PatriciaDooley

Distinguished
Feb 1, 2010
11
0
18,510
Hi, I have scoured your site trying to figure out two questions. This question has to to with random shut downs. The symptoms are that the computer suddenly shuts down for no apparent reason. I do not get a blue screen or anything. It is as if I turned it off. When I restart the computer all is normal. I cleaned out the fans and other dusty areas already. It mostly occurs when I am playing WoW. But not always, sometimes I can do WoW for HOURS(and never have a problem) other times, just minutes. Most of the time I can surf the web with no problem, but once in a while it just shuts down too. I believe, from your forums, it is the Power Supply, BUT, my temps are not running abnormally hot. I did finally find a discussion on voltages. I don't mind messing with my computer to figure things out, but am a little timid to stick a voltimeter into the comp(Afraid I will short the comp out). I did download 2 programs that were suggested on this site--HWMonitor2 and Everest Home Edition. Below is an abbreviated report from HWM. Please note that the Voltage for the +12v is 14.61 and I can get it to hit 14.70 without the comp shutting off. I also noticed that the +5 is running at 3.49. All of the voltages are staying consistent to what is listed below (whether they are high or low or normal). With the most fluctuation in the +12 (+/-.5degrees variance).

I have also heated the GeForce video card to 80C but the power never turned off there. (I know they run hot). The other temps stay pretty regular to what you see below.

I could also flash the BIOS, but again am timid to try that also. My BIOS are standard settings. I am not over clocking (not that I would know how to), and I don't believe, underclocking. MSI does has an online BIOS program, but it flashes and if that is not the problem and the comp shuts down, i could cause some serious problems!

Any ideas?

Thanks for all the work you do in answering these. They helped me figure out my monitor problem (won't stay on but light stays green). Thanks to pics you all posted and the various theories, I figured out it was the LCD itself going bad (would stay on after it "warmed up") (versus the capacitors). Both problems did start about the same time. I did switch out the LCD with a VGA monitor until I can buy a new one, but the power still shut off today after changing them out.

Thanks in advance for your help.

I am running windows xp sp2, with an MSI K9N Neo v1.0 mainboard. GeForce Nvidia 8600GT video card

HWMonitor version 1.1.5.0
Mainboard Model MS-7511

Hardware monitor Fintek F71882F
Voltage 0 3.33 Volts [0xD0] (+3.3V)
Voltage 1 1.34 Volts [0xA7] (CPU VCORE)
Voltage 2 2.21 Volts [0x8A] (VIN2)
Voltage 3 1.43 Volts [0x7A] (VIN3)
Voltage 4 3.49 Volts [0x53] (+5V)
Voltage 5 14.61 Volts [0xA6](+12V)
Voltage 6 1.42 Volts [0x59] (VIN6)
Voltage 7 3.33 Volts [0xD0] (VSB3V)
Voltage 8 3.14 Volts [0xC4] (VBAT)

Temperature 0 31°C (87°F) [0x1F] (TMPIN0)
Temperature 1 38°C (100°F) [0x26] (TMPIN1)

Fan 0 3145 RPM [0x1DD] (FANIN0)
Fan PWM 0 75 pc [0xC0] (CPU)
Fan PWM 1 70 pc [0xB2] (System Fan 1)
Fan PWM 2 50 pc [0x80] (System Fan 2)
Fan PWM 3 50 pc [0x80] (System Fan 3)

Hardware monitor AMD Athlon 64 X2 3800+
Temperature 0 35°C (94°F) [0x54] (Core #0)
Temperature 1 28°C (82°F) [0x4D] (Core #1)

Hardware monitor NVIDIA GeForce 8600 GT
Temperature 0 62°C (143°F) (GPU Core)

Hardware monitor ST3200827AS
Temperature 0 37°C (98°F) [0x25] (Assembly)
Temperature 2 37°C (98°F) [0x25] (Air Flow)


 
If the 5 volt line were running at around 3.5 volts, I doubt if the computer would boot. Likewise, the high 12 volt reading.

A simple - and safe - way to check the 12 volt output is to measure between the yellow and black wires on a power connector for an IDE drive.
 

PatriciaDooley

Distinguished
Feb 1, 2010
11
0
18,510
I checked the BIOS but could not find any reading for the +12v. I found the +3.3v and it was running right at that. but I saw no +5v or +12v. With the +3v there were temps and other items, but not other voltages labeled as such. And no where for me to expand the information. I had the choice to change the cpu temp and the fan speed, but those were the only two options on the same page that gave the temps and voltages.

If you don't think I will make a mistake with the voltimeter, please describe (in very simple layman terms) how to do it. My husband can bring one home from work.

Thanks again for helping me.
 

Paperdoc

Polypheme
Ambassador
BEFORE you do this, try using the BIOS Setup screens to find the info. The manual I looked at on page 3-23 shows where the info is found. From the Main Menu of BIOS Setup, cursor to the right to select H/W Monitor and hit Enter. On that menu go to the last item, PC Health Status, and hit Enter to see voltage readings.

jsc pointed to a simple and safe way to check the voltages with a voltmeter. Your PSU (power supply) box has many groups of wires coming out of it. Among them will be several sets of four wires ending in a 4-pin female plastic connected called Molex. It is about ¾" wide and looks like this:

http://www.electrical-res.com/EX/10-18-22/Molex-style-power-connector.jpg

It is commonly used to supply power to IDE hard drives and some optical drives. In it, both black wires are Ground. The Red wire is the +5 VDC line; the Yellow wire is the +12 VDC supply.

Disconnect the power cord to your computer case and open it up. Look for one of these 4-pin Molex connectors, preferably one that is just hanging there unused. Pull it out a bit so you can reach it easily. Leave the cover off, reconnect the power cord, and boot up.

Now you can use the pin probes of a voltmeter to check this. Usually the meter has a red probe for the + signal, and a black one for the - signal. Connect meter - probe to either black wire's little metal socket in the connector, and the meter's + probe to the yellow wire's socket. That should be close to 12.0 volts. Move the meter's + probe to the red wire's socket and check to see how close that is to 5.0 volts. Remove the probes, shut down, disconnect the power cord, and re-arrange the wires inside so you can close up and reconnect power.
 
  • Like
Reactions: capturer

PatriciaDooley

Distinguished
Feb 1, 2010
11
0
18,510


I looked at my BIOS again. I don't have what you say I should have. I even looked in my manual (p. 3-21) and the page that I should be looking at has about 5 options on my H/W monitor page. I have only (in White) 3 that are accessible by me:

Chasis intrusion [disabled]

Cpu fan temp [40]
cpu fan speed [62.5]

Then I have
PC Health status (in White)
**************
info in grey. I cannot access this area and it only gives me the +3.3v along with 2 temps, fan speeds, and a couple of other things, but only one voltage (3.3)

I was given the box for the motherboard (mainboard) and all the cords/plugs/adapters..., and 2 discs--MSI drivers and Utilities and HDD backup. Could there be something disabled elsewhere in BIOS that makes this inaccessible. Or is this a sign of a faulty BIOS program?

I had read that I could flash the BIOS but would need the HDD backup to do it. I have been afraid of completely ruining my machine if I try that. I do not know if the Drivers and Utilities was loaded. I had Windows xp and Microsoft office put on it before I bought it. Obviously there are drivers and utilities are on there, but I don't know if they came from the MSI disc.

I also should have "Cool n Quiet" on there, but I don't have that ANYWHERE. And I have gone through every page in the setup.

Any ideas?

I will try the voltmeter when hubby gets home (if I can't get BIOS to tell me the voltages). Now that you describe it, it does not seem very intimidating.
 
+1 to jsc and paperdoc - -1 for hunter 315
First - HW monitor is not reporting the +5/+12 correctly - The statement "No software reads voltages correctly" is BS. Try some of the other programs like speed fan, also check the MB website for a utility. Writing a program to convert voltage to a displayed value is easy and accurate, the problem with the MB programs is knowing WHERE to get the data and the algorimn to convert the reading. Like sex, need to know which hole to use.

The BEST way is to use a DVM (Voltmeter - a cheap one from radio shack will do) and follow what Parperdoc outlined - It is very safe and simple. A DVM comes in very handy and is easy to learn. I would leave the AC input to the PSU plugged in to the wall Outlet, BUT TURN the PSU rocker switch off. This leaves the Computer case at ground potential.

Jot down the voltmeter readings for the +5 and +12 V - Find a program that reports voltages the same as meter readings (with in 0.1 V at least).

While monitoring the +12, run "furmark" to load the +12V and not the drop in voltage. The +5V should not drop below 4.75V (standard TTL Logic min voltage) and the +12 V should not drop below 11.4 V ( Myself, I use 11.6V as a min value). For furmark, just google it.

If the software allows logging to a text file, use that to record values as you might not catch a low read befor power off, hopefully it will stil be intack after a shut down/reboot.

Also uncheck atomatic reboot on error. This way (hopefully) you will see the error message for what caused the reboot. Will not get an error message if PSU is shutting down on its own.
 
For K9N2G
See page C7 For MSI utility that will display +5 and +12 V
http://www.msi.com/index.php?func=downloadfile&dno=6796&type=manual

There is also a K9N2V, which may be different.

For Bios and manual, see
See http://www.msi.com/index.php?func=searchresult&keywords=MS-7511&table=download

As to Bios G = ver 1.5 dtd Oct 19 2009
For V Ver 2.5 same dtd
Updating Bios - Best way is from a floppy (or USB thumb drive - NOT FROM Windows And is safe, and fairly easy, as long as you follow procedure
 

PatriciaDooley

Distinguished
Feb 1, 2010
11
0
18,510


I already have that unchecked. There is no error that is showing up on anything. There is also NO reboot. It shuts off like it was turned off on purpose (like on shut down). Even in the logs I have found for the system, there are no errors listed, just shut downs.


I will try to do the other stuff. Hubby should be home in an hour or so (fyi, I am the electronics "guru" in the house...we would have a whole new computer by now, if he had his way--he'd rather replace it than fix it--and, although I like that idea, I want a laptop AND a desktop, not another desktop).

the MSI website (for the MB) has an online update for all of its drivers and BIOS. BUT it wants to flash it also and warns me that it could cause damage, would I like to continue? LOL So, I have not done it yet. Since I have the HDD backup disk, should I go for it? Or just wait.

Why am I not getting all of the info on my BIOS page that I should be getting...like it shows in the owner's manual? Would a new install of the drivers and utilities disk that came with the MB be worth it? Or will that erase everything like reloadin XP does?

Thanks again.
 

PatriciaDooley

Distinguished
Feb 1, 2010
11
0
18,510



My K9N Neo V1.0 . It is not the 2 or the G or the F or anything, so I am wary about using those others to do anything. This is why I was tempted to do the live update online from that website.
 
The past several years I've bought Gigabye Mother board. They have had problems with updating the BIOS under windows. Do not Know if MSI has had a simular problem. Here is the KICKER, if while doing the Flashing (rewriting the Biois code), your computer decides its time to save power and turns off - YOU may be getting that new computer sooner than you want. NOTE the warning - DO NOT POWER OFF while flashing/ updating the BIOS. Myself, I'd probably say It's worth a shot - If murphy bites me - Well I'd tell my better half, Honey I think your right - we need to get that new one.

As to manuals, Sometimes they provide a manual that covers several models, and some features may not be available for a specific model.

Added
A 2nd option is to replace the PSU and cross your fingers and hope that is what it was. (PS buy one that you would use for your next build)
 

PatriciaDooley

Distinguished
Feb 1, 2010
11
0
18,510
Well, gents, after doing the voltage with a voltmeter, I am not sure what to think. I am going to follow Retiredchief's suggestion and buy an new PSU that I can use with the next computer.

12V DC
normally stayed in the 12.12-12.17 range. It did have several quick fluctuations down to 9.48, 11.52, 11.83 over the 3 minutes of watching it.

It also did some wild crazy bargraph things (I am using a BK precision Pocket DMM3-1/2 digit Model 2700), but there was no change in the voltage number. The bargraph went from zero to max to the negative sign back to max in milliseconds for a couple of seconds.

5V DC
This normal range was from 5.13-5.15, but then did jumps down to 4.79, 5.05, 4.93, 5.07 in the 3 minutes I watched it (there were more jumps, these are an example). I also got the bargraph craziness here, but to a much crazier extent

I started running the virus scan (AVG) to tax it. After 5 minutes of running the scan I started getting the bargraph craziness.

After getting these readings, I tried to max out the computer. With the AVG scan running in the background, I opened my game WoW and it did nothing special. So, I got off there and went surfing. Got to Land's end and it did not like all the pics and shut down.

I turned on the computer (it does not automatically reboot) and started AVG up again. I started getting to this page to type all this in and it shut down on me again.

I am currently running this without AVG scanning.

Observations:
1. intense loads seem to make it waffle.
2. It is NOT heat related seeing as the computer was completely open during all this checking.
3. noise from the unit is not relative to the voltage being read (I can hear it processing or not processing and there was no apparent correlation of that to the voltage).

Questions:
Are these fluctuations normal when the comp is trying to work? So, that I would see slight drops in power every once in a while? This crazy bargraph thing? I don't even know what to ask about that.


Should I still run Furmark?

Thanks again for all your input. Thanks too retiredchief for the simple description of using the voltmeter.

added:
I will not try the BIOS update since I can't guarantee it won't shut down the computer due to overloading it. I am narrowing what causes it, but not always.
 

theholylancer

Distinguished
Jun 10, 2005
1,953
0
19,810
run linx or prime for cpu

then run 3d mark vantage for gpu

if it crashed on cpu then chances are that is your error, if it crashed on gpu then that is a more likely issue,

if it doesn't crash on any of the above, grab OCCT and run a cpu+GPU burn in test and stress the whole system, and if it crashes then it would most likely be a psu issue.


Also, what exact brand and make of psu do you have (serial numbers or other unique IDs please)? As the damned thing sounds like it's not outputting clean power (hence the fluctuation) and thus is providing unstable and bad power

if that is the case then you don't even need to do further testing and it's likely a psu issue

and wow is not exactly hard on the system... unless ur raiding icc-25 in 1920*1200 with all details maxed and 8x AA with the mages and shammes and pallies going nuts with their heavily animated deals

 

Paperdoc

Polypheme
Ambassador
The missing info in your BIOS screens is NOT an issue with drivers that did not get installed or updated. What is shown in those screens is entirely controlled by the BIOS itself. So if stuff is missing, what it likely means is that the manual was written for a version of the BIOS you do not have. Most likely scenario is the manual is more recent than your BIOS.

Flashing a BIOS is not too complicated, but it has risks you need to prepare for. First question is: can this process be done for your mobo? I looked through the manual for it and there is no mention of how to update the BIOS. So look on the MSI website for clear information on whether this can even be done on your machine. IF the answer is yes, the next thing to check is whether it is already done. That is, in your manual it tells you how to read the POST messages that show you the model number of your machine and the BIOS version and date. Compare this info with any downloadable BIOS update available from the MSI website. If you already have the latest BIOS, don't bother re-doing it.

Now, assuming you decide you CAN do an update and need a more recent BIOS, you need two items that you download. One is the BIOS file itself, and it MUST be specifically for your mobo version. The other is the utility software used to install or "burn" it. Basically, the process is to wipe out the old BIOS code stored in the EEPROM part of your BIOS chipset, then write new code into it. Sometimes there is more than one choice of utility. I prefer the type that you install and run from a floppy or USB stick (it will need to be a bootable USB stick that loads and runs a small OS, OR your existing BIOS needs a built-in screen to do this). My mobo and its utilities have a nice system. The software loaded from a bootable floppy and you could make a couple of disks. Then I ran the first one and used it to write to that floppy a copy of the existing BIOS, for backup in case I needed it. The second floppy contained the utility and the NEW BIOS file - running this actually erased and re-placed the BIOS code.

I agree with RetiredChief - I really don't like doing this job from within Windows, even if your mobo allows that as an option. I much prefer a way to do it from a basic simple OS that is not trying to do twenty-seven other jobs at the same time.

The risk elements are of two types. One is that the BIOS code supplied contains an error or it gets installed incorrectly, but you can still boot into BIOS afterward. In this case you can always re-run the BIOS update utility and re-install the new code to see it that works. Or, if you have the old BIOS version code, you can re-run and install that older code to get you running again.

The other risk is that you really cannot tolerate a power failure in the middle of running the update. If only a part of the new BIOS code is installed, it almost never can start up and work for anything. In these cases there are two recovery options. One is to replace the mobo etc completely. The less drastic that MAY be possible is to get a pre-programed new BIOS chip from the mobo manufacturer and replace the old one.

Neither risk is common. Mobo makers warn especially of the power failure problem just to be sure that users don't do silly things while running the update process. In your case, however, as RetiredChief pointed out, your machine has a history of unexplained random power-offs, so one might suggest you run a slightly higher risk that others.

IF you update your BIOS in this manner, you MUST remember to do two steps immediately after. First, you follow the procedure to Reset the new BIOS. This is the sequence: shut down and disconnect power; remove the quarter-sized BIOS battery from its mobo holder; move the BIOS Reset jumper on the mobo to the other pin pair and leave it on there for 5 to 10 seconds, then move it back to its initial storage position; replace the quarter-sized battery; and close up and reconnect power. This rests the BIOS. Now as the second step you boot directly into BIOS Setup screen and from the main menu choose the option to load the manufacturer's default Optimized Settings. This will put your new BIOS into a very stable condition that works so you can start and run properly. After that's working you can consider coming back to BIOS Setup and changing any configuration settings you need.

Of course, there is no reason at all to suspect that the BIOS itself is the source of your odd power-down symptoms, although I suppose anything is possible there. The BIOS update by "flashing" or "burning" really is aimed at getting you the latest improved BIOS, and maybe adding any new BIOS features that are not already in your machine. You'll have to decide whether it is possible with your machine, and whether you want to do it, balancing potential benefits vs. risks.
 

PatriciaDooley

Distinguished
Feb 1, 2010
11
0
18,510


I will work on doing those two (three) things.



here is the psu data
Mustang Switching Power Supply
model LC-B350 ATX (max 350W)
04/06 (month/year)

just googled the make and model...I am not alone with this problem with this PSU, but I am sure if I googled the MB I would find things there too, and the GPU, etc. So, I am scanning articles as I await some replies from you all.

Thanks again for all of your help and willingness to walk me through it.



ROFL umm...no, I am not running a 25 man raid or ANY raid for that matter, even my display options are way low, (for now--since this issue has started, just in case it was the GPU. I do love the cool AoE visuals, and being able to see detail to great distances so I hope to fix that! LOL) although SW still causes serious lag on the weekend evenings! Although I do wish I had the brain matter it takes to run an icc-25! (too old and too slow in the processing areas of the old gray matter).
 
Based on your voltage readings (DVM), Yes I would replace your PSU BEFORE attempting to update Bios

Your reading of 9.48 is less than the spec of min = 11.4 (I prefer a min of 11.6)
The 5 V - 4.75 -> 5.25 is OK, this is based on standard TTL logic Vcc. (my min for the 5 V is no lower than 4.8V.

Check out the ratings on psu, and for wattage try to estimate Max wattage (of next Build)and go minium of 20% higher.

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Review_Cat&recatnum=13
 

theholylancer

Distinguished
Jun 10, 2005
1,953
0
19,810
O god that PSU is deathly!!

I would not want to use the computer before ya replaced it!!! o_O

that unit is most likely a no name that is rated at 350 watts, and with the 8600 GT you are putting loads it most likely can't sustain...

get a good psu, maybe a antec 300 watt or something...

and meh wow is getting easier, the only thing you really need to raid is time and a guild willing to do it, marrogar is simply prince mala from kara with different animation imo, you run when he does the bone storm and avoid his coldflame which is like the infernals from price mala. I do like the gunship thing, that one is nice and new lol.
 

Paperdoc

Polypheme
Ambassador
I agree your PSU is under intense suspicion. The digital voltages do have too much variation. The wild bargraph readings you describe sound to me as if its outputs - both 12V and 5V - are showing random substantial drops, maybe right down to zero. Unless you have reason to think that data is incorrect - for example, if you know that the connections of probes to test points was intermittent - I would recommend you replace the PSU before proceeding. In the meantime, stop using the comp. With that kind of power you are lucky you don't have hardware damage already, and I REALLY worry the failure might happen in the middle of a disk data write and corrupt key areas, destroying all your access to the HDD.

I fully agree you MUST not consider updating your BIOS in this situation. AFTER you have the machine running cleanly with a new PSU you could come back to the question of whether there is any advantage to be obtained from that procedure.

One other thing to check here. A few years ago there was a whole raft of problems traced to widespread use of certain electrolytic capacitors in power supply circuits that were OK when first made but deteriorated over a couple years. The commonly-noticed symptom, other than PSU failure, was that some of them showed up as cans with bulged tops (from overheating caused by excessive internal conductance). BUT this same problem also inflicted itself on capacitors used on some MOTHERBOARDS in the Voltage Regulation area close to the CPU. I had a mobo fail this way. The capacitors typically are metal cylinders standing on end, ¼" to ½" diameter, ½" to 1" tall. Some are covered with plastic sleeves so the metal can is not so obvious. Their tops are supposed to be flat. On my failed board, about 7 out of 11 such units had rounded bulged-up tops. I do not know whether this problem would show up as voltage drops on the +12V and _+5V lines for Molex connectors coming out of the PSU, but I suppose it might if the mobo circuit elements were to pull too much power from the PSU rails from a temporary short in a cap. So, look over the larger cylindrical capacitors on your mobo for rounded tops, just in case.
 

PatriciaDooley

Distinguished
Feb 1, 2010
11
0
18,510
WOOHOO! I got the new psu...a Thermaltake TR2 600w. I tried and TRIED to crash it. No can do. I even checked the voltage on it. MUCH more stable.

Thanks for ALL of the help. Your advice helped me to try things I was wary about.

BUT you all never mentioned that the most complicated part of building a PC was the PSU! OH MY the number of cords and attachments looked like an octopus (especially compared to the old 350w).

Now, I have to find things to add so I can use all those new cords! LOL

Should I still leave the BIOS alone?


Thanks again everyone!
 
As my better half says - If it isn't broke don't fix it. To bad I don't heed.

If you are having a problem that could be associated with the BIOS - Yes.
You can also look at the change log to see if there is any benifit to updating the BIOS. Bear in mind they do not always mention every thing that the "New" bios update fixes - They don't want to advertize a previous boo-boo. They also include performance tweaks that are not specified.

If none of the above applies - don't

PS I just updated the BIOS on my new system from F4 -> F7 this evening along with Put a Blu-ray writer in. Didn't heed my own advise, but then I some times do just to see. I've never "Bricked" a Bios. Came close on my P965-DQ6 C2D E6400 @ 3.2 GHz, Used a bad FD (had some Bad sectors), but was able to recover. My i5-750 (Gigabyte P55-UM4P) you can put the Bios file on the HDD instead of the FDD

Also on BIOS updates - Never Update to one that just came out - wait tell it has been available for at least a Month - LET OTHERS beta test it.
 

theholylancer

Distinguished
Jun 10, 2005
1,953
0
19,810


heh yeah its the psu and the case (A good case is large enough that you can put those wires somewhere safely, additional 5.25 inch bays is a good place as long as you have the ends capped, and of course, the airflow) that is the most important part that most people don't check when they buy a new computer, the easiest part for manufs to skimp on (or a builder to forget) since the psu is enclosed, and the case ppl don't notice at all, and if your computer dies due to either adding too much stuff or not dusting it out and have heat issues for ventilation and they would tell you either the computer wasn't designed for additional stuff, or that it's wear and tear and tell you to buy a new one and it's not their issue that you "broke" it

Yeah bios updates are good if they are "tested" in the wilds for a month or two, and best if you know what to do if you do brick your mobo, some boards become decorations unless you send it to the mauf or knows how to solder with a new bios chip lol




 

Great choice for an inexpensive dmm.

What you were seeing here is a side effect of the instrument's sample time. Digital instruments do not read instantaneously or continuously. (Everyone seems to forget that.) A cheap dmm might have a sample time as slow as 4 times a second. Bar graphs in the display respond much more quickly. So here's what happens:
You measure.
Blink. Instrument samples and displays. Pause.
Blink. Instrument samples and displays. Pause.
OK. You get the idea. :)
With digital instruments, you have no idea what is happening in the "pause" interval. This is where the bargraph helps.

About flashing the BIOS, as RetiredChief says, "If it ain't broke, don't fix it." But if you need to, do not do it from Windows. Follow directions exactly. If you do not understand something, ask. Do not guess. I always flash late at night. I do not have a UPS and I think that the power grid is a little more stable late at night.

And congratulations. You done good. You successfully diagnosed and repaired a problem and learned something more about your computer. Altogether, a successful outcome. :D