Renesas USB 3.0 PCIe x4 issue + solder question.

[Wobbler]

So finally got my zenith extreme + 1950x, on latest BIOS (0801) and up to date Windows 10 PRO 64 bit. After realizing that I had to up the soc voltage a bit from the auto setting(< 1) to get system stable with rated memory clocks. it been working fine apart from this issue....

Renesas Electronics USB 3.0 Host controllers

This device cannot start. (Code 10)

{Device Timeout}
The specified I/O operation on %hs was not completed before the time-out period expired.

On the newest driver that the StarTech.com (PEXUSB3S44V) offers + windows tries to change to universal drivers that won't work either. It's plugged in to the x4 slot on the MB, bios was on AUTO I changed it to Gen. 2 no difference. So I gather error has something to do with the MB PCIe controller rather than the add-on cards drivers.

Heck when I plugged the USB add-on into the x8 slot on boot it tries to test NVram and gets stuck with code: 92, Funny thing is my SATA 3 HW raid x2 add-on card (PEXSAT34RH) works on the first x16 slot (noctua blocking rest of the slot). Then again I had some other Test NVram code: 42 errors at random on restart, stuck at the first grey screen, when trying different voltages and cpu overclocks... but Graphics card works fine on the second x16 slot.

Oh and one more thing. Is it normal for USB-C solder joints(2) back of the MB to be just lightly touching, as they are in symmetry I gather is might be but looks suspicious being so thin/small connection.

[Wobbler]

Hmm... something odd happened.

Reinstalled windows, installed latest drivers 0/4. Left the computer on for overnight, it went to sleep. woke it up this morning, installed 1 thing from the old "AMD_ChipsetDriver_Win10_V900008_20170726" OEM something, that the 17.5 didn't install. Then Rebooted because NVidia failed to detect correct display settings. during boot it got stuck at loading circle (fast boot UEFI) had to change boot settings in BIOS, tried to go to legacy... booted back to BIOS, revert setting and started normal boot and windows loaded. The odd part is somehow that course of events resurrected 1/4+1 of the USB controllers on the add-on card, as the software started informing about superspeed USB ports available also checked device manager.

Tried reinstalling drivers again remove reinstalling only 1/4+1 controller loads, removed the PCIe card and reinstalled... back to 0/4. So tried to redo by uninstalling the AMD drivers reinstalling in same order, nothing... only thing left to test is reinstall windows and update etc.

For some reason I can't uninstall 17.5 need to install 17.3 on top of it to uninstall.

[xeromist]

Wobbler wrote:
Is it normal for USB-C solder joints(2) back of the MB to be just lightly touching, as they are in symmetry I gather is might be but looks suspicious being so thin/small connection.

If they are both ground pins then it's fine. You'd only have to worry if one was current/data and one ground. In that case you'd have a short and the port wouldn't work correctly.

[Korth]

USB-C connector pinout here. It's okay to bridge the V+ pins (red A4-B9 and A9-B4) and to bridge the GND pins (green A1-B12 and A12-B1). In fact, the pins are mounted to a common electrical substrate, they're already soldered together underneath. Visible solder bridges on the topside won't open/short any electrical paths unless they bridge other adjacent pins, they just look sloppy.

[xeromist]

Korth wrote:
USB-C connector pinout here. It's okay to bridge the V+ pins (red A4-B9 and A9-B4) and to bridge the GND pins (green A1-B12 and A12-B1). In fact, the pins are mounted to a common electrical substrate, they're already soldered together underneath. Visible solder bridges on the topside won't open/short any electrical paths unless they bridge other adjacent pins, they just look sloppy.

Yes, that's a better explanation and probably more useful if he wants to actually identify the pins concerning him. Thanks. 🙂 +rep

[Korth]

And I thought I was the only neurotically obsessed nerd who carefully scrutinizes every part, pin, and point on my mobos, lol. Most people don't peer into the bottom of their USB receptacles unless they're looking to fix some kind of problem.

[Wobbler]

Can't really check now what pins = what solder point but, here is a crude drawing from memory of what solder points were connecting (circled) behind the board near the USB-c port. They did look really sloppy compared to the clearly purposed jointed solder points, like robot lifted and dragged thin solder tin to to next point.

Also asked StarTech support about the USB cards driver problems and he commented that "It must be a compatibility issue with the other USB controllers", Renesas drivers being from 2002 just the installer being updated, so AMD/ASmedia perhaps, need to see if I can disable some of the USB ports in BIOS to test.

[Korth]

So this is all about a StarTech PEXUSB3S44V 4-Port PCI Express USB 3.0 Card, built around a Renesas µPD720202 chip.
Release note - Manual - Datasheet - %20USB%203.0%20Controller%20Card.zip"]Driver - Firmware Update

Note that Win10 is not explicitly supported. WinXP, Win2K, Vista, Win7, Win8/8.1 are supported.
Of course the StarTech support staff aren't going to readily admit the "compatibility" problem is caused by their own product or their own outdated drivers.

I see that similar USB3.0 PCIe adapters made by Anker, BYTECC, Corn, IOGear, ORICO, Rosewill, SUNIX, SIIG, Syba, Tek Republic, Vantec, and VisionTek all have reviews/complaints about incompatibility with AM4/TR4 or Win10. The only products which have no compatibility complaints are those from acomdata and HooToo (because, I think, nobody has actually bought any of them) and those from ASUS (when used on ASUS motherboards, anyhow).

[Wobbler]

Lol again magically one of the Renesas USB controllers came to life when I had issues with Graphics card, NVidia newest drivers signal 4:2:2 screwed up TV hdmi input (turned colors to red/orange on all devices connected to that hdmi) had to reset TV to get back to normal. And after nights shutdown the USB controlled stopped working again.