Music from BIOS, 550W is STILL not enough power? What am I doing wrong?

[ASSVTLightning]

I visited this forum a long time ago when I encountered problems with my 3.3 V line (see thread here for background info).

My rig was originally configured like this:

- P4 2.4b o/c @ 2.9GHz, Asus P4PE, 512MB 2700 RAM @ 400DDR
- Gainward TI4600 Ultra 750 Golden Sample (o/c @ factory settings)
- 3 IDE ATA Hard drives (2X 7200, 1X 5400 RPM)
- 2 IDE DVD/CDRW optical devices
- 2 PCI cards (including ATA100 controller card)
- 3 case fans (excluding 2 fans on PSU, 1 fan on CPU, and 1 fan on video card)
- cold cathode tube
- 410W PSU with 2 Y-connectors to add additional 4-pin device plugs

I was getting low readings on my 3.3V line, so I swapped out the 420W PSU for a generic 500W unit. My 3.3V line recovered, but that PSU couldn't keep pace for long and eventually I replaced it with an Antec TruePower 550W PSU. Smooth sailing for a year... UNTIL NOW.

I replaced the dying 5400 RPM HD with a new 200GB 7200 RPM HD (also IDE) and everything is cool for about a week. Then one morning I boot and I hear this ringtone melody from the mobo speaker. It's not a typical POST error beep, it's an actual melody of 3 beep tones in rapid succession, repeated every 10 seconds or so. I swear I thought it was my cell phone when I first heard it. Then my heart stopped when the error message "NTLDR is missing. Please press ctrl-alt-del to continue" appeared. If anyone else has ever troubleshot this problem before, you know how painful it can be. I used the recover console to reload the boot files, and finally the computer booted to Windows XP Pro. I thought I was done. Warm boots didn't yield any beeps. But a cold boot brought back the beeps (even though I could still boot to Windows XP fine without the NTLDR error). I haven't been able to identify what the Award BIOS beep melody means, and the beep code legends I find on the web are way simpler than the one I hear. So now I'm really frustrated. Finally I read somewhere that sometimes the cause of the NTLDR error message is just a loose or faulty IDE cable. While changing IDE cables I did some experiments:

- When I booted with only the boot hard drive connected (directly to the mobo IDE channel 1), I could boot normally without any beeps or errors.

- When I booted with 2 drives on the same mobo IDE channel, the beeps went away and I could boot normally without any beeps or errors.

- When I booted with 1 drive on the mobo IDE channel and 1 drive on the Promise ATA100 controller, the beeps went away and I could boot normally without any beeps or errors.

- When I booted with all 3 drives (2 drives on the mobo IDE channel and 1 drive on the Promise ATA100 controller), I get the funny beeps again and BAM! The "NTLDR is missing" error appears again! As soon as I take aware the 3rd HD, the computer is fine.

- Just for chits and giggles, I switch 2 of the HD's from IDE to ATA controller card and same problem.

So does this mean my power supply doesn't have enough juice to run three 7200 RPM HD's? I must have only had barely enough power to run 2 7200 RPM HD's and 1 5400 RPM HD for a year if that's the case. I just don't understand why my computer is demanding so much power... I don't think 3 ATA drives is excessive (think of all the raid and SATA setups with 4+ HD's). Is the overclocking hogging all my juice? Could my house's physical power outlet be faulty?

In case it matters, here's what my Motherboard Monitor says when I only have 2 HD's connected:

CPU Temp: 37 deg C
Mobo Temp: 20 deg C
CPU Fan: 2360 RPM
Core: 1.68 V
+3.3V: 3.3 V
+5V: 5.00 V
+12V: 11.37 V
-12V: -11.88 V
-5V: -4.98 V

Is it my 12V line that's flaky here? My temps (CPU fan seems to spin ok?) Any help at this point would be greatly appreciated... I am at wit's end already.

 

[MercenaryForHire]

Clock down to stock speeds for starters, and put a UPS between you and the wall. It could also be your drive temps. Google "DTemp" and see if your drive is supported.

- M4H

 

[Big Lar]

If in fact Mobo Monitor is reporting the correct +12v, it is too low, tho to believe a software is hard to do. A volt meter check would be good. Also Digitaldoc has a plug in with a digital readout for voltage.

 

[mechBgon]

edit: what he said ^

Also, use a digital multimeter to test your 12V line. Software is not necessarily accurate, or even relevant with regards to the PSU in particular. Your PSU may be delivering precisely 12.00V to the end of its cables... what the motherboard does from there on is not the PSU's problem. Eh?

A 550W TruePower is certainly enough power for your rig. If you aren't getting any progress with the CPU and RAM set to their stock speeds, try bumping the memory voltage to 2.8 volts as a troubleshooting step (if it isn't there already), and use Memtest86 to see what kind of results you get when testing your RAM. If you're fortunate enough to have a spare motherboard, or have been waiting for an excuse to get a new motherboard, try a new mobo Good luck.

 

[Gerbil333]

Your +12V line is dangerously low. I'd inspect that if I were you. Try the PSU in a different computer if you can. There's no reason your rig should eat up that many watts.

I thought my K7N2G-L was making a POST beep error when I put a new hard drive in it (and not just a few beeps -- a lot of beeping), but the hard drive turned out to have a speaker in it to report its own errors...I replaced the drive, and no more problems. The new hd was a Seagate Barracuda 7200.7 btw. I was only replacing the old drive, a Maxtor, to make the computer quieter.

It sounds more like something is wrong with your PSU unit though.

 

[ASSVTLightning]

I'm about to test turning the wick down on the CPU to see if that frees up some juice. I'll also try that memtest86 too. Voltmeter says 11.8.-11.96 depending on which plug I test, so I guess it is a tad low. DTemp show both drives hovering around 31 deg C - is that ok?

If it turns out to be not enough power? What are my options?

1) Just run 2 drives
2) Run 3 drives but without overlocking CPU/memory
3) Remove other devices?
4) Other?

Is there such a thing as running a 2nd PSU?

Is my TruePower 550 defective? Or am I overloading the system the way it is? Doesn't seem like that crazy of a rig to me, but what do I know.

 

[Overkast]

Maybe the PSU is defective. But I have the same PSU and don't have any problems with it (yet).

Here's my system specs:

P4 2.4c @ stock speed
1GB Mushkin 3500 L2 RAM
Intel D875PBZ mobo
2 SATA WD 36GB Raptor HDs
Radeon 9600Pro 128MB
Antec TruePower 550w
DVD-ROM
CD-RW
5 case fans (8 total including CPU and 2 PSU fans)

I think you should start off by not OC'ing your CPU and see if that makes a difference.

Do you think there's a problem with the Master/Slave jumpers on the back of your Hard Drives?

 

[Markfw]

OK, here is a real shot in the dark... Maybe your motherboard doesn't like all of those drives. I know it's a longshot, and you probably don't have the money to replace it, just to try. If turning off the OC doesn't help, then ????

 

[ASSVTLightning]

Thanks for the suggestions and ideas folks - keep em coming. I don't have another mobo to test here, but what did you mean my mobo might not "like" the drives? Jumpers on HD's seem correct (they detect and show up correctly, although the drive letter assignment in Windows XP show the slave HD on the mobo IDE channel AFTER the 2 partitions from the HD on the ATA100 controller card, which I thought was odd).

I just booted at "stock" CPU speeds and Vcore voltage, and believe it or not, my voltage readings are all the same! They fluctuate as usual, but within the same ranges I saw when it was overclocked.

I was trying to add up my power usage, because I have a "max" of 24 amps available for my +12V line (according to the label on the PSU). I borrowed this table from another thread, just to plug in some basic numbers, but I'm not sure what to estimate for the HD amps:

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 3.3V _ _ 5V _ _ 12V
Motherboard _ _ _ _ _ _ _ _ _ _ 3A _ _ 2A _ _ .3A
Fans (some actuals) _ _ _ _ _ _ 0A _ _ 0A _ _ 2.39A
Memory (actuals) _ _ _ _ _ _ _ _ 0A _ _ 6A _ _ 0A
Video _ _ _ _ _ _ _ _ _ _ _ _ _ _ 3A _ _ 0A _ _ 0A
Sound _ _ _ _ _ _ _ _ _ _ _ _ _ .5A _ _ .5A _ _ 0A
Hard Drive - IDE (actuals) _ _ _ 0A _ _ _ <------ ????
CD-RW (actuals) _ _ _ _ _ _ _ _ 0A _ _ 1.5A _ 1.5A
DVD-ROM _ _ _ _ _ _ _ _ _ _ _ _ 0A _ _ 1.5A _ 1.5A
Floppy _ _ _ _ _ _ _ _ _ _ _ _ _ _ 0A _ _ .8A _ _ 0A
KB & Mouse _ _ _ _ _ _ _ _ _ _ _ 0A _ _ .75A _ 0A
Processor _ _ _ _ _ _ _ _ _ _ _ _ 0A _ _ 10A _ _ 0A

Totals _ _ _ _ _ _ _ _ _ _ _ _ _ 6.5A _ _ 14.05A _ 15.69A

Can anyone tell me what Amp value I should be using from this spec sheet? I see a couple values but not sure which to plug in:
Hitachi Deskstar Data Sheet

Same for this, what value do I use? The 662 mA?
Maxtor DiamondMax Plus Data Sheet

 

[Bozo Galora]

Has nothing to do with your PSU. NT (boot) loader is corrupted or missing.
Maybe a virus, or your switching HDD around musical chairs, or XP lost track of all your crap.

link

12V has plus or minus 10% by spec or 10.8 as low
but its prac impossible to get valid reading under complex load

maybe one of the controllers on your HDD's is getting flakey, so XP boot cant access for its bootloader portion.

or FAT32 NTFS conflict, or you recently used Part Mag or Ghost

 

[Gerbil333]

In that case, can you try all the drives in another machine?

 

[ASSVTLightning]

Originally posted by: Bozo Galora
Has nothing to do with your PSU. NT (boot) loader is corrupted or missing.
Maybe a virus, or your switching HDD around musical chairs, or XP lost track of all your crap.

link

12V has plus or minus 10% by spec or 10.8 as low
but its prac impossible to get valid reading under complex load

maybe one of the controllers on your HDD's is getting flakey, so XP boot cant access for its bootloader portion.

or FAT32 NTFS conflict, or you recently used Part Mag or Ghost

I spent all week researching and fiddling with the "NTLDR missing" error, including most of the suggested fixes on that thread you linked me too (I found that at the top of the Google heap earlier this week, another GREAT resource!). A lot of the suggestions in that thread don't apply to me, as I wasn't trying to upgrade one OS to another, my boot drives have always been NTFS (I do have one data drive I converted from FAT32 to NTFS a long time ago), and I'm not doing any crazy dual boot scenarios. All I did was: I bought a new (Hitachi) HD, used MaxBLast to copy my old HD partiton to the new HD, switched boot order by jumper (these are on the same mobo IDE channel), booted with new HD, then wiped out old boot HD and turned it into a data drive. Maybe MaxBlast didn't like my Hitachi and screwed up the ghost? or maybe it's XP getting confused by the boot drive swap?

Other notes: My BIOS has virus protection on the boot sector, and NAV Pro 2003 didn't find anything when I did a manual scan 2 days ago. When I first loaded the ntldr and ntdetect.com files from Windows XP Recovery Console, I honestly thought those files were lost or corrupted. But now I see that I can predictably get the "NTLDR missing" error as soon as I connect all 3 HD's back up (suggesting a hardware problem). This idea is reinforced by the fact that as soon as I take away one of the hard drives, the PC boots like a champ (without me having to restore the ntldr and ntdetect.com files!). I don't see how the physical data could change/corrupt itself with the addition of a hard drive, then fix itself when I take that drive away again. And it doesn't really seem to matter which drive I take away, it really seems like a 3rd drive is "pushing it over the edge."

EDIT: As soon as I thought I had it all figured out, now my PC is booting fine overclocked with all 3 drives again! But I have another computer plugged into a different power strip on the same wall outlet, and it's currently off so I don't know if that's why my PC is fine now. I may get a UPS today (any affordable models that you recommend?), I'm still leaning toward a hardware or PSU problem. I do have another older PC I can test the drives in - it only supports ATA66 though. I'll keep posting updates to my situation, but don't hesitate to post more suggestions for me to try too.

 

[mechBgon]

APC SmartUPS Pro 620VA comes to mind as a good one with automatic voltage trim/boost and the PowerChute auto-shutdown software & serial cable so it can gracefully save your stuff and shut down your system if the UPS is going to run out of power and you're not there.

 

[ASSVTLightning]

OK dumb question... I'm researching UPS info now, but why would a UPS help me in this situation? Do these clean up or condition the power coming into my PC? Because I honestly don't care for the backup capability (and I already have a high-rated surge protector).

 

[mechBgon]

Heh heh... maybe when I get to work Monday I'll clip some of the power sag/surge events from PowerChute's logfiles to help answer why you'd want an AVR-equipped UPS Is it the cause of the trouble here? probably not.

Are some/all of your drives Dynamic Disks in Windows Disk Management, or have they been in the past? I know that Dynamic Disk can do interesting stuff... I reformatted my primary hard drive at work and reinstalled Windows on it, but in Disk Management it still lists its old self as... now what did it call it... "offline" or "failed" or some such. I'm not sure which of my two drives is giving Windows this flashback, since they're both Dynamic, but that's an example of some Dynamic funkiness. In my case, it doesn't cause a problem.

 

[ASSVTLightning]

I'll have to get back to you on the Disk Management, I'm still in UPS-research-mode right now. (Boy this thread is taking on a life of its own.)

Let me elaborate on my power environment. I have a townhouse and use the first floor as my office. I have 4 power outlets (2 sockets each for 10 available sockets) in my office space. I have the following items plugged in throughout the room:


OUTLET #1:
- 2 surge protector power strips
- 1 PC
- 1 laptop and dock
- 1 21" CRT
- 1 19" CRT
- 1 Altec Lansing 4.1 computer surround system (the subwoofer is as big as my desktop chassis)
- 1 pair harmon/kardon stereo computer speakers
- 1 printer
- 1 scanner
- 1 cordless phone
- 1 cable modem
- 1 wireless router
- Other USB peripherals like webcam, card reader, etc that don't have their own power source

OUTLET #2:
- 1 laptop
- 1 paper shredder

OUTLET #3:
- 1 halogen floor lamp

OUTLET #4:
- 1 20" TV (CRT)
- 1 home alarm system power brick thingie


I notice that whenever I turn on the TV, my 21" CRT flickers once then recovers. Whenever the washer/dryer kicks on nearby, the halogen light dims briefly. Whenever I run the paper shredder, the computer speakers crackle with noise. Also, when I try and recharge some Ni-Mh batteries in outlet #2, I notice they never charge to full capacity or stop charging altogether, and I have to plug them into the bathroom outlet down the hall to get a full charge. And... maybe the clincher here... if I am doing some heay duty HD work (like accessing/copying files between 2 drives and winrar'ing at the same time) while listening to MP3's, I notice my Audigy 2 sound card drops some audio segments and my music gets clipped every couple millisecond until the HD's are done working. But this only happens when I'm really straining the drives, not in normal everyday usage.

So does this sound like I am overloading any circuits? Would a UPS with line conditioner and voltage reglation help? I thought my surge protectors already had built-in noise suppression and line conditioning, but maybe not as good as a UPS would offer.

Do I need something like this too?
APC Line-R

 

[mechBgon]

On a tangent regarding the audio skipping: out of curiosity, what PCI slot is your Audigy2 in? For the P4PE, the "sweet" slots look to be PCI slot #1 or PCI slot #5, both of which share an IRQ only with eachother, so put the Audigy2 in one of those slots if it isn't already, and leave the other one empty.

edit: and yeah, the AVR Trim/Boost ensures that your system's power is within a safe window. You can judge for yourself once I get a snip of those logfiles.

 

[ASSVTLightning]

Re: the tanget, my Audigy 2 is in slot 4 right now, with 1,2,3 available, and the Promise ATA100 Controller card in slot 5. I may move things around and see what happens. How did you come to find out about the P4PE "sweet spots?"

Can't wait to see your logfiles.

 

[mechBgon]

Originally posted by: ASSVTLightning
How did you come to find out about the P4PE "sweet spots?"

I gots over 100 mobo manuals in PDF format for popular mobos (so far)

 

[ASSVTLightning]

Wow!

BTW, I see CC has a $50 rebate on the APC BX 1000VA. Is this model any good (or good enough for what I need)?

APC BX1000

The only negative I see to this one is that it doesn't have broadband network protection (only phone line protection looks like). The RS1000 does the broadband I think.

 

[mechBgon]

The automatic voltage regulation is what I think you want, since it doesn't wait for the power to go out... it smooths out the sags and bumps for you. So lessee... does the BX 1000VA have AVR... sweet, it does! And with a 1000VA conversion circuit, that shouldn't have a problem with your system + two big CRTs either. That looks like a good buy to me.

 

[ASSVTLightning]

Cool, I'll pick one up tonight maybe, unless anyone thinks I should try and find one with the RJ45 support (vs the RJ11).

 

[mechBgon]

Originally posted by: ASSVTLightning
Cool, I'll pick one up tonight maybe, unless anyone thinks I should try and find one with the RJ45 support (vs the RJ11).

It's strange, but the RJ45 support is getting tough to find in APC's lineup My older BackUPS Pro 1000VA has it but new ones seem not to. Not sure what's up with that.

 

[mechBgon]

Okie dokie, here is some PowerChute log from one of our servers:

100000 7/16/03 10:08:41 *** PowerChute PLUS Version 5.2.1 started ***
100200 7/16/03 10:08:48 Communication established
200003 7/16/03 10:52:12 UPS on battery: Blackout
100300 7/16/03 10:52:54 Normal power restored: UPS on line
100000 7/16/03 12:57:15 *** PowerChute PLUS Version 5.2.1 started ***
100200 7/16/03 12:57:23 Communication established
100100 7/17/03 12:04:43 *** PowerChute PLUS Version 5.2.1 stopped ***
100000 7/17/03 12:08:13 *** PowerChute PLUS Version 5.2.1 started ***
100200 7/17/03 12:08:20 Communication established
100403 7/17/03 12:39:40 Self-test at UPS passed
200006 7/27/03 02:38:07 UPS on battery: Deep momentary sag
100300 7/27/03 02:38:07 Normal power restored: UPS on line
100403 7/31/03 12:37:58 Self-test at UPS passed
100100 8/1/03 11:07:06 *** PowerChute PLUS Version 5.2.1 stopped ***
100000 8/1/03 11:10:19 *** PowerChute PLUS Version 5.2.1 started ***
100200 8/1/03 11:10:26 Communication established
200700 8/6/03 01:25:38 UPS enabling SmartTrim
200006 8/6/03 04:34:00 UPS on battery: Deep momentary sag
100300 8/6/03 04:34:00 Normal power restored: UPS on line
200006 8/6/03 04:36:48 UPS on battery: Deep momentary sag
100300 8/6/03 04:36:48 Normal power restored: UPS on line
100403 8/14/03 12:36:46 Self-test at UPS passed
100403 8/28/03 12:35:03 Self-test at UPS passed
200007 9/8/03 04:59:44 UPS on battery: Large momentary spike
100300 9/8/03 04:59:44 Normal power restored: UPS on line
100403 9/11/03 12:34:26 Self-test at UPS passed
200700 9/21/03 00:24:34 UPS enabling SmartTrim
200700 9/21/03 21:15:08 UPS enabling SmartTrim
100403 9/25/03 12:31:03 Self-test at UPS passed
100403 10/9/03 12:29:43 Self-test at UPS passed
200700 10/14/03 00:19:53 UPS enabling SmartTrim
100403 10/23/03 12:28:19 Self-test at UPS passed
200006 10/28/03 14:53:09 UPS on battery: Deep momentary sag
100300 10/28/03 14:53:09 Normal power restored: UPS on line
200006 10/28/03 14:53:15 UPS on battery: Deep momentary sag
100300 10/28/03 14:53:15 Normal power restored: UPS on line
200700 10/29/03 00:02:48 UPS enabling SmartTrim
200700 10/29/03 00:37:24 UPS enabling SmartTrim
200700 10/31/03 18:36:29 UPS enabling SmartTrim
100403 11/6/03 11:26:57 Self-test at UPS passed
200700 11/7/03 23:51:54 UPS enabling SmartTrim
100403 11/20/03 11:25:37 Self-test at UPS passed
200700 11/22/03 09:29:42 UPS enabling SmartTrim
100403 12/4/03 11:24:19 Self-test at UPS passed
100403 12/18/03 11:23:01 Self-test at UPS passed
100403 1/1/04 11:21:42 Self-test at UPS passed
200006 1/5/04 05:46:17 UPS on battery: Deep momentary sag
100300 1/5/04 05:46:17 Normal power restored: UPS on line
100403 1/15/04 11:20:18 Self-test at UPS passed

 

[ASSVTLightning]

OK, I picked up an APC Back-Ups XS 800 with 540 watt support, thinking this would fix my problem (since it has automatic voltage regulation). Well it hasn't fixed the problem (and might be aggravating it) but at least the Parachute software is offering me additional clues. Here's what I observed:

1) I charged up the unit for 8 hours (per the included instructions) in another room on a different circuit.

2) I plugged the UPS into my computer's power strip and moved only the monitor and computer's power cables over to 2 battery-backed outlets on the UPS.

3) With those 2 devices plugged in, I booted up, installed the Parachute software, performed a self-test, and monitored the system's performance. With only the 2 devices plugged in (21" CRT and overclocked computer w/ 550-watt PSU) it said I was only drawing 270 watts. Well under the rated 540 watts of the UPS and 550W output of the computer's PSU. I have have observed no "power problems" for the past 2 days, and my voltage has been a steady 120-122 V. I have electrical noise sensitivity on it's highest setting too. While I realize 2 days is not enough data to base any conclusions, a couple things happened that are worth mentioning:

First, I have cold-booted 3 separate times since plugging my PC into the UPS. 2 of the 3 times, I got the BIOS music and it failed to detect the hard drive on the PCI ATA100 controller card. So adding the UPS in-line made my problem worse somehow (but I'm not sure why as it's conditioning the signal and regulating the voltage!!).

The second thing I observed (thanks to DTemp) is that the same hard drive (which has been operating perfectly for over a year) just gave me a SMART failure/warning message and when I click on the details it shows a problem with the "Spin Up Retry Count." The values shown say 222 when the threshold is 223. I didn't know what this attribute means so I looked it up:

Count of retry of spin start attempts. If the HDD can't start its spindle on the first try (to make HDD work) it makes another try - and so on while the spindle will not rotate normally. This attribute stores a count of start retries.

So what does this all tell me? The whole reason I purchased a new hard drive was because one of mine started clicking one day and failed. Now it looks like I'm slowly killing off another hard drive! Why can my system only handle 2 IDE hard drives? Shouldn't my 550W Antec TruePower PSU be able to handle all 3 of my drives? Does everyone agree that my +12V reading of 11.37 V is probably the culprit? What can I do to boost or help that line? Or should I be looking at something else?