Persistant pc problems

[PhoenixUK]

Just opened 18 instances of firefox, each one with 3 tabs and flash game running in each tab.

That's on my media PC with e2200 and 2gb ram.
Apart from massive slowdown due to pagefile access ( ran outta ram ) I can't really see any problem.

The CPU is at about 60% load.

I bet its either your power socket as Jon have said above or you just have VERY VERY bad luck. Or your PCs had some rubbish chinese PSUs which couldn't cope with the load ( which most likely is the case as PCs don't just die randomly, it would have been crappy PSU that blew up and took all the components with it ).

If you want a new build I would grab 955BE with 8GB ram and seasonic/be quiet PSU, should be more than fine with that kinda stuff.

 

[ben500]

Just opened 18 instances of firefox, each one with 3 tabs and flash game running in each tab.

That's on my media PC with e2200 and 2gb ram.
Apart from massive slowdown due to pagefile access ( ran outta ram ) I can't really see any problem.

The CPU is at about 60% load.

I bet its either your power socket as Jon have said above or you just have VERY VERY bad luck. Or your PCs had some rubbish chinese PSUs which couldn't cope with the load ( which most likely is the case as PCs don't just die randomly, it would have been crappy PSU that blew up and took all the components with it ).

If you want a new build I would grab 955BE with 8GB ram and seasonic/be quiet PSU, should be more than fine with that kinda stuff.

Thanks, power socket sounds possibility, pc is run from extension which I would concede is probably overloaded, I wasn't aware this would have any effect, I do however have a belkin surge and powersave in line. You've certainly given me some things to think about that's for sure.
I normally only open one instance of firefox and run as many as eight or ten tabs each with a related window with flash game running, when the browser's are affected it isn't normally one, they tend to fail one after another until stabalising when I don't re-open dead browsers and run minimal amount of windows. I will repeat the above for seamonkey,kmeleon etc, ie doesn't allow me to have more than one instance of the game running as latest window open takes id of previous window and not new game.

 

[Mattus]

If your PC uses good components and is configured correctly, then it shouldn't crash no matter how much load you put on it.

I would suggest not buying any more crappy prebuilt PCs from Acer and getting a decent system from OcUK. Doesn't matter if you build it or they do, as long as it has quality components in it.

 

[ben500]

If your PC uses good components and is configured correctly, then it shouldn't crash no matter how much load you put on it.

I would suggest not buying any more crappy prebuilt PCs from Acer and getting a decent system from OcUK. Doesn't matter if you build it or they do, as long as it has quality components in it.

Again sound advice, problem being is I don't understand enough to know exactly what my requirements are, I bought this latest Acer because I thought it would be able to handle it, so far it has, it is a much lesser specked Acer that has recently died out on me.

 

[PhoenixUK]

Just if you wanted to get a new PC:

Product Name Qty Price Line Total
AMD Phenom II X4 Quad Core 955 Black Edition 3.2GHz (Socket AM3) - Retail AMD Phenom II X4 Quad Core 955 Black Edition 3.2GHz (Socket AM3) - Retail £149.99
(£130.43) £149.99
(£130.43)
Gigabyte GA-MA770T-UD3P AMD 770 (Socket AM3) PCI-Express DDR3 Motherboard Gigabyte GA-MA770T-UD3P AMD 770 (Socket AM3) PCI-Express DDR3 Motherboard £76.99
(£66.95) £76.99
(£66.95)
Thermaltake V9 Case VJ40001N2Z - Black (No PSU) Thermaltake V9 Case VJ40001N2Z - Black (No PSU) £56.99
(£49.56) £56.99
(£49.56)
OCZ Platinum 4GB (2x2GB) DDR3 PC3-10666C7 1333MHz Dual Channel (OCZ3P13334GK) OCZ Platinum 4GB (2x2GB) DDR3 PC3-10666C7 1333MHz Dual Channel (OCZ3P13334GK) £46.99
(£40.86) £93.98
(£81.72)
Seasonic Silverpower 400W Power Supply Seasonic Silverpower 400W Power Supply £45.99
(£39.99) £45.99
(£39.99)
Samsung SpinPoint F1 320GB SATA-II 16MB Cache - OEM (HD322HJ) Samsung SpinPoint F1 320GB SATA-II 16MB Cache - OEM (HD322HJ) £34.99
(£30.43) £34.99
(£30.43)
Sapphire ATI Radeon HD 4350 512MB "Low Profile" DDR2 TV-Out/DVI/HDMI (PCI-Express) - Retail Sapphire ATI Radeon HD 4350 512MB "Low Profile" DDR2 TV-Out/DVI/HDMI (PCI-Express) - Retail £34.49
(£29.99) £34.49
(£29.99)
Sub Total : £429.07
Shipping cost assumes delivery to UK Mainland with:
DPD Next Day Parcel
(This can be changed during checkout) Shipping : £11.75
VAT is being charged at 15% VAT : £66.12
Total : £506.94

This should laugh at any amount of flash games you throw at it while dancing, singing and cooking you a dinner at the same time.

 

[westom]

Thanks, power socket sounds possibility, pc is run from extension which I would concede is probably overloaded, I wasn't aware this would have any effect, .

Power is not your problem. In fact, put an incandescent bulb on the same power line. That bulb must dim to less than 40% intensity before the computer shuts down. A requirement defined long before PCs exists and defined in numbers from Intel specs.

Low power does nothing destructive to electronics. If power gets too low, the system simply shuts down without any hardware damage. The Belkin also does nothing. Read some numbers on its box. Is that number something like 500 volts? That means it does nothing different from the extension cord until voltage exceeds 500 volts.

Not yet clear is whether you have hardware damage or just blame system lockups (a software problem) on hardware failure? IOW the details of what has failed and why must be provided. Are you saying each computer simply will not boot anymore? Software cannot cause hardware damage.

Are you having a software bottleneck? Load Task Manager. Put that window in a corner. Monitor categories such as CPU and other columns (use VIEW to get them) such as Memory Delta and PF delta. Is the disk drive constantly access?

Currently you have not even said what is failed or detailed any symptoms. As predicted, numerous 'it could be this or could be that' replies result.

Finally, did you review various system logs such as the system (event) logs? Windows will record a problem and work around it. Leaving you to solve it later. Just another example of getting information and numbers so that the next replies can be useful.

 

[JonJ678]

I think you've misread part of this westom. I'm not suggesting insufficient power, as this would blow the fuse in the extension cable rather than kill the computer. I'm suggesting particularly jagged mains ac from a badly wired socket which the psu can't output clean enough power from, leading to dead components.

 

[ben500]

Ok I'll try and clarify, I have three dead systems, first is a Dell 4700 series which packed up just over four months ago, fan was running nuts for a while prior to failure, now when I switch on the fan runs fast and loud and system will not boot, second is acer which until yesterday apart from browser problems was running fine, yesterday I was not watching system as it is secondary system and was running whilst I was active on main, I noticed that graphics quality had become really poor, the mouse was not reactive and I did a hard shut down, the system will no longer boot up, when I switch on I get the red symbol described above on the screen but nothing else, I have switched the graphics card but no result there, power light comes on and dvd drive appears to start up but no more activity than this that I am aware of, no beeps.

The third is a gaming machine which originally would light up all the flash lighting but not boot up, I have tested the power supply on another unit and it is fine, I suspect that the motherboard and or processor may be the problem with this one, but not sure how to determine which.

Essentially I have two problems one browser based and the other would appear hardware based as I understand it. I have not viewed event logs in answer to that question, as to be honest I don't recall how to.

I will monitor task manager tomorrow when I am due to play another session, is there anything I should be keeping an eye out for by any chance?

 

[ben500]

Looking at your spec pheonix, I suspect between the three dead pc's I have that I have a good deal of that equipment of equal or greater spec, I shall have a look, I'm sure I've got a decent graphics card in the gamer, I have 4g spare ram, and an excellent psu for starters and am wondering if I can use these and water it down to bring down the cost a little, I'll need to check just exactly what I have and compare with your recommendations.

Thanks for all your comments folks I appreciate your time helping.

 

[ben500]

Is there a way to avoid/minimise bottlenecking?

 

[westom]

I'm suggesting particularly jagged mains ac from a badly wired socket which the psu can't output clean enough power from, leading to dead components.

Is the extension cord spitting sparks where a wire is badly connected? If not, then no 'dirty' power exists. What is 'dirtier' than power from a power cord? Power from a UPS in battery backup mode.

A badly wired power cord (that is not spitting sparks) can cause lower voltage. Lower voltage is not harmful to anything electronic. If voltage is even lower, the only problem is a house fire. As noted previously, extension cord could lower voltage so much that an incandescent bulb is at 40% intensity. All computers must (even long before PCs existed) startup and operate just fine at that voltage.

You don't define 'clean' power. Low voltage? Noise? Harmonics? What electrically is this 'dirty' power? A subjective 'dirty' is how junk science gets promoted. What is this 'dirty' that is measured by numbers? If no sparks are coming from the extension cord, then the worst is low voltage. How low? Resulting hot power cord causes a potential house fire. Fire puts the computer at greater risk.

 

[westom]

Is there a way to avoid/minimise bottlenecking?

You want a solution before a problem was identified. That is wild speculation. Always first - identify the problem. Fixing comes later. To answer your question without identification means answering for maybe 15 different bottlenecks. If not specific - with numbers - you will not get any useful answers. Only get speculation.

Meanwhile software bottlenecks do not harm any hardware. Your previous post describes apparent software failure - ie system will not boot. And nothing that says what has failed.

For example, the Dell has diagnostic lights on the back. What do those report? What exactly do you mean by not power up? Will it power up with the comprehensive hardware diagnotic CD installed? Just not boot Windows?

Another machine's power supply was OK. How (and with details and numbers) do you know? A completely defective supply can work just fine in one system. And a perfectly good supply can fail in an otherwise perfectly good system. Nobody knows if that supply is good in that system. Without a multimeter (with nothing in a computer disconnected or removed), then we have no idea what is wrong with that third system. Swapping parts says near zero to nothing about why hardware fails. Answering that question is simplified if numbers or diagnostics are provided.

Defined are three machines that you believe are hardware damaged. Software (and bottlenecks) do not cause hardware damage. A computer works no harder with 40 windows open. It just works longer. Its CPU is working only as fast as its clock pulses (that continue at the same constant pace) with one or 40 windows open. Better equipment only means its functions will work faster - at the same speed with one window or 40 open windows. 40 Windows put no additional strain on hardware other than make it work longer - which is perfectly normal for every machine.

Which brings us back to the results from an autopsy. What exactly has failed? Your reply suggests hardware has failed. Fine. There are maybe 30 hardwares in a system. Which hardware has failed?

So you power it on. In all three systems, does the BIOS not even show anything on the screen? First all computers execute a program called a BIOS. Long after that BIOS program starts, the BIOS then load Windows and stops executing. Windows takes over. You must detail how much of (if any of) the BIOS executes. What happens when you keep pressing F2, F5, or F8 (whateven key is unique to that motherboard BIOS) constantly every second for maybe one minute after power is applied (to get other BIOS software loaded and executed)?

You have two task. To identify the hardware failure. To identify what (if any) software bottlenecks exist. Software does not cause hardware damage.

 

[JonJ678]

You don't define 'clean' power. Low voltage? Noise? Harmonics? What electrically is this 'dirty' power? A subjective 'dirty' is how junk science gets promoted. What is this 'dirty' that is measured by numbers? If no sparks are coming from the extension cord, then the worst is low voltage. How low?

Bit repetitive there, but I take your point. You however have also failed to define noise, and can probably guess that by dirty I meant noisy. Capacitive relationships between the main power lines and their environment is what I have in mind, the ac causing a build up of charge in the nearby building materials which discharges semi-randomly back into the power lines.
Less exotically, damaged insulation in the cables leading to occasional voltage transfer between the lines, but inadequate to cause a true short circuit.

I disagree with the idea that the worst that can happen is an annoying but safe low voltage. if you feed a psu wildly fluctuating voltages, you'll still get 12V out, but the variation on it can easily be out of atx spec. Particularly an issue when using cheap power supplies where the voltage regulation circuitry won't be that good to begin with.


You're definitely my favorite newcomer to these boards westom, good to have you here.

 

[westom]

I disagree with the idea that the worst that can happen is an annoying but safe low voltage. if you feed a psu wildly fluctuating voltages, you'll still get 12V out, but the variation on it can easily be out of atx spec.

Let's review what a power supply does. AC electric goes through a filter. then it is rectified (converted from AC to DC). Then that voltage not at more than 300 volts is filtered by large electrolytic capacitors. Then the regulator turns that DC voltage in a radio frequency. Then that voltage is put through a transformer. Then the radio frequency is rectified. Then it goes through more filters. Meanwhile any variation in that is made completely irrelevant by a regulator operating at radio frequencies.

No matter how massive the variation on 230 volts, those so many conversions, filtering, galvanic isolation, and regulator responding at radio frequencies makes any AC variations completely irrelevant. Either supply provides all the power required; or it is shutdown by the power supply controller.

Noise can occur in many forms. For example, a power supply has a line filter up front to make noise of so many frequencies or modes irrelevant. Notice many more filters take out noise. Noise is typically single digit or a few tens of volts. If noise is a problem, well, your AM (medium band) radio would probably be more harmed by no station reception.

Computer supplies must be some of the most robust devices in the building. Having said that, only the computer assembler is responsible for providing those necessary functions. Since the computer assembler is responsible, then many supplies may dump inferior supplies into a market devoid of technically knowledgeable assemblers. One function routinely missing is that first line filter. The computer may even violate Federal standards. And everyone but the computer assembler is innocent.

The computer assembler who only understood dollars and watts has no idea that his system was defective due to his parts selections. Then the customer has electronics failure. Customer then blames failure on the "usual suspect" (rather than perform an analysis) such as surges, heat, or other popular myths. What a mess.


OP has problems but provided no useful facts (yet) to begin an analysis. At best, each appears to be hardware failure. But nothing has yet been defined as defective. Everything in each system is still undefined.

 

[ben500]

Sorry Westom but what exactly do you require from me? Bearing in mind I am just a layman coming here for help not lectures, I have no idea how to perform diagnostics on a machine to determine the cause of the problem, if I had perhaps I wouldn't be here asking for assistance.

 

[westom]

Sorry Westom but what exactly do you require from me? Bearing in mind I am just a layman coming here for help not lectures, I have no idea how to perform diagnostics on a machine to determine the cause of the problem, if I had perhaps I wouldn't be here asking for assistance.

View the Dell, for example. Does anything come up on the screen? Is there any action on anything? Lights? Noise? Will it boot from its Diagnotics CD-Rom? Do the diagnostic lights on the back show anything?

Same questions apply to each machine in order to, step by step, determine what has failed and maybe get all three machines working again.

These above questions just to establish if your problems for each machine are in software or hardware. That was never crystal clear and that was the point.

 

[JonJ678]

Then westom, I put to you this. Why are there reports online of computers repeatedly failing on one power socket, but running fine on any others?

The OP is not using very good hardware. The psus will be cheap and nasty. Does your own post not suggest that these psu's will be less capable of coping with dirty input voltage? Does this not lend credance to the idea that its a dodgy power input that is killing things?

I do not have any cases bookmarked, but my memory is pretty solid and I haven't picked the idea of a faulty power socket killing computers out of thin air.

Tying the signal onto rf isn't going to do anything about noise on the input, it just vastly reduces the noise picked up internally. I suspect you know this, so wonder why you bring it up. It might help to know what your technical background is, I'm a materials scientist masquerading as an engineer. Undergrad student. Consequently I have a pretty sound knowledge of solid state, and enough electrical knowledge that I don't need rectified defined for me. Theres a definite gap between electrons and mechanics which I'm not so confident with.

Please increase the technical content of your posts, as currently simplifying it for my benefit is doing nothing but confuse the issue. For example I'm very skeptical of the psu 'turning dc into radio frequency', I rather suspect you mean adding a rf waveform to the currently still messy dc to facilitate cleaning the signal later on. Precision would be appreciated more, even if I have to google a few of the terms.

Cheers man

apologies to the op, ive hijacked your thread somewhat