doh!
is that surge protector the same one that all the blown PSUs have been plugged into? I'd be getting shot of it if it was.
PSU's keep burning out
- Thread starter Summermoon
- Start date
doh!
is that surge protector the same one that all the blown PSUs have been plugged into? I'd be getting shot of it if it was.
- Joined
- 1 Nov 2009
- Posts
- 36
- Location
- MOS UK
Yes it was the same one. IT also turns off all teh other peripherals and just leaves the PC's 5v rail working. (or3v can't remember)
Well about 2-3yrs ago we had a warn out trip switch, so that was the reason for buying the ups, cos the power use to go off at least twice a day. My mum thought it was going cost big money to get it fixed, because she was convinced the whole house needed re-wiring, and I kept saying "Its only the trip thats buggered", but you know what parents are like, they never listen. Anyway after about a yr later she finally got someone in to look at it, and I was correct.
Anyway, if I hadnt had my ups for that yr, my pc would have been buggered with the power always off and on, and never once my ups failed to kick in when the power died.
Im still using the ups now even tho my power prob is sorted, cos its protecting me if the mains under or overvolts during high winds or whatever and the odd power cut every now and again.. I have yet to try it during a storm
Its hard to think how fast the change over from mains to battery must be for a electrical equipment to not notice the change over, cos there must be a split second the equipment isnt getting any power??
Anyway, if I hadnt had my ups for that yr, my pc would have been buggered with the power always off and on, and never once my ups failed to kick in when the power died.
Im still using the ups now even tho my power prob is sorted, cos its protecting me if the mains under or overvolts during high winds or whatever and the odd power cut every now and again.. I have yet to try it during a storm
Its hard to think how fast the change over from mains to battery must be for a electrical equipment to not notice the change over, cos there must be a split second the equipment isnt getting any power??
Hello Summermoon,
Which case do you use? . . . is the case interior hot? . . . does the PSU sit at the top or bottom?
So you physically saw it with your eyes? Or just did what most do - know it cleans power because hearsay said so?
A typically UPS output in battery backup mode is some of the ‘dirtiest’ power a computer will ever see. Why do they recommend no power strip protectors or motorized appliances on a UPS? Because its power is so 'dirty' as to be potentially harmful to those other devices. Since all computers are required to be so more robust, then a UPS harmful to a power strip protector is also ideal perfect power for every computer power supply.
No power supply must harm its load. If a power supply damaged a supply tester, then the supply is missing required functions. Required functions so that any power supply failure cannot cause damage to a load (supply tester, motherboard, etc).
Power supply tester is a useless tool. It will report some supplies as good when that supply is defective. Cannot identify some conditions that could predict a supply failure months later. Does not report with numbers. Essential to measuring a supply and obtaining useful replies here is the far more useful 3.5 digit multimeter. Costs almost same money. Numbers from the meter could have also identified other reasons for supply failure quickly - including a dodgy power receptacle.
UPS has only one function - to protect data from unexpected power loss. Anything that UPS would solve must be already done inside the computer power supply. UPS has a power supply equivalent or less robust than one required for all computers. What does that UPS supply do? Read its spec numbers. It does not claim to protect the computer from transients. That protection is a myth based in hearsay; not based in manufacturer specifications.
No, its because if a ups failed, it could because of the powerstrip(s) thats plugged into it. I have 3 power strips connected to my ups, I have about 8 devices plugged into the ups together, so I will have to tell porkies if my ups failed during a storm or whatever.
Im letting my mum borrow my ups soon for 3weeks for her chicken egg incubator, just incase the power dies
Edit: A pc is not robust at all
Last edited:
This is the spec of my belkin 1200va UPS
# Total Energy absorption - 420 Joules
# Maximum Surge / Spike - 6,500 Amps / 6000 Volts
# AVR response time <220ms
# AVR SmartBoost increases 10% of input voltage if input is 187~264VAC
# AVR Voltage Detection tolerance is +/- 6VAC
# Transfer to battery 187~264 VAC +/- 6%
# Output voltage on battery 230V +/- 5%
# Total Energy absorption - 420 Joules
# Maximum Surge / Spike - 6,500 Amps / 6000 Volts
# AVR response time <220ms
# AVR SmartBoost increases 10% of input voltage if input is 187~264VAC
# AVR Voltage Detection tolerance is +/- 6VAC
# Transfer to battery 187~264 VAC +/- 6%
# Output voltage on battery 230V +/- 5%
So you do not own and use an oscilloscope. Cleanest power from a UPS is when it connects the load directly to AC mains. Square waves (called modified sine waves) with spikes when powered from its battery. And that 'dirty' power is ideal for all computers.
Plug one UPS into the output of a second. When the second is disconnected from AC mains (goes into battery backup mode), then the first UPS may also go into battery backup mode. Why? Because power output by the second UPS (in battery backup mode) is so 'dirty' as to upset the first.
Reason so many claim a UPS 'cleans' electricity: hearsay. Most are only educated by retail propaganda. Even the UPS manufacturer does not claim that 'clean' electricity in numeric specs. If Phil2008 uses technical knowledge to 'know', then he posted those numeric specs. Why no specs? A typical UPS outputs power so 'dirty' as to be harmful to some power strip protectors and small electric motors.
Phil2008 - your 'experiment' proves nothing. Those power strip protectors damaged by a UPS also would not report their failure (that type of failure) on its indicator light. Your example says nothing. A UPS in battery backup mode can be some of the 'dirtiest' electricity a computer will ever see. Post UPS spec numbers to reply in a technically responsible manner. Power is 'cleanest' when a UPS connects the computer directly to AC mains - which is how most typical UPS work.
Your numbers say no protection for a power supply. Surges that might damage that supply are hundreds of thousands of joules. Your Belkin uses only 140 joules and never more than 280 in surge protection. How does it absorb surges that are hundreds of thousands of joules.
All computers must work normally - this is perfectly ideal power - from 180 to 265 volts. So what did that UPS do? Well if voltage drops very much lower, then the computer simply powers off - without any hardware damage. What does that Belkin UPS do? Provide temporary power so that unsaved data can be saved. And does so typically with some of the 'dirtiest' power that computer will see.
Your specs say that Belkin does nothing for the OP's possible problem.
Did the OP's power supply destroy his power supply tester? Well that UPS also would not have averted that failure.
How does that Belkin absorb typically destructive surges? Its 140/280 joules are near zero. Just enough above zero so that you will claim it does surge protection. That hearsay conclusion recommends a UPS to 'clean' power. How does near zero surge protection protect the OP's power supply? It doesn't.
At the end of the day you cant expect a ups running of 2 12v batts to produce 240v as good as mains power, but it produces it well enough to keep your equipment going during a powercut or storm, but the switch over from mains to batts may not be quick enough to stop surges getting to your equipment tho, but it must be fairly quick for the pc not to die on switch over??
Last edited:
A UPS is slow switching between a direct AC main connection and battery power. Another function in a power supply - to maintain uninterrupted power during those long milliseconds when a UPS outputs no power.
Meanwhile, that is not relevant to the OP's problem. He is suffering power supply failure. A UPS does not protect his supplies from destructive anomalies. Anything a UPS might do must already be inside that supply. Apparently his problem has even damaged a power supply tester. No UPS or power strip protector would avert that.
Power supply must be so robust as to withstand 'dirty' UPS power without damage. And so the relevant question: why the OP has so many failed power supplies.
Agree,, 1 thing my ups wont keep going is my yamaha amp, it starts off good when it switches over, but amp dies within 2secs
But westom, saying all that, my ups has been a godsend these last few years and I would have had a busted pc by now if I hadn't of bought 1. Even tonight we had a powercut for a few secs and my ups kicked in and kept my pc and a few other things from going off. Also whats this AVR purpose then if it dont provide clean or cleanish power, cos at the end of the day its false advertising and thats not allowed??
- Joined
- 1 Nov 2009
- Posts
- 36
- Location
- MOS UK
The Case is an Xblade with the PSU at top.
There are 3 120mm fans(Rear Side and front) NB and SB 50mm fans, a Zalman CNPS 9900 LED PWM for the CPU. Two Antec 70mm directional fans.
The case never seemed to get hot CPU ran at 29.c-33.c when OC'd to 3.82Ghz.
There was a hefty build up of dust at the front before the filters and the case switch was shall we say a fluff ball of dust.
The Surge protector is a Oneclick Intellipanel.
The psu's in question are respected enough that the OP would need to be astonishingly unlucky for them all to self destruct in such time frames. Since it's only one box causing problems and the rest of them are fine, and considering how resilient atx psu's tend to be to input, the electricity supply isn't likely to be the problem.
This leaves overheating and possibly a rogue motherboard. Overheating's the more likely one if the case/airflow supports the theory, or if it's full of dust.
The first failure killed your cpu, but not the motherboard. That strikes me as unlikely, the currents will have gone through the board to get to the cpu and motherboards are considerably more fragile than processors in general. So it's possible that the motherboard was damaged, but not killed completely. A faulty motherboard can do all kinds of mad things, killing off attached power supplies isn't beyond the realm of possibilities.
Hope for overheating, as persuading Asus to take an rma of a motherboard which seems to be working but is suspected of killing power supplies will be an experience.
We're discussing mains ac here, which isn't at constant voltage. As such volts x amps isn't going to get you power delivered. It's possible you can just sub in rms voltage instead and it'll work out ok, but you might need to take phase differences into consideration explicitly. Over to westom for that one.
Square waves aren't modified sine waves, they're a large number (in the ideal case, an infinite series) of sine waves superimposed. Unless by "modified" you mean "adding loads and loads of other sine waves".
"Dirty power" patently isn't ideal. At best you're arguing it's sufficient for some computers. Achievable overclock is certainly a function of power supply as well as other parameters. Which probably means that one's stable 4.4ghz processor won't be stable when running off the ups, which may be a source of concern.
If I'm being pedantic it's because you occasionally claim to be a chartered electrical engineer, and it seems strange that as such you would misuse terminology like this. I'm not very convinced by your attitude that numbers are all important either, as its not one I've seen in my colleagues or lecturers, but this may well be an eccentricity of yours rather than common across the discipline.
edit:
Just seen a post detailing the case. I'm going to say that overheating is therefore unlikely to be the problem, and put forward rogue motherboard as my best explantion for what's happening.
Where do you think you might be going wrong here?
This leaves overheating and possibly a rogue motherboard. Overheating's the more likely one if the case/airflow supports the theory, or if it's full of dust.
The first failure killed your cpu, but not the motherboard. That strikes me as unlikely, the currents will have gone through the board to get to the cpu and motherboards are considerably more fragile than processors in general. So it's possible that the motherboard was damaged, but not killed completely. A faulty motherboard can do all kinds of mad things, killing off attached power supplies isn't beyond the realm of possibilities.
Hope for overheating, as persuading Asus to take an rma of a motherboard which seems to be working but is suspected of killing power supplies will be an experience.
We're discussing mains ac here, which isn't at constant voltage. As such volts x amps isn't going to get you power delivered. It's possible you can just sub in rms voltage instead and it'll work out ok, but you might need to take phase differences into consideration explicitly. Over to westom for that one.
Square waves aren't modified sine waves, they're a large number (in the ideal case, an infinite series) of sine waves superimposed. Unless by "modified" you mean "adding loads and loads of other sine waves".
"Dirty power" patently isn't ideal. At best you're arguing it's sufficient for some computers. Achievable overclock is certainly a function of power supply as well as other parameters. Which probably means that one's stable 4.4ghz processor won't be stable when running off the ups, which may be a source of concern.
If I'm being pedantic it's because you occasionally claim to be a chartered electrical engineer, and it seems strange that as such you would misuse terminology like this. I'm not very convinced by your attitude that numbers are all important either, as its not one I've seen in my colleagues or lecturers, but this may well be an eccentricity of yours rather than common across the discipline.
edit:
Just seen a post detailing the case. I'm going to say that overheating is therefore unlikely to be the problem, and put forward rogue motherboard as my best explantion for what's happening.
Where do you think you might be going wrong here?
Last edited:
AVR is already inside your computer. That was posted previously with numbers. If you did not understand what those numbers were saying, well, this is the subjective version: AVR is already performed inside every power supply. Incandescent bulbs must dim to less than 40% intensity. That is normal and ideal power to any computer. AVR is already inside.
So what is that expensive AVR doing? Nothing useful. Is it 'cleaning' power? No.
Your UPS demonstrated its only function. To provide temporary power during a blackout. To protect unsaved data. Blackouts do not damage hardware. So what is a UPS doing? Protecting data (and saving you time doing a reboot). Those other functions (AVR, surge protection) are near zero or already exist elsewhere. And look so good to those who only read sales brochures. Fluff to promote higher profit sales.
UPS does nothing for the OP's computers. Will not even cure symptoms. Unfortunately insufficient facts identify the OP's subjectively defined problem.
Hard numbers would go a long way to creating an immediate and useful reply. For example, if he had provided voltage numbers from each damaged supply's output wires (before removing the failed supply), then better answers are possible. Currently, we can only speculate that the wall receptacle is defective. Numbers from that receptacle would also go a long way to solving this problem without speculation.
Overheating is routinely blamed when one does not know what else to blame. All computer power supplies must be perfectly happy in a 40 degree C room. Which is why heat is also a useful diagnostic tool. If finds defective hardware that is still working in a 20 degree room. However many will blame heat rather than learn the failure that a diagnostic tool - heat - is reporting.
This UPS is called a modified sine wave. And its output is square waves with a spike between those square waves. Are they lying? No. From high school math. A square wave is nothing more than a sum of different sine waves.
Some of the 'dirtiest' power a computer will ever see comes from a typical UPS in battery backup mode. To promote the myth of clean power, many UPS manufacturers will describe the output as modified sine wave, stepped sine wave, or pure sine waves. All repetitive waves are sums of sine waves. So they are not lying. And a UPS does nothing for the OP's problem.
I did not see this in his post. First power supply killed a CPU and not the motherboard? That is very unlikely because the CPU is behind a completely different power supply that is on the motherboard. Typically that power supply fails - not the CPU.
Those power supplies also explain why anything on AC mains do not harm the CPU. First AC mains voltage is increased to well above 300 volts - and converted from AC to DC. Then it is converted to high voltage radio frequency. Then pushed through galvanic isolation. Then converted to DC again and filtered more. Then it is delivered to the CPU's power supplies which do these power supply functions again.
More likely to cause CPU damage is ESD or some other human intervention.
But with so few hard facts, then we can easily speculate hundreds of reasons for his failures - mythical and real. Hard facts (ie numbers) would go a long way to resulting in a useful reply. So far, none (including mine) have provided the OP with anything that is actionable.
so there is more integration with the PSU and the surge protector than with a normal setup? I'd chuck it right now, it sounds like the 'common link'