View More Comments
hendrik@palaver.p3x.de ⁨2⁩ ⁨days⁩ ago
You could plug in some USB fan, light, … and see if there’s power on there. Maybe prepare a harddisk on another computer, boot Linux and look at the debug logs. See if the USB chip gets recognized. And maybe learn something about the topology… If it’s just a single controller or multiple ones in there.

First thing I’d do is reset the BIOS. Maybe the previous owner disabled all USB ports for security reasons. Or they threw it out because they don’t work.

source
- glitching@lemmy.ml ⁨2⁩ ⁨days⁩ ago
  has 5V on all ports I measured. UEFI doesn’t detect keyboard or mouse and they don’t light up.
  
  source
- glitching@lemmy.ml ⁨2⁩ ⁨days⁩ ago
  gonna try measuring the voltage and report back. tried clearing PRAM, isn’t a settings thing. yeah, was busted already, that’s how I got it.
  
  source
j4k3@piefed.world ⁨2⁩ ⁨days⁩ ago
Use linux-hardware.org to search for a relevant kernel scan. I think that should have some kind of potential hardware identification for the USB hub and PCI trees. That may clue you in about the architecture and devices. I’m no expert on motherboards or the evolution of hardware devices. However, it could be as simple as a USB hub chip on the board. I only know that these things exist from around 5 years ago when I was researching how I might make a USB 3 to PCIE adaptor for an Intel Core Duo machine. I was looking into adding a faster external drive port to libreboot hardware.

From what I recall, the CPU has a ton of connections on its various I/O pins, (like you indicated checking for bent pins). However each register port has a ton of compromises to make at the board design level. Like, the hardware spec of the CPU says it has way way more stuff than is actually possible to implement in practice. The base CPU spec may say it has a half dozen USB ports and three 32 bit PCI lanes or whatnot, but when one gets into the weeds of hardware design, one quickly finds stuff like one USB port is hardware available in practice in the same hardware register as something like the Ethernet controller, but if you use it, the Ethernet speed is halved, and the other five hardware USB ports are only available if one disables all the PCI lanes. So instead, the hardware designer uses a GPIO pin against the enable pin of a USB hub to PCI converter chip that sits on the same PCI bus as all other external hardware. This was also very common in older generations of laptops. There was often a small daughter board that had a set of ports for USB, SD cards, Ethernet, etc., and this worked in a similar topology where it was a PCI device.

The first rule of troubleshooting is “thou shall check ground”. Beyond this, if you happen to discover that the USB ports are handled by such a controller, if you look up the datasheet and pinout, the Enable pin in the first to check after confirming power is present. Also be sure you have the kernel module or driver necessary to identify this device. Worst case scenario, if you identify there is an external controller, you could simply replace it. There is no programming in such a chip. It is just a simple hot air rework swap job. Alternatively, maybe try an external card or adaptor on one of the other ports or on PCI.

source