Windows 7 Intermittent USB 3 Issues Causing Boot Problems and Mouse Malfunction on Windows

[seekermeister]

It took about 6 tries to boot to Windows a little while ago. First try - couldn't get the login field to open for a password. second try - got the password entered, but after a prolong period, it popped an error simply saying access denied. third try - booted into safe mode okay. fourth and fifth tries got no cursor response from the mouse, and couldn't login from the keyboard. Sixth time, reached desktop, but still no mouse.

I changed the battery and swapped it around between the 2 front USB 3 connectors...no joy. The same happened when I connected using the back 2 USB 3 connectors. Finally got it to work using the USB 2 connectors on the motherboard.

Don't really know if the boot problem was due to the mouse problem or if it was just a coincidence. Checked the Device Manager and it says that both of my 2 USB 3 host controllers and both USB 3 root hubs are okay and have no conflicts.

Anyone with any idea of how this kind of problem comes about and how to fix it?

 

[Digerati]

Yes, Voltage in and out.

If so, it might explain some readings given by SpeedFan that I have never understood. I shall attach a screenshot to show what I mean. It lists 3 different Vin values. If they mean voltage in, then where are they measured, and what should they indicate to me?

I like SpeedFan but it has many limitations that are common to all 3rd party monitors and that is properly matching sensor to label, and they often give data on things we just don't need, or have any control over. And that varies CPU to CPU, and more importantly, motherboard to motherboard. For this reason, when it comes to HW monitoring, I generally like the monitor provided on the utilities disks that came with the motherboard, or from the motherboard's download page. I do use CoreTemp to monitor my CPU temps, however.

If you are not overclocking, don't worry about those voltages. The voltages you need to care about are 12VDC, 5VDC and 3.3VDC as those are provided by the power supply. The other voltages you see are motherboard generated from the PSU supplied voltages needed by the CPU, RAM, and other devices.

Have you checked out Link Removed from the makers of CCleaner. It does not run full time in the background, but it is very informative.

 

[seekermeister]

I have used other software for monitoring. The only reason that I've stuck with SpeedFan is because I can run it continuously and have a visual temp reading by just glancing at the system tray.

I would go with some other software, if I found something better in terms of convenience, but so far that hasn't been the case.

 

[Digerati]

I have used other software for monitoring. The only reason that I've stuck with SpeedFan is because I can run it continuously and have a visual temp reading by just glancing at the system tray.

That's why I use CoreTemp - it sits in my tray and I can have it show the highest core temp, or temps for each core. Plus, there is a CoreTemp gadget that is nice.

Again, SpeedFan is a good program, but as noted, it often does not properly match sensor to label. It might show the system temp as the CPU temp, for example. However, you can use Speccy, SIW, or Everest to correctly match sensor to label, then edit the labels in SpeedFan to properly reflect the correct information.

 

[seekermeister]

Actually, I also use CoreTemp, but with the gadget, because I didn't see an option for it to run from the tray. Core temps are good, but I kind of prefer the aggregate outer temp. Having both at the same time is even better.

 

[Digerati]

Not sure what you mean by "aggregate outer temp" but as I noted, you can have CoreTemp display in your tray, the temp of the warmest core at that moment, or the temps of all cores. I don't see the need to know all four (in my case) core temps. All I want is the highest temp. To display all four in my system tray adds to the clutter already there. That said, the gadget displays each core temp, and more.

 

[seekermeister]

Another reason that I use the gadget, because it displays all 6 of my cores in a neat fashion.

Aggregate outer temp is simply a phase I coined, because I didn't know what to call it otherwise. SpeedFan gives this temp reading, instead of the core temps. As I understand, it is measure by a sensor on the outside of the CPU, instead of reading the cores separately. It is always much higher than the core temps on my system, but also a more stable reading than the core temps which fluctuate fairly rapidly when in use.

I've never been sure which temp correlates best with the manufacturer's (AMD) temp limit, but I think that it is by TjMax, which relates to actual core temps.

 

[Digerati]

SpeedFan gives this temp reading, instead of the core temps. As I understand, it is measure by a sensor on the outside of the CPU, instead of reading the cores separately.

Ummm, no. Outside the CPU implies and assumes the motherboard maker provided a sensor for the CPU. They don't because AMD and Intel put temp sensors in their CPUs and the Chipset makers provide a method for reading it. The hardware monitors simply piggyback on to the existing motherboard sensors (as monitored in the BIOS Setup Menu). If the sensor were outside the CPU, ambient temps, the temps of air being blown onto the CPU, would interfere. This is true even if the sensor is located on the die case, and not inside each core. In other words, "outside" is totally inconsistent across CPU and motherboard lines, and each environment they operate in. So I go for core temp, and I am comfortable with just monitoring the warmest in my system tray - though I can easily see all with the gadget.

 

[seekermeister]

I guess so, but I do remember that with an AMD system that I had years ago, that there was an external sensor outside of the CPU located under it. That is no longer the case, at least visibly, but I had assumed that some such sensor existed somewhere.

Since SpeedFan's CPU temp is consistently significantly higher than any core's temp, I still don't see how it arrives at that figure, unless it is reading something other than the sensors that you mentioned.

 

[Digerati]

And different readings is why some folks dropped SpeedFan. Again, you should be able to use Speccy or something similar to help you match up sensor to label. The system temp is typically the chipset.

 

[Mike]

There was an effort, some time ago, and perhaps even now, with some motherboard manufacturers, to put a temperature sensor between the heatsink and the top of the processor die. This is no longer really being used by anyone, as the cores themselves should provide far more useful information. The thought, at the time, was that you could measure the efficacy of your fan/heatsink set up to make sure that it was working properly, and that the top of the CPU die was not overheating. However, if the core temperatures are around specifications, this is a redundant concept. Measuring the core temperature is far more effective. The ambient temperature around your RAM, CPU fan, and motherboard in general can be very important, as this is a good indicator of proper airflow. There is a good suggestion that less airflow is needed with SSDs, as conventional hard disk drives generate a large amount of heat, and multiple stacks may require a fan enclosure themselves. Such is not the case with SSDs, but if the ambient temperature in the case around your RAM gets too hot you can and will blow out a stick of RAM over time. In general, good airflow throughout the case will prolong the longevity of your system for a long time. As time has gone on, miniaturized devices have been constructed with no need for a fan at all. There are heatsinks being developed that have no CPU fans. As heat requirements decrease, this becomes more plausible. Here is an example of a heatsink with no fan:

 

[seekermeister]

That Sandia cooler is intriguing, but I guess it isn't available yet, because the only one that I found Googling was some kind of axial fan for $195 apparently for an airplane, because it mentions alerting the pilot if there is a problem. Not sure how that would be applied to an airplane, but it certainly is too expensive for a computer.

Actually that fan is not at all similar to what is shown in that video, so I guess it is too early to expect to find one on the market. One thing that puzzles me about it, is that it appears that it would have to work in the horizontal plane so that it could ride on the air, but most computer fans operate in the vertical plane so I must misunderstand something about what they intend to do with it.

 

[seekermeister]

I thought it worth making note that all of the boot problems may have nothing to do with the CMOS after all, because I finally reconnected the H: drive, that disappeared along with C: a while back, it reappeared and was accessible, but that was by hotplugging it with Windows running. However on the next reboot it disappeared once more. Considering the possibility that it was just a bad connection, I moved it to the SATA slot next to it. BIOS detection hung for a while and Windows saw it enough to install the driver for it, but otherwise it was invisible.

When I opened Disk Management, it took much longer than usual to load the virtual disk manager, and when it did open, it popped a dialog saying that the drive needed to be initialized. I Selected okay, but after some more delay, it popped another dialog saying that it couldn't do it because of an I/O error.

Wanting to separate the drive as far as possible from any motherboard issues that might exist, I installed it in my external case. Again, the BIOS and Windows saw it and Windows installed the driver again, but Disk Management couldn't see it.

Backtracking my thoughts a bit...back on the first reboot, there was an abnormal high pitched hum which was barely audible during shutdown. In fact I didn't noticed it at all, until I heard the pitch of it change a couple of times, and somewhat later, just before shutdown finally occurred, the hum died away in a peculiar fashion. The hum never returned after that.

Also worth mentioning, the drive in question is the one that I posted about recently, which was an RMA for my 500GB drive. In testing it back then, though it passed the benchmark and error scans, on the speed map it had a long line of red blocks (very slow), which was the same blocks that had produced errors on the scans prior to having sent it in for RMA.

Apparently what ever they did to "fix" them wasn't entirely successful, because I had set it up to use for a backup archive for True Image, but when TI's validation of the first backup file hit those slow blocks, TI and the system froze. I continued to use the drive, since the first backup seemed to fill the area with the slow blocks, and the next backups were okay.

I'm going to send this drive back once more, but if I get trash back again, I shall be looking for another brand of drive.

 

[Digerati]

Since I have done some work at Sandia Labs at Kirtland AFB in Albuquerque, the Link Removed intrigued me too. Too bad the use of after-market coolers on "boxed" retail Intel and AMD CPUs that come packaged with supplied OEM coolers voids the CPU warranties.

As for possible CMOS issues, note battery performance is significantly affected by temperatures, and a borderline battery may be affected even more. Even a few degrees difference in the ambient (room) temperatures can make a difference to a weak battery. And a symptom of weak batteries is inconsistent recognition of connected devices. So, before ruling out the CMOS (and its battery), just like power supplies, I'd still swap in a known good one before spending money on anything else.

 

[seekermeister]

Oh, of course I will, when they arrives. But I think it also curious to note that all of the string of recent BSODs started occurring just a few days after I installed the drive in question. The last BSOD occurred just a few minutes ago, when I powered off the drive in the external case to replace it with another drive. Maybe it is just some kind of coincidence that it coincided with that operation, but I really don't believe in coincidences.

 

[seekermeister]

As far as an aftermarket cooler voiding the warranty on a CPU, that doesn't mean to much to me, after AMD returned a CPU to me with one of the pins bent flat in a position that couldn't have been an accident. Maybe it was just their way of showing their annoyance at me for sending in an RMA that wasn't really defective.

The only way that I would RMA a CPU now, is if it were DOA when I first purchased it, and I could send it back to the dealer, rather than AMD.

 

[Digerati]

But I think it also curious to note that all of the string of recent BSODs started occurring just a few days after I installed the drive in question.

Certainly something to keep in mind, but it would be more convincing if the problems started immediately, and not days later.

As far as an aftermarket cooler voiding the warranty on a CPU, that doesn't mean to much to me

And that's perfectly fine and I have no problems with that, as long as (1) users are aware of the terms of their warranty agreements, (2) folks don't recommend aftermarket coolers without noting it violates warranties, and (3) folks are upfront and honest with AMD and Intel when seeking RMA replacements. Omitting pertinent information when requesting an RMA is deceit, and deceit for personal gain is fraud and fraud is a serious criminal offense.

I have noted both AMD and Intel are very liberal with their RMAs when folks are honest about the circumstances, most likely because they want our return business. And that's great. But I have also seen both AMD and Intel refuse RMAs for users who have a history of returning CPUs that have been "abused". It is rare but it happens.

It is important to note, however, that OEM supplied coolers are indeed excellent coolers, more than capable of providing adequate cooling, even with mild to moderate overclocking. They have to be as they are they are guaranteed to support their CPUs for 3 years. No Aftermarket cooler covers CPU replacement. Their only real downside, IMO, is they are not the quietest coolers out there. They are MUCH quieter than OEM coolers of the past, but for HTPC applications, a cooler designed for silent operation may be better.

Also note that motherboard makers design motherboards placing heat sensitive and heat generating devices surrounding the CPU socket in anticipation of the expected downward firing OEM cooler fan - including the nearby chipset. Many aftermarket coolers fire sideways, or don't fire at all (they use liquid). If there is not adequate case cooling, these sensitive devices may be neglected good cooling. It is the case's responsibility to provide good cooling through the case. The CPU cooler need only toss the CPU's heat up into the flow.

 

[seekermeister]

Thanks for the lecture, you can come down from your podium now.

 

[seekermeister]

Certainly something to keep in mind, but it would be more convincing if the problems started immediately, and not days later.

I said a few days later because my memory is not what it once was, and couldn't pin point the exact moment events began to happen, however I do remember that when I first formatted the drive with Disk Management a BSOD occurred immediately. That BSOD and the one that I mentioned earlier today do tie together and relate directly to that drive. All of the rest of the BSODs could be for other reasons, but they did occur between these two, and were not something typical of previous behavior. For me, that is sufficient reason to believe as I do, until I'm proven wrong by a continuance of BSODs without this drive installed.

 

[Digerati]

Podium? I don't think personal insults were necessary. I am just a volunteer here trying to help folks with their problems. Sorry my help is not appreciated. Good luck with your problem.

 

[seekermeister]

I suppose that my comment could be taken as an insult, but it wasn't intended as such. However when anyone elaborates as you did in regards to a comment that I made that didn't indicate anything untoward in my actions or thoughts, you made it appear as though it did. Your response could only be taken as a reaction to my post which you felt was unethical or illegal, or you just took it as an opportunity to state your feelings in general on the topic for the benefit of everyone, including me, as a professor might do in a classroom. I chose the latter as your intent, because you have been very helpful otherwise. If I had chosen the former as your intention, then I might have responded in a fashion that would have left no doubt about my response being an insult. However you have the right to think about my statement in any fashion that you choose, but when you express those thoughts to me and the world, then I have the right to respond in return as I feel best.

The bottom line is any comments in any thread (other than those in general discussion or debate forums) should be confined as much as possible to dealing with the issues that the thread was either started on, or along the path that discussion sometimes evolves into. Now that I have violated that principle myself, I shall step down from my podium.