Replies

If you are just building a system and are experiencing BSODs then you should NOT install any application software. You should check & test all hardware.

The base system was bought from Dell Outlet, which is re-certified equipment with the same warranty as new. While I have some diagnostic software for the 32-bit XP class machines, I am new to 64-bit 4+ GB RAM machines. Long ago, I used to build systems from parts, and did hardware diagnostics at each stage. But, I wanted a ready-build system to avoid the that stage of the work. I put in a bigger power supply and a video card, with plans to add more RAM, and another hard disk, any possibly another optical drive. Due to my work schedule, I only work on the Win 7 system on the weekend, so all this has been taking a while.

But, I got weird BSODs from the beginning. I made the OS/software aspect as pure as possible, uninstalling all the crapware, e.g. Windows Live, and started getting utilities that show what is going on with a Windows 7 64-bit system.

An attempted RAM update didn’t pan out, so I started looking into nitty gritty hardware issues, like I should have done from the beginning, and found that CPU-Z reports the CPU voltage to my i3 550 as over 2 volts, while Intel specs say 1.4V max.

A preliminary look at the Dell forums found that about the same time I bought my machine, it was becoming wee know that there was a problem with the system itself. (http://en.community.dell.com/support-forums/desktop/f/3514/p/19370432/19844134.aspx)

E.g., one poster said they ”just spoke with a tech support supervisor who informed me that they are aware of the problem with the 580 series, and have a “back end team” working on a fix. It seems to be a RAM and/or motherboard problem that they hope to have a solution for by May 24.” And ‘‘Dell called today to follow up and told me it’s a “global problem” and now hope for a permanent fix in “three or four days.” One other person sugested that it may be a heat sink issue. (Which has me wondering if the Thermalright Archon CPU cooler might be worth getting.)

My first post to the Dell forum was about CPU-Z and other monitoring utilities reporting 2 volts vs. the max of 1.4 V. The first reply I got was that those utilties aren’t designed for Dell, so aren’t necessarily accurate. (I had tried SIV with the -DELL option, and it got the same readings as the others.)

(For those that don’t know, Dell’s BIOS is a stripped down version, so there are no readings or settings in the BIOS that let me see voltages or RAM timings.) I’m going too ask how I can determine the voltage, then. How would Dell do it?

I don’t know how to proceed to get “official” help from Dell. I’m within the warranty period, but don’t know if they have a real solution yet. I have some software licensed per machine tied to that system now, so I’m not going to want to send it back. What I’d like to know is what Dell would do, and let me do it myself. E.g., if it needs a new motherboard, send me one, and I’ll switch it myself.

[…] ran memtest for a whole week day and night. It threw 2 errors in that time. I replaced both memory sticks and it is perfectly solid…….

Memory errors can be very hard to find and I would not totally rule it out with only an overnight test run.

It seems there are two types of RAM problems, actually bad modules, and good modules in slots that just don’t work properly due to timing or other issues. (E.g., which slot the motherboard “wants” them, in what order, single vs dual channel running, timings that may be autodetected not quite correctly with no way to tweak them in the BIOS.

Did you have actual bad modules, or maybe replaced them with a higher quality or better matched set?

The general theme that has me wondering is that I don’t find reports of crashes happening while running the various stress tests that are out there. In other words, I don’t see someone saying that they were running Prime95 on its toughest settings, and had a crash then. It’s mostly that they were doing nothing demanding, and then it crashes.

Back in the days of NT4, flaky hardware wouldn’t let the system even boot. Windows 7 has services with delayed starting to help speed up boot times, and many Microsoft items in Task Scheduler running in the background, some doing things I would consider trivial, like keeping stats on how “reliable” the system is. (I can runAppCrashView and BlueScreenView to see how many crashes I’ve had.)

I remember reading in Scott Mueller’s books about the shift away from ECC to non-ECC RAM, which means that non-ECC RAM will have occasional errors. Maybe Windows 7 is extremely sensitive to those?

On the other hand, I do have an application that works fine under XP, but has problems with Windows 7, causing BSODs.

It may be that the problem in this specific thread is actually hardware based, but wanted to sound out the theme of mystery BSODs with no clear cause.

I’ve been setting up a Windows 7 system, duplicating what I have on my XP system, slowly and carefully, and have noticed this random unrelated BSOD problem, and wanted to post on this forum about it, after seeing the same theme over and over again on specific problems discussed all over the Internet.

It seems that many have have done some action, then get a delayed BSOD. Yet, if we boot again, and wait a while, if the system doesn’t BSOD within a few minutes, then it will stay running stable for quite a while.

The BSOD seems unrelated to anything we have done. People suspect viruses, replace RAM, video cards, motherboards, etc., to try to track down the “cause”. My suspicion is that some Microsoft-y something in Task Scheduler to “help” us may be the culprit. On my XP system, I never used Task Scheduler, to avoid having anything run without my initiation, to not sudden drops in system resources. Since Task Scheduler is a more integral part of Windows 7, and is useful to bypass UAC prompts for startup menu items, I am trying to learn to live with it.

Trying to research every mystery Windows 7 BSOD post on the Internet would take more time than I have, but wanted to raise the meta-question here that maybe there is something going on beyond our actions that cause these seemingly random BSODs.

For just avoiding the UAC prompt, I wonder if the program offered at “UAC Trust Shortcut-Bypass the UAC prompt while running a program” http://bestofnawaz.wordpress.com/2010/08/01/uac-trust-shortcut-bypass-the-uac-prompt-while-running-a-program/ would be more reliable and simpler to use than the whole MS ACT 5.6?

When I test the programs within ACT they work fine with no prompt, but the UAC prompt is back when I try to run outside ACT.

I’m having the same problem, and am wondering if that Matching Information screen info is somehow not matching what the UAC screen thinks it sees. The MS page http://technet.microsoft.com/en-us/library/cc721928(WS.10).aspx notes:

By default, the Compatibility Administrator automatically selects the basic matching criteria for your application. As a best practice, you should use a limited set of matching information to represent your application, as it reduces the size of the database. However, you must also make sure you have enough information to correctly identify your application.

I’ve noticed discrepancies between what the UAC screen says it wants to run and the detailled criteria found in the Compatibility Manager screen. Do we actually need any criteria to be saved in the database beyond just the file name?

I’ve also had it seem to work then after adding another fix, then not working.

The other thing I noticed, that needs further investigation, is that the BSOD I get is not while doing something, but at the logon screen before I log on. But, it seems that Window 7 doesn’t have a logon screen saver like XP, so I don’t know what is being attempted that is leading to a crash.

Now that I’ve had time to actually boot back into the Windows 7 system, there are more usb*.* files in system32drivers, than when I first looked, and the relevant ones have a local date of 3-24-11, which matches the UTC date of “25-Mar-2011” in KB 2529073. (I’ve seen the advice in passing that several re-boots might be needed for some changes to really take effect. This is a different world than XP or NT.)

Further investigating found Microsoft’s own “Steps to follow before you install Windows 7 Service Pack 1 from the Microsoft Download Center” (KB 2505743) where step 3 has three, not just two, updates to install before SP1: “Check Windows Update to see if you are offered the option to Install Windows Update KB2454826, KB2534366, or KB2533552”.

I didn’t look for or install KB 2454826.

The SP1 update itself appeared to work. The way I determined that there was a problem was comparing the dates of my USB drivers to the “25-Mar-2011” date shown in KB 2529073.

One thing I didn’t do was run the System Update Readiness Tools, as seen in Susan’s updated advice. Also, advice I’ve seen elsewhere suggests doing the update from a “Clean Boot.” I’m going to try these steps after I restore the backup I made before installing SP1.

If the issue is USB drivers, I wonder if the system can’t install the newer drivers by way of some Windows 7 update mechanism, as opposed to a special update?

There is one entry for WOPR 2003 itself, but not the individual programs, i.e., Word, Excel.

Trying WOPR’s own “Add/Remove WOPR Components” does remove the three items from Excel’s Tools menu, though.

The COM Add-Ins dialog box has no “Add-Ins available” listed.

The process that happened after I clicked on WOPR2003.exe on the CD.

It added things like:
HKCUSoftwareMicrosoftOfficeExcelAddinsEnveloperCOM03.dsrEnveloperXl
HKCUSoftwareMicrosoftOfficeExcelAddinsWOPRImageExtractor03.XlConnect
HKCUSoftwareMicrosoftOfficeExcelAddinsWOPRTaskPaneCustomizr03.XlConnect

And I can see them in Tools-Add-Ins-Automation.

Meaning that its one of those DLL Server type situations:
HKCR
+ CLSID
+ TypeLib
+ WOPRImageExtractor03.XlConnect

The usual startup directories show:

C:Program FilesMicrosoft OfficeOFFICE11STARTUP – empty

C:Program FilesMicrosoft OfficeOFFICE11XLSTART -empty

C:Documents and Settingsrd23Application DataMicrosoftAddIns
Name Manager.xla
XSFormatCleaner.xla

C:Documents and Settingsrd23Application DataMicrosoftExcel
Excel11.xlb

C:Documents and Settingsrd23Application DataMicrosoftExcelXLSTART – empty

Those are the ideas I was trying out before posting here.

What I’m trying to say, politely, is that *I* didn’t put the WOPR items in the Tools menu, so I don’t know what is needed to get them out.

The “usual” methods don’t seem to work — it is not the normal add-in that I have any control over.

(And the WOPR install routine didn’t present me with a choice if I wanted them in Excel or not.)

Being in Tools-Customize, then dragging the WOPR items off the Tools menu gets them off for the current session only.
When I exit and re-start Excel, they are back.

That’s why I was trying the tricks in the old 2001 thread about the matter, with only partial success.