Strange USB anomaly: "Unknown Device"

[BonzaiDuck]

BACKGROUND : I built this SLI rig in 2008 on a lark. It used the EVGA 132-CK-NF78 780i motherboard with an E8600 processor. I had been able to overclock it effortlessly to about 4.2 Ghz with an idle voltage of about 1.34V. I had installed an OEM VISTA-64 Ultimate OS. Never, never, ever -- had any problems or "strange things" going on with it. I've reset the BIOS to stock values -- planning to pass on the system to my brother (today's his birthday), after I make sure everything is ship-shape.

With absolutely no hardware changes other than the BIOS clock-speed and voltage, I installed an OEM Windows 7-64, which also mostly went forward without a hitch. Only after discovering the anomaly I'm about to describe, did I install a chipset-network upgrade (~2010) for the board. But these boards didn't really "need" a chipset driver: supposedly Win 7 "takes care of it." The anomaly persists.

The system uses a Silverstone Commander USB fan-controller. USB communication allows fan control through nVidia's "nTune" Performance program, and the latest version for Win-64 had been posted at nVidia, which I installed.

THE ANOMALY: Regardless of starting with a cold-boot or waking from sleep state, I get a message that the system was "unable to install a driver" for a USB device. But the only USB device connected to the system is the ESA Commander, other than the front-panel USB ports. The Commander shows up in the Device Manager tree under "Human Interface Devices" (HID) -- identified as "USB Input Device" and showing "ESA FW Update" under "details->bus-reported device description." So the Silverstone is still working. Another HID device also appears, shows to be "working properly," but has no identifier.

The Silverstone is connected directly to a motherboard USB port -- of which there are only two ten-pin plugs. The second plug is used to connect the front-panel USB ports (two) which also seem to work fine. One of the two connections is only "half-used" with a 4-pin USB cable, while the other has a ten-pin plug. [At this moment, I'm not sure which is which, but I can find out. It is likely I used the two-port cable from a rear-panel USB plate to connect the ESA, leaving the other USB plug insulated and unconnected.]

Why would there be no problem under VISTA-64, but this shows up with a Windows-64 installation? I havent' tested every single USB port, but both front-panel and those at the rear I/O plate seem to be working fine when I plug in a USB flash drive.

How can there be an "Unknown Device" when I've accounted for every active USB connection in the machine -- all which seem to be working?

Anyone have any insight in this little "irregularity?"

 

[Iron Woode]

I have seen issues like this before with Windows 7 64bit.

If there is no hardware id tag for the unknown device, then try a google search with the motherboard model, version of windows and unknown usb device and see if others are having the same issue.

 

[BonzaiDuck]

I have seen issues like this before with Windows 7 64bit.

If there is no hardware id tag for the unknown device, then try a google search with the motherboard model, version of windows and unknown usb device and see if others are having the same issue.

Well . . . I appreciate your response on this and every little thing helps. But I tried running similar search strings the day before, turning up nothing. And I tried again with your own suggestions, and still couldn't find an explanation.

I should also mention that I had temporariliy plugged in the USB communications connection for a CyberPower UPS before installing Windows, discovering this problem after the installation. Didn't install the CyberPower software, although Windows recognized the battery, and the connection was working. I was planning to move this machine upstairs with an APC UPS -- thus the temporary nature of the CyberPower.

I had to shut down the machine when I accidentally broke a blade off a side-panel fan, so I unplugged the CyberPower at that time. When the system first booted, the "Unknown Device" was gone. But eventually, after I let the machine go into sleep, it woke up with the Unknown Device showing again in the Device Manager tree, and a popup message again informed me that "Unable to install device" etc.

And that's where I am on this. I've only "disabled" the device, but it still shows up in Device Manager.

 

[vailr]

Latest bios for that board: P10
ftp://ftp.evga.com/bios/NF78P10.bin

 

[Iron Woode]

I am trying my best to help.

if no usb devices are plugged in does the issue go away?

perhaps vailr is on to something. Do you have the latest bios?

 

[mikeymikec]

nvm

 

[BonzaiDuck]

I am trying my best to help.

if no usb devices are plugged in does the issue go away?

perhaps vailr is on to something. Do you have the latest bios?

I'm beginning to think Vailr's observation about the BIOS version could -- COULD -- be the case.

Also, the one thing I haven't attempted to do is to disconnect the USB cable to the ESA fan-controller. The controller has its own microprocessor, and should continue to work with the current settings.

I'll review the notes and "fix-lists" for the newer BIOS versions. And, yeah, I first took a look at that and discovered my current BIOS is P06 (or was it 08?)

This is what led me to this "situation" with the 780i board and Windows 7. I'd built my brother's system in '08 with an mATX 610i (nVidia chipset) motherboard. Because of distractions unforeseen during the day I'd set aside to do the work, I was confused about the order of things I needed to do when I was upgrading his processor from an E2180 to an E6700 (Wolfdale-"Pentium"). I've been flashing BIOS's for 20 years! I'd be the first to tell you not to flash a BIOS that addresses compatibility with a new processor -- AFTER you installed the processor! But I did.

So Bro gets this 780i system. I've probably done BIOS flashes four-dozen times for a dozen machines or more, through several states of "BIOS update technology," -- never with any problem. This very last time was carelessness, a matter of interruptions and distractions, and fatigue.

So -- that's the next step, I think . . .

On the up side, I have four sticks of DDR2-800 RAM and a brand-new E6700 processor as "spare parts." The 610i motherboard -- is a casualty.

I'll remember to report back here the results of an update to version P10.

 

[BonzaiDuck]

OK! Apparently, updating/flashing to the newer P10 BIOS didn't fix anything per the "Unknown Device." Soon, I will cold-boot the system after unplugging the Silverstone ESA Commander, and see what develops.

Now -- without eliciting any stern remarks or "I-told-you-so's", or any long dialogs about OS-version ethics -- I can add here that I was looking for a Win7-64-Pro SP1 install disc and OEM license for Bro's system. It was either that, or VISTA-64 -- a retail-box which I still have.

I'd avoided buying "branded" OS discs until now, but the assurances that it was simply an OEM SP1 install with either a Dell or HP logo on it prevailed. Other than presets making a Dell web-page the Internet Explorer default Home-page, the assurances by the reseller seemed accurate.

In connection to the ESA controller, I discover through web-searches that Dell had produced a computer model with nVidia mobo that had an "ESA I/O" board, indications suggesting that it controlled fans and lights. There was a link for a firmware update for it, but no indication that the board was the same as the Silverstone.

And there are no firmware updates available from Silverstone.

Still yet planning to get to the bottom of it, I'll come back here after disconnecting the internal USB cable from the Silverstone Commander. We'll see what develops.

 

[BonzaiDuck]

OK . . . I've tried booting with the ESA Commander USB connection unplugged -- "No cigar" -- "Unknown Device" still shows up.

What was worse -- after updating the BIOS, I found that going into a sleep state -- hybrid, S3+S1, S3 or S1 (any of those) -- would cause the computer to freeze on wake-up. The display would come back, the HDD light would stop flashing, the mouse and keyboard would be frozen, and I would have to press "RESET."

Apparently, this is a problem not unknown to Windows 7 64-bit. After trouble-shooting that issue, reading countless forums, advice and other available material, I finally decided to reinstall Windows.

Which -- I did. I was very careful about setting things up in proper order. Installed my Kaspersky after the chipset and network drivers, then allowed the full boat of some 62 Windows Updates. After that, I corrected a "WMI Error Code 10" with Windows "Fixit." Now, all my event logs are "in the blue."

The "Unknown device" still appears in the Device tree, as I said. Now -- the only thing that was connected when I disconnected the ESA Commander was the cable to the front-panel ports. This should NOT cause a problem, nor do the ports seem to be malfunctioning. Something else is going on with that.

Now -- I'm testing the sleep state, but only by putting it into hybrid sleep for a few minutes and then waking it. Right now, that seems to again be normal.

But my purpose with this computer upgrade is to reduce household power-consumption. I'm going to feel edgy about this setup until I'm sure that the system no longer freezes after indefinite periods of sleep.

 

[vailr]

Maybe: re-install Vista, then run a driver capture program such as DriverMax:
http://majorgeeks.com/DriverMax_d5260.html
Or: just use Windows Device Manger to try and narrow down exactly what the "unknown device" is, and the driver being used in Vista.
DriverMax can also be directed to take whatever saved Vista driver was working before, and then (maybe?) re-install the same driver in Windows 7.

Note: take screen captures of Device Manager under Vista, then compare that with Windows 7's Device Manager.

 

[BonzaiDuck]

Maybe: re-install Vista, then run a driver capture program such as DriverMax:
http://majorgeeks.com/DriverMax_d5260.html
Or: just use Windows Device Manger to try and narrow down exactly what the "unknown device" is, and the driver being used in Vista.
DriverMax can also be directed to take whatever saved Vista driver was working before, and then (maybe?) re-install the same driver in Windows 7.

Note: take screen captures of Device Manager under Vista, then compare that with Windows 7's Device Manager.

At this point, I'll file that advice away for further reference. But there aren't any other devices hooked up, other than the onboard USB controllers. There's the "Standard Enhanced USB to PCI USB Host Controller," and the "Standard OpenHCD USB Host Controller." Neither of those is "not working properly." What's left are "USB root hubs" and the "Unknown Device."

After going through this second Windows 7 installation pass, I'm not eager to fiddle with VISTA-64 to "see what gives."

At least one development proves promising, but I'm not out of the woods entirely. that's the Wake from Hybrid Sleep feature -- it's working again. I really need to find out if it will "work" after the machine has been asleep for the good part of a day. My theory -- departure from my usual and longstanding perfectionist behavior with this technology -- I can forget about the glitch on the device tree if everything else works and I have what seems to be full USB 2.0 capabilities on the ports I'm using.

I'm beginning also to wonder if it doesn't have something to do with the "branded" OS, but I can't see how it would.

 

[vailr]

@BonzaiDuck:
I'd still recommend using DriverMax in Windows 7, to see whether there's any driver updates to consider downloading.
Note: the free version has a limit of 2 driver downloads per day, but does give sufficient version information for downloading any additional drivers via Google search.
Also: recommend adjusting the program settings, so that it doesn't auto start at every system boot time.

 

[tweakboy]

I have solution. Dont use hibernation or sleep.

A lot of these mobos no matter what brand dont like hibernation especially when you have SSD you shouldnt use any power savings. If you want stability.

 

[BonzaiDuck]

Thanks, Vailr and Tweakboy.

From Vailr's wording, I was almost thinking there wasn't a version of the mentioned software that worked with Win 7, but he seems to indicate there is.

I was "in and out" today -- me old "moms" goes through bananas like a gorilla on a food binge, so I had to make a grocery-produce run. Let the eVGA board and newest Win 7 installation go into sleep. Couple hours -- it comes back, tip . . . top.

And Tweakboy, the whole point of this is to reduce power consumption. I'm carrying two other people in this house, I build all their hardware and do all the maintenance with our LAN. After seeing how much the electric bill is reduced with improvements made so far, I can now see that S3 or hybrid sleep will reduce wattage draw by almost 200W for Bro's machine.

We leave these things on 24/7. My computer and home-theater are integrated. Moms is always doing genealogy and other s*** on her machine, and Bro has . . . certain bidnis . . . . Add to that the WHS server.

I have to come up with a more energy-efficient scheme. I think I can cut our power-bill by $60 to $100/mo. Sheesh. I don't even worry about the "trickle-wattage" my friends talk of in reference to other "electronic equipment."

 

[Iron Woode]

I have one more suggestion:

unplug the front panel usb cable. Have nothing at all plugged into the usb connectors on the motherboard itself. Just the plain motherboard and see if the unknown device goes away. You could have a bad onboard usb connector.

 

[BonzaiDuck]

I have one more suggestion:

unplug the front panel usb cable. Have nothing at all plugged into the usb connectors on the motherboard itself. Just the plain motherboard and see if the unknown device goes away. You could have a bad onboard usb connector.

Yes -- and I will do that. But I must observe here that both the front-panel plugs work properly when I stick a flash-drive in them.

Not happy with Vailr's software recommendation. As soon as I downloaded, installed and ran DriverMax, it made me . . . uncomfortable. You have to "create an account" to use it. I didn't feel like trusting its information about drivers, or where to get them. It didn't reveal anything new about the system, or so I concluded. Suddenly, I wanted to uninstall it.

After I did that, my web-browser wouldn't connect to my home page. I tried a restore point -- no cigar. Had to reinstall the OS (as I write this -- Windows updates are completing.) This is becoming so routine, I should start a business solely to serve people with old nVidia chipsets!

 

[BonzaiDuck]

Does Windows 7 support "legacy" USB 1.1? The vague identifying information in "Details->Properties" window for the "Unknown Device" shows "Standard USB Host Controller" for "Manufacturer."

It's just a hunch. But when I set up the BIOS on these older boards, I always leave the default USB setting to include USB 1.1.

 

[Iron Woode]

Does Windows 7 support "legacy" USB 1.1? The vague identifying information in "Details->Properties" window for the "Unknown Device" shows "Standard USB Host Controller" for "Manufacturer."

It's just a hunch. But when I set up the BIOS on these older boards, I always leave the default USB setting to include USB 1.1.

as far as I know it does.

I always set legacy usb in the BIOS.

I have also seen pinouts for USB different on the front panel from whats on the board.

 

[tweakboy]

Thanks, Vailr and Tweakboy.

From Vailr's wording, I was almost thinking there wasn't a version of the mentioned software that worked with Win 7, but he seems to indicate there is.

I was "in and out" today -- me old "moms" goes through bananas like a gorilla on a food binge, so I had to make a grocery-produce run. Let the eVGA board and newest Win 7 installation go into sleep. Couple hours -- it comes back, tip . . . top.

And Tweakboy, the whole point of this is to reduce power consumption. I'm carrying two other people in this house, I build all their hardware and do all the maintenance with our LAN. After seeing how much the electric bill is reduced with improvements made so far, I can now see that S3 or hybrid sleep will reduce wattage draw by almost 200W for Bro's machine.

We leave these things on 24/7. My computer and home-theater are integrated. Moms is always doing genealogy and other s*** on her machine, and Bro has . . . certain bidnis . . . . Add to that the WHS server.

I have to come up with a more energy-efficient scheme. I think I can cut our power-bill by $60 to $100/mo. Sheesh. I don't even worry about the "trickle-wattage" my friends talk of in reference to other "electronic equipment."

Im with you, I understand now sorry. I mean the times you leave house for while or arent gonna use comp, putting on sleep pretty much turns off comp,,, except couple watts are still drawn from USB stuff you have connected. Use sleep, but at night, TURN off comp and turn off the PSU switch. Wake up turn on PSU power and boot up. Since you have SSD booting shouldn't be 4 minutes like mine was and be quick.

thanks gl

 

[vailr]

Not happy with Vailr's software recommendation. As soon as I downloaded, installed and ran DriverMax, it made me . . . uncomfortable. You have to "create an account" to use it. I didn't feel like trusting its information about drivers, or where to get them. It didn't reveal anything new about the system, or so I concluded. Suddenly, I wanted to uninstall it.

After I did that, my web-browser wouldn't connect to my home page. I tried a restore point -- no cigar. Had to reinstall the OS (as I write this -- Windows updates are completing.)

Sorry: guess you didn't read the disclaimer:

"Limitations: This program is advertising supported and may offer to install third party programs that are not required for the program to run. These may include a toolbar, changing your homepage, default search engine or other third party programs. Please watch the installation carefully to opt out."

And for future reference: an unwanted "Home Page change" is fairly easy to fix. No need for a complete OS re-install.
I use: Hijack This (free):
http://majorgeeks.com/Trend_Micro_HijackThis_d5554.html

 

[BonzaiDuck]

Sorry: guess you didn't read the disclaimer:

"Limitations: This program is advertising supported and may offer to install third party programs that are not required for the program to run. These may include a toolbar, changing your homepage, default search engine or other third party programs. Please watch the installation carefully to opt out."

And for future reference: an unwanted "Home Page change" is fairly easy to fix. No need for a complete OS re-install.
I use: Hijack This (free):
http://majorgeeks.com/Trend_Micro_HijackThis_d5554.html

MajorGeeks is a good resource. I've heard of Hijack, but never fully realized the advantages. OS topic maybe, but when some advice recommends a complete OS reinstall annually, I've been able to keep my various Windows version installations from the time I've built the machine, or at least three years and even longer. that -- without any major software conflicts or performance degradation.

But I'm not putting any gold stars on my forehead. When building a system, I would go through install and reinstall at most twice. Here, something turns up that looks like a hardware problem, but I also took a cheap route on the OS version -- which leaves me with less confidence about what I'm actually seeing reported.

Since you guys are so helpful in coming back to me in this thread, let me throw out a question and explain my purpose afterward: Are the USB controller(s) linked to the South Bridge or North Bridge? [File that away and address it after reading further if you wish. Suffice to say -- I don't have a little schematic stored accurately in my brain. I've seen it -- I just don't remember it . . . ]]

OK -- back to the main issue. Here are some "Properties" information for the USB "Unknown Device:"

USB\
VID_0000
PID_0000\5-2DE4EB9C-0-4

This is from the "Plug and Play ID" (in other words USB\VID_000\PID . . . . )

Now . . . I'd have to recheck the BIOS to see if there is a complete "Disable" option for the USB configuration, as opposed to "1.1 & 2.0" and "2.0" exclusively. If such is the case, and if there is a damaged port, then I could find a PCI (or PCI-E ?) USB card with an internal port or ports. The ones I've seen usually have one internal, but that's a minor problem.

Otherwise I could just disable it, I suppose. But those USB PCI cards were cheap some time ago, so they'd be just as cheap now.

 

[BonzaiDuck]

UPDATE:

With more specific information drawn from the "Unknown Device" Properties table, I ran another web search. Apparently, this is a problem with Windows 7 reported by people from 2010 to 2012, for both desktops and laptops.

http://answers.microsoft.com/en-us/...-code-43/aacd4dfc-2a7d-4090-b6b7-db5d9a15db1b

One respondent says "I've been building computers [for years] . . . . I am stumpted . . . "

Probably no different in a more generalized way than the other respondents to that particular thread, the problem resides on my "embedded hub" #0001 and is specific only to one port on that hub [I think it is port #0004, but I'd have to look again for the precise port #.]

In my case, I'd be reasonably sure that my port #0004 will not respond with some external device hooked up to it, since all the other devices and extensions I'm using are working. Or put another way, I just don't happen to be using the port which shows "Unknown Device."

So I suggest the problem is not that the "standard USB host controller" has gone bad on my motherboard, or it is less likely since so many other people with a variety of machines also report the problem -- with either Windows 7 x64 or Windows 8, as a matter of fact.

And again -- going back to my early response to the discovery, there aren't any further updates to my chipset drivers from eVGA beyond the last one posted at eVGA for Windows 7 (32 and 64).

Also, one more point. Of the two "devices" using my USB ports (the internal motherboard USB plugs), neither is using their ports to "power" the devices. That is, the only device (extension) that requires power is getting that power from a direct connection to a UPS Molex connection.

 

[tweakboy]

Ignore the thing. In device manager disable the device....... Its just winblows being winblows.

It will go away soon. Trying using compm few days then go to device manager and enable the device and question mark might go away. Personally your gonna deal with it until you upgrade. Its pointless it comes and goes ,,,,,,,, dont stress yourself as long as all your USB stuff work. gl

 

[BonzaiDuck]

Ignore the thing. In device manager disable the device....... Its just winblows being winblows.

It will go away soon. Trying using compm few days then go to device manager and enable the device and question mark might go away. Personally your gonna deal with it until you upgrade. Its pointless it comes and goes ,,,,,,,, dont stress yourself as long as all your USB stuff work. gl

I agree completely. Although I don't see it "going away" in some haphazard way. They'll have to offer some sort of driver update through Win/MS Update, I'd think. It isn't an "intermittent" failure at the moment.

But I don't think it's a "hardware failure." I entertained that idea before reflecting on the forum posts all around with people stymied by it. It looks like a driver imperfection specific to older chipsets, but howsoever that could be, it's "WinBlows."

Also -- this. Used to get the latest Windows OS as soon as released, until XP. Stuck with Windows 2000 for a couple years. I don't think we upgraded to XP until at least the first or second SP. I think it was 2003 or 2004. After that experience, I might try a "release candidate" on an available machine, but defer "investing" in the new OS until the first SP release.

Anyway, I THOUGHT I saw somewhere that someone was having this same sort of troubles with Windows 8! Just likely it is how they address legacy motherboards and chipsets. IMHO . . . .

But it's just a single port on this machine, and I can disable it. All the event-viewer logs are "in the blue" now. Oh . . . . And there's that "WMI Error code 10" glitch (an imperfection in Win7 install disc) you have to resolve with the "FixIt" button. Got rid of that . . . too . . .

How many mainstream users -- even in 2012 -- do you think examine their Dev Mgr tree or their App and System Event logs? Not that many, or so I guess . . .

 

[bionicled]

What brought me to here was that my usb key worked last night but in the morning when I plugged it in the file sys went from FAT32 to RAW and the device manager the same "unknown Device" listed with the ESA FW Update.

I dont know if the 2 are connected or not. I'm still learning.