Insufficient current to power strip?

[drewp29]

So I have an e8400 running on a Gigabyte x38-ds4 motherboard that I have had overclocked to 3.92GHz for at least 11 months now, and everything has been fairly normal with the overclock (i.e. prime95 stable, good benchies, etc.). Now, yesterday I received a new dual external HD enclosure, and I added 1 out of 2 hard drives to the enclosure and everything proceeded without a hitch. That is, until I tried to format the drive. The OS (xp x64) gave me the message: This drive cannot be formatted because Quick Disconnect is enabled. And it gave me the option of canceling or disabling QD. So I disabled QD thinking I would just enable it after the format. But when I went to enable QD (Drive management, policies tab) the options for Write Caching and QD were grayed out. So I installed Intel Matrix Storage manager (ICH9R chipset), which required a reboot, and upon rebooting, I can no longer get past the bios memory check with the system overclocked. However, if I turn the new external drive off, I can boot normally (OC'd) and the computer is fine. If the external is on, I get random blue screens upon which the reboot will power down and reset my OC settings to stock.

So, I think I have determined that the computer needs to be plugged into a wall socket instead of the power strip since I have a 500GB external, a 250GB external, the new enclosure (1TB - 1 drive, second drive will arrive today), and the monitor (19" WS) plugged into the same power strip as the computer (1000W Antec Quad PS, 80GB OS drive, (2) 750GB Samsungs, and (2) 1TB Samsungs).

Does this sound plausible, or has anyone else experienced something like this? The power strip is a decently good one, but I think it is at its limit for current supply - maybe? Any thoughts?

 

[Jim_L9]

I think your OC has affected the port you are using for the new dual external HD enclosure. The power strip should be good for up to 15 amps and your power supply should be able to deal with any normal voltage drop caused by the load of your PC.

 

[drewp29]

Its a USB 2.0 external, as are the other two . . . any reason why the new enclosure would be different than the other two? Any suggestions on how to fix it? Thanks for the reply!

 

[terr281]

I would have to agree with your hypothesis over Jim's...in that the strip, the single plug of the outlet, or the single outlet itself is the issue.

Further, this is the easiest problem to check for. See if the problem still exists if you...

1. Plug the computer into the other outlet plug.
2. (By using an extension cord) Plug the computer into another outlet somewhere in the room.

If the problem still exists, then power isn't the cause. But, we have had similar circumstances occur with computers in our business environment. (Too many things in one strip, computer not behaving properly, plug the computer in a different outlet and it works fine.)

 

[drewp29]

Thanks terr281, I'll check this on my lunch break . . . then I'll post the result.

 

[drewp29]

Didn't get to it on the lunch break, it'll have to wait until tonight.

 

[Jim_L9]

When you did the OC did you lock the PCI and PCIe buses?

 

[drewp29]

The Gigabyte Bios has the option to manually set the PCIe frequency, so I set it to 100MHz . . . I have played around with setting it higher while adjusting the corresponding voltage with mild success on graphics benchmarks, not enough to warrant the stress on the video card. However, I have read a few other places that the Gigabyte pci bus speed may not be fully locked . . . that's a good thought, I'll have to check it in Everest when I get home. Thanks! I would think since there is no option in the bios that it would be locked, since they tout the Gigabyte boards as overclockers. A good thing to check regardless.

 

[drewp29]

So I looked for a long time for a hardware monitor that would show me the current pci and pcie bus frequencies, to no avail. Clockgen was the best bet, and I need to look at my MB to determine my PLL id to find a copy that will work. CPUID's website says clockgen will be back soon, and gives no more info. If anyone knows this, and what version of clockgen to use, please let me know.

At any rate, I determined it is not the power strip, so I looked to other solutions. I disabled the legacy usb detection in the bios, after updating the bios from F3 to F4e (a beta bios . . . I know, I know), and it allowed me to boot with the new external powered up. Eureka! Or so I thought. I then found that the first few minutes of use in Windows was fine, but after accessing several drives, the OS became sluggish and explorer started using upwards of 70% cpu usage constantly. I have never had more than 30-35% usage for any normal Windows activities (unless in a game, etc.). So I thought the Intel Matrix Storage Manager might be the culprit, uninstalled, rebooted, same issue arose. So I went back into the bios and enabled legacy USB detection. Still the same issue on overclocked booting, but turning the new external on after windows initialization didn't seem to give me issues like before regarding stability of the OS. The OS was still sluggish after a few minutes of activity, and CPU usage >70% again.

So, I thought, screw the beta bios, maybe it's the problem. This morning I replaced F4e with F3, and upon boot, the LOGO SCREEN came up?! I haven't had that option enabled since I got the system, so I chalked it up to a full reset of my motherboard setting. I went into the bios and reset everything to normal overclocked settings, reset, and the computer turned on and immediately off. So now I have to reset the CMOS with the jumper, and start from scratch.

The weirdest thing is what the default bios settings were after the logo screen cam up. The cpu clock frequency was set to 100MHz x 9.5x multi (stock is 333 x 9), and the CPU voltage step variations (amount between voltage amounts when raising or lowering voltage) were not even the same. The voltage changes were .0(0)5 steps instead of the usual .0(0)25 steps. Anyone else ever had this happen? I am positive I flashed with the correct bios, its the same exact disk I used before.

So the saga continues - most difficult addition of an external hard drive EVER!

Edited: should have ben .005V and .0025V steps in bios

 

[drewp29]

Its just a freaking USB 2.0 HD, not anything crazy like SCSI or SAS or ieee1394, I mean (in the words of Gob Bluth) - Come On!

 

[drewp29]

Reset CMOS, Overclocked, everything is good, now just to figure out the external problem. Anyone have any ideas? Help would be appreciated, thanks in advance, and sorry for the extremely long posts . . .

 

[terr281]

IMHO, long posts that describe the problem and the already attempted solution are better than short ones...

With it not being power, I would try <no particular order>: (Pulling at straws, btw...)

1. Disconnect the other USB drives and connect the new one. Problem still there?

2. Are any of the drives prior to the new external (especially the internal ones) in a Raid configuration? (Could the new drive changed something with your raid setup via auto detect, etc... then installing the Intel Manager...caused the Intel Manager and something else to constantly try to acquire the same resources...the raid drives.)

3. Windows needs some particular files for bootup from one of your external drives, and when the new one is connected, your MB's USB throughput is maxed out. Therefore, Windows cannot get it/get it accurately, thus the non-bootup and cpu utilization.

4. Something wrong with the connection between the computer, new enclosure, and hard drive?

a. Remove hard drive from enclosure, put the new hard drive (you mentioned in first post) in the same slot. Does it work?
b. Put the 1st hard drive in the other slot in the enclosure. Does it work?
c. Try a different USB cable between the computer and the enclsoure.
d. Plug the enclosure into a different USB port. (Preferably one that you know works from one of the other drives.)
e. Try the original hard drive in original slot in enclosure with the same cable in a different computer. Same issue?

...That's all I can think of right now.

 

[drewp29]

Thanks again, I have some ideas to try now . . . no raid configs in my system, but there are some files that windows creates on one of my external drives (the larger of the two) every time I install Windows (about every 6 months, just because, well, just because its Windows). I usually segregate these to a folder so I don't have to look at them every time I explore the drive, and I've never seen this cause any issues, but perhaps since I am adding a new storage device that it hasn't previously dealt with it needs these files again. Hmmmmm, maybe.

Otherwise, I will try a different USB cable, in a port I know is in working order, and then I'll try the drive in my other machine to test whether its computer or enclosure problems. If all else fails, and it seems to be the computer's problem I'll reinstall the OS (its about time anyhow).

I had uninstalled Intel's Storage Manager because the slow down started after I installed it, but maybe windows is still utilizing certain files that did not get removed . . . a reinstall of XP may well be in order.

Thanks again, I'll let you know what I find out.

 

[drewp29]

So, I think I have it - there seems to be some sort of addressing or priority problem between my new external drive and my old 250GB external. If I have the 250GB off the OS is like butter, but as soon as I turn on the 250GB drive and peruse the files on both the new and old external drives, the CPU overusage starts and continues even if both externals are disconnected and off. I don't know if it is hardware or software, and I can't check it with my other computer because its running XP 32bit instead of x64 and the hardware is totally different. So, I guess I'll redo my OS at some point, probably over the weekend and start fresh to see if it is a symptom of Intel Matrix Storage Manager or just my MB. Thanks terr281 and Jim_L9 for the input. If you have any ideas for how I can solve the conflicting externals (I don't know if this is an IRQ issue or a PCI bus addressing issue, probably the latter) and I don't know how to manage the pci addressing. I do know how to manually set IRQ addresses while loading Windows, which can be a huge mind drain - I wish you could still do it in the OS, but alas, M$ decided we don't need this function anymore. I could rant, but I won't . . . again, thank you very much.

Edit: Psyche!!! I found it . . . the REAL problem. One of the programs I got to do hard drive benchmarking, from downloads.com no less (although I don't think its the actual file's fault) was STUCK on the desktop. The file could not be deleted and access was denied if I tried to change any properties. So while I thought it was a hard drive issue, it was just that after so many times windows explorer tried to refresh the desktop without being able to refresh that file it would go into an endless loop. So I booted to safe mode and it allowed me to delete the problem file and the OS is fine now. The thing is, I would have never found this out if I hadn't decided to clean up my desktop and get rid of random icons cluttering things up . . . you live and learn I guess.