The Google tells me I'm not the only one with this issue but I haven't seen anything that looks like a solution. I have a very stock setup, 8GB RAM, 750GB WD Black HDD, HDMI monitor, KBD/Mouse connected via KVM all running under Windows 7 Home Premium. I have a USB 3.0 hard drive cradle, external powered with 2 bays. If I put a drive in one of the bays and try to run a CHKDSK or create an image backup, the process fails due to the drive disconnecting. Most of the time the drive immediately reconnects but the damage is done at this point. When it occurs, Windows sends the pop-up window announcing that a new drive has been detected. FWIW, Windows then considers the drive dirty so a CHKDSK needs to be run resulting in an endless cycle eventually leading to madness.
This seems to happen only under heavy load. If I simply leave the drive in the cradle and even if I peruse it, there is no disconnection. It's not the drive itself, have tested several both 3.5 and 2.5. It's not the cradle, I've been using it for quite a while on different machines and only just put it on the NUC. It would not appear to be old or missing drivers as I just updated them from the Intel site today. I believe I am on the current BIOS version, updated last week. There are no messages in the Event Log that indicate a problem.
The feel here is that either the driver or the USB hardware simply can not tolerate a sustained high data rate. Has anyone ever solved this problem?
Hello Shamrin, thanks for joining the Intel Community.
Would you please try the following workaround to make it work?
How to install
***** WARNING *****
Do not run this driver's installer (Setup.exe) from a USB storage device (i.e. external USB hard drive or USB thumb drive). For proper installation, please copy driver files to a local hard drive folder and run from there.
After installing this USB 3.0 driver, your USB keyboard and mouse may not work in Windows*. This is only a driver installation related problem and is currently under investigation. Follow these steps to solve this problem:
1. During boot, press F8 to access the Advanced Boot Options menu.
2. Select "Last Known Good Configuration (advanced)".
Test the system and let me know the results
Thanks for the reply Sylvia. Alas, the new driver did not solve the problem. I ran an image backup, backing up a drive in the USB 3.0 cradle to the internal SATA drive in the computer. At the time the drive went offline it was averaging a read-rate of 848Mb/s and Write-rate (to the internal) of 792.Mb/s.
The Windows event logs are sort of interesting. The backup software reports that it started it's process at 20:27:54. From the System Log, I see 4 warnings at 20:27:22 saying "The speed of processor [X] in group 0 is being limited by system firmware" and that it's been in that state for 71 seconds; it's hard to see how that's relevant since it precedes the backup, but the coincidence is interesting. It looks like the external drive fell offline at 20:28:07 so the whole thing didn't last too long.
This morning I ran a CHKDSK /R on another drive in the cradle. I started that at 9:02 and it ran until 9:05 when the drive went offline briefly and the job failed. This time there were none of the processors speed warnings so maybe that's a red herring.
Let me know if there is any other data that I can provide to help.
Sorry for the delayed response, I have been away on holiday.
I'm missing the relevance here. The thread you mention points out an issue where USB 3.0 interferes with connections on the 2.4GHz wireless band. The problem I'm having is on the USB3.0 port itself and it is disconnecting when used to transfer data. I do not have any wireless devices connected to the computer. Maybe I'm being thick but the referenced link appears to be orthogonal to the problem.
I am having the exact same problem with a D54250WYB the 2 x USB ports on the same side as the LAN keep disconnecting randomly. I RMA'ed the board and the replacement does the same.
So I swapped the board for a lower D34010WHY and have exact same problem on LAN side USB ports.
Ports on Non-LAN side work without issue with zero disconnects.
This appears like a manufacturing issue to me, likely an issue with no fix hence silence from Intel.
Experiencing similar problems with my NUC5I3RYH. Windows Eventlog is full of
Disk 1 has been surprise removed.
An error was detected on device \Device\Harddisk1\DR11 during a paging operation.
This is causing corruption of the files being created. I'll try connecting it to the front USB ports and see if that makes a difference
In this case, try accessing the BIOS, select advance and then select devices, once your there make sure the xHCI mode is set to Smart Auto.
If you don't have the option for the xHCI then you will need to update the BIOS to the version 0246 or later so you can get that option in the BIOS.
My NUC was set to Smart Auto, so I changed it to ACHI. I am still getting the USB surprise removed issue. I have no Intel drivers installed at this time and I am running Build 10586 of Windows 10. I have now done the following in the advanced power options: Turn Off Hard Disk After(Never) and USB Selective Suspend Setting(Disabled). I am testing to see if the disconnect happens again.
I am starting to believe this is a manufacturing issue. If this does disconnect then I am going to connect to the front USB ports. Dear Intel what is the difference between the front and back USB ports? Something must be different.
Edit: it looks like the combination of XCHI set to enabled, and Turn Off Hard Disk After(Never) and USB Selective Suspend Setting(Disabled) in the power settings works for me. My drive has not disconnected in a day.
This explanation of the various BIOS-settings when it comes to USB 3.0 (xHCI) seems legit:
XHCI Mode = Disabled - The on-board USB 3.0 port function like a 2.0 port
XHCI Mode = Enabled - The on-board USB 3.0 port function like a 3.0 port
XHCI Mode = Auto - The on-board USB 3.0 port function like a 2.0 port before OS USB 3.0 driver load. If you reboot the OS, the on-board USB 3.0 port again function like a 2.0 port during this reboot BIOS phase before OS USB 3.0 driver load.
XHCI Mode = Smart Auto - The on-board USB 3.0 port function like a 2.0 port before OS USB 3.0 driver load. If you reboot the OS, during this reboot BIOS phase, BIOS is "Smart" enough to avoid downgrade the USB 3.0 port back to 2.0 functionality before OS USB 3.0 driver load. So Smart Auto is faster than Auto on 2nd boot onward, but Enabled is fastest once you are sure the OS has the USB 3.0 driver installed, because it avoid the switching. Making the on-board USB 3.0 port function like a 2.0 port is mainly to support OS installation or to support OS that does not have build-in USB 3.0 driver, so that the USB keyboard would still work if the user plug-in a USB keyboard or any other USB devices into the USB 3.0 ports before the OS is installed with the USB 3.0 driver come with the motherboard.
Personally I would verify that you use the latest BIOS-firmware along with this setting:
XHCI Mode = Enabled
and finally that you have installed the proper drivers when it comes to USB in Windows:
https://downloadcenter.intel.com/product/76977/Intel-NUC-Kit-D54250WYK Downloads for Intel® NUC Kit D54250WYK
Unless they are included with your OS.
Another troubleshooting thing you can apply is (except for having latest BIOS and XHCI = Enabled) is to boot of a Ubuntu 15.10 live-image and see if you get the instability when in Linux or not.
Could be anything from broken implementation by Intel (not the first time - look at the screw up Intel did with the sleep mode and D54250WYK who bricked the device, this wasnt fix not until BIOS version 0040 or so, 0041 is the currently latest version) to bad physical interface on either your NUC or your USB-device, bad cable, bad physical connection or hopefully - just a bad driver (which the test with Ubuntu would verify which of these possibilities can be the root cause).
Any news on this? I just purchased the NUC5PPYH (N3700 Braswell CPU version) and am getting disconnects on the rear USB ports. This is on a new install of Windows 10. As reported elsewhere rhe front usb ports seem to be fine. Very annoying.
Please updating Windows and keep Windows up to date and update make sure you have the newest driver or chipset drivers installed; you can download then at the following link:
https://downloadcenter.intel.com/product/87740/Intel-NUC-Kit-NUC5PPYH Drivers & Software
You can try also updating the BIOS; you can download the latest BIOS vision at the following link:
https://downloadcenter.intel.com/download/25674/BIOS-Update-PYBSWCEL-86A- Download BIOS Update [PYBSWCEL.86A]
BIOS update instructions:
http://www.intel.com/content/www/us/en/support/boards-and-kits/000005724.html F7 BIOS Flash Update Instructions for Boards and Kits
Do you have too many devices plug into your USB ports? Many devices plugged into the USB ports can be pulling more power than the onboard USB chipset can output. May want to try a powered USB hub.