Do I need a UPS?

[Panos Beltis]

A friend of mine says i need a UPS for my rig, because of the voltage outbreak and it can destroy my computer he says, do I really need it? Or I can just close from behind the switch whenever the voltage drops out/when I'm not at home?

 

[gangrel]

IMO, it's the cheapest form of insurance for your hardware you can get.

And over and above that...a very short power drop takes down your system. If you're in the middle of doing something, whatever you're doing just went POOF as well.

 

[GearUp]

Good modern power supplies have much of the same protection but I still use UPS. It will shut down the computer properly on loss of power and 1-2 second interruptions have no affect at all. I tend to shut down computers during lightning storms but anticipating power outages isn't usually possible.

 

[jsmithepa]

Depends on quality of your electricity. If you have to ask, I say you don't "need" it but it would be nice.

 

[westom]

A friend of mine says i need a UPS for my rig, because of the voltage outbreak and it can destroy my computer he says, do I really need it?

Incandescent bulbs can dim to 40% intensity. Due to something called a power supply, even voltage that low is normal and sufficient for every properly designed computer. If voltage goes lower, a computer simply powers off. Low voltage do not damage hardware.

But low voltage is potentially harmful to motorized appliances. If your voltage is that low, then the refrigerator, furnace, and washing machine need a UPS.

AC utility must either provide sufficient voltage or cut off power - to protect your motorized appliances. Only advertising and hearsay get many to wildly speculate - then fear voltage variations. Circuits long required in computers (long before PCs) made voltage variations irrelevant.

UPS purpose: temporary and 'dirty' power so that unsaved data can be saved. If it does anything else, then the recommendation also cites a manufacturer specification number that says it does that.

 

[GearUp]

I'm not an electrician but if there is any point in reporting the tests of power supplies I expect that it goes beyond performance. I only recently appreciated what impedance actually means in practical situations. I'm not familiar with "power outbreak" as such.
I'd like to see what affect a UPS coupled with power supplies has in some tests. For example does ripple improve, I'd guess not. I know that power regulation in older computers was not that impressive and potentially could damage components. When combined with a "brown-out" I would expect the situation to be worse. One of the issues with power outages concerns the return of power. I would prefer the UPS to handle a surge rather than the power supply which is directly connected to components, fast as the PS might be in protection. I can stop using the UPS but I might not have a suitable PS on hand as a replacement.
Power in my area has been quite good, whereas the OP probably has more serious issues. I found that an auto battery charger is very useful for charging the battery in my UPS so I am less inclined to stop using it due to expense and time consumed in replacement.

 

[westom]

I'd like to see what affect a UPS coupled with power supplies has in some tests. For example does ripple improve, I'd guess not. I know that power regulation in older computers was not that impressive and potentially could damage components.

What happens to cleanest or 'dirtiest' AC power? First converted inside a PSU into high frequency spikes that exceed 300 volts. Then superior filters and regulators convert that 'dirtiest' power into the rock solid and stable, low DC voltages; as all 'properly designed' power supplies have always created.

Since most computer assemblers know nothing about electricity, then many third party power supplies were dumped into the market missing essential functions. They knew a computer assembler (not the supply manufacturer) is responsible for meeting industry and government standards. This (and not a brownout) explains some damage.

Power restoration does not create an overvoltage. Everything is demanding most current. So voltage increases slowly. This is hard on motorized appliances. And is ideal for electronic appliances. In fact, we sometimes design an inrush current limiter into some supplies to intentionally make voltage climb slowly; to increase electronic life expectancy.

UPS does not avert power transients. Read its specification numbers. Some 'dirtiest' power seen by electronics comes from a UPS in battery backup mode. Some call it a pure sine wave UPS. Yes. Square waves, stepped waves, and spikes are nothing more than a sum of pure sine waves. They did not lie. They used subjective reasoning to manipulate naive consumers who failed to read specification numbers - such as %THD.

Brownouts do not damage electronics. Tom MacIntyre demonstrated same:

We operate everything on an isolated variac, which means that I can control the voltage going into the unit I am working on from about 150 volts down to zero. This enables us to verify power regulation for over and under-voltage situations. ...
Switching supplies ... can and will regulate with very low voltages on the AC line in; the best I've seen was a TV which didn't die until I turned the variac down to 37 VAC! A brownout wouldn't have even affected the picture on that set.

Die as in stop operating - not fail. All designs are tested this way since low voltage must never damage electronics.

An AC utility demonstrates what a typical UPS output looks like when switching from AC mains to battery. Cleanest power on left is provided direct from AC mains. Dirtier power on right is from a UPS battery:
http://www.duke-energy.com/indiana-business/products/power-quality/tech-tip-03.asp

No problem. Because robust protection must already exist inside each computer's PSU. These facts and numbers often contradict popular urban myths created by hearsay, subjective reasoning, and advertising.

 

[GearUp]

I appreciate the information from someone knowledgeable but what I would like to know is the quality of the UPS power when not on battery. The battery simply assures proper shutdown so data is not lost and system files are not corrupted and for me is typically within a minute.
Historically CRT TVs I am familiar with would stay off when a power outage occurred. This seemed to be critical to larger and older TVs. Actual failure may only be quickened by such an event but good TVs could last over 30 years. TVs and other equipment showed obvious affects when power shortages occurred. More recent TVs would turn on when power was resumed, presumable because of the current limiter you mention but still questionable policy. Presumably power grids have also changed in that time but a few years ago my refrigerator noticeably struggled to recover from an outage (yes motorized and OK).

 

[westom]

... but what I would like to know is the quality of the UPS power when not on battery. The battery simply assures proper shutdown so data is not lost and system files are not corrupted and for me is typically within a minute..

View two waveforms in http://www.duke-energy.com/indiana-business/products/power-quality/tech-tip-03.asp . That is a UPS output voltage before (left) and when (right) in battery backup mode. A cleaner sine wave on the left - which is UPS output power when not in battery backup mode.

At risk is unsaved data. Only obsolete filesystems from 20+ years ago would cause saved data / OS file corruption when power was suddenly lost. A modern filesystem does not lose data due to a sudden power loss. Only unsaved data is at risk.

Current limiter impedes power only when electronics first turn on. It does nothing during other and normal operation. Electronics power off only when voltage drops below a number. For example when incandescent bulbs dim to less than 40% or during 37 volts in Tom MacIntyre's example.

Power grids operate as was true 50 years ago. Because motorized appliances still have same needs. Meanwhile, electronics power supplies have become even more robust.

 

[Panos Beltis]

So...do I really need it? Because I'm gonna do an upgrade on my rig soon, and I will change to a better PSU thanks to many people here that recommended for me. And I'm afraid If a power outbreak is gonna fry my computer forever.

 

[westom]

So...do I really need it?

If anyone says yes or no, then you must consider it a lie. Because the answer did not say why.

Your question was answered many times with those 'whys'. Please read what was posted. All explained in layman simple concepts.

A voltage outbreak does not exist. Also obvious because the expression comes without numbers.

 

[GearUp]

... but what I would like to know is the quality of the UPS power when not on battery. The battery simply assures proper shutdown so data is not lost and system files are not corrupted and for me is typically within a minute..

View two waveforms in http://www.duke-energy.com/indiana-business/products/power-quality/tech-tip-03.asp . That is a UPS output voltage before (left) and when (right) in battery backup mode. A cleaner sine wave on the left - which is UPS output power when not in battery backup mode.

At risk is unsaved data. Only obsolete filesystems from 20+ years ago would cause saved data / OS file corruption when power was suddenly lost. A modern filesystem does not lose data due to a sudden power loss. Only unsaved data is at risk.

Current limiter impedes power only when electronics first turn on. It does nothing during other and normal operation. Electronics power off only when voltage drops below a number. For example when incandescent bulbs dim to less than 40% or during 37 volts in Tom MacIntyre's example.

Power grids operate as was true 50 years ago. Because motorized appliances still have same needs. Meanwhile, electronics power supplies have become even more robust.

Many SSDs do not lose any data now. The problem is if the drives are damaged. Other equipment can fail as well. Power strips provide some similar protection depending on quality that others seem to dismiss. The article you reference is far from definitive.
Mine self-test once a week which should avoid the issue where it is failing. The oscilloscope output says nothing about the condition of a new unit only that a better one fixes the issue. That may be but one incident suggests that two computers are plugged into one UPS (not clear). These are business units I assume. Battery backup is only provided on one side of mine for one computer. I have seen the non-battery side of mine does interfere with the other briefly which may not be a good indicator. I am advised to replace the battery within 3 years and find that correct but charging on my own is doing very well. The units themselves need to be repaired somewhat less often or replaced. That prompted me to consider dispensing with them but I'm doing OK so far.
I back up my files every two weeks or more often if I've been particularly productive. A surge protector can be a good alternative for most people. The specifications vary greatly for them.
There was much discussion several years ago about wave form regarding compatibility of power supplies but I've seen nothing about it since.

 

[gangrel]

And westom's comments only concern whether you'll completely fry your system, or at least parts thereof. That's not the whole story. A power interruption will take your system down. Even if it doesn't do damage, I really don't like the thought of losing whatever work I'm doing at the time.

Now, one point. I don't do gaming rigs. I doubt I've got any system drawing more than 250 or 300 watts...and probably less. That simplifies my UPS needs, and keeps the cost down...I'm not dropping $150 or $200 on a 1200 VA unit.

 

[westom]

And westom's comments only concern whether you'll completely fry your system, or at least parts thereof. That's not the whole story.

Previously posted:

UPS purpose: temporary and 'dirty' power so that unsaved data can be saved. If it does anything else, then the recommendation also cites a manufacturer specification number that says it does that.

UPS does not protect hardware or protect from a mythical power outbreak. Only hearsay or speculation claim that protection. That's the whole story.

 

[GearUp]

And westom's comments only concern whether you'll completely fry your system, or at least parts thereof. That's not the whole story. .

Previously posted:

UPS purpose: temporary and 'dirty' power so that unsaved data can be saved. If it does anything else, then the recommendation also cites a manufacturer specification number that says it does that.

.UPS does not protect hardware or protect from a mythical power outbreak. Only hearsay or speculation claim that protection. That's the whole story.

Your limited and confused evidence involves best case scenarios. Improperly maintained or old or cheap equipment comparisons are used both to support your claim and discount the opposition. I expect most of my power supplies will do fine without a UPS. However much existing equipment is likely to benefit and will do so for some time.
Consumers could easily test your beliefs at risk of warranty and their equipment but we need to trust in other sources and experience. This doesn't include the rather limited risk of lightning strikes. It is really a cost to benefit consideration.
The UPS does not, as some might conclude, provide power strictly through the battery. That would make no sense. It simply maintains a battery charge until it is needed. Less demand on the power supply protection circuitry is required due to UPS processing. If that is not true I am glad to hear of valid evidence.
You may also believe that cold temperatures do not cause "colds" which some limited research supports, but a recent study confirmed that in fact it has some affect.

 

[westom]

Less demand on the power supply protection circuitry is required due to UPS processing. If that is not true I am glad to hear of valid evidence

If a UPS does something useful, then its relevant specification number is cited. Even its surge protection number, typically hundreds of joules in numeric specs, is near zero protection. Inferior to what power strips and to what typically exists inside household appliances (including original and less robust PCs).

What is UPS processing? What number defines it? When not in battery backup mode, a computer connects directly to AC mains.

A typical UPS has one function: temporary and 'dirty' power during a blackout so that unsaved data can be saved. Even low voltages (called brownouts or sags) are already made irrelevant by what every computer PSU is required to do.


BTW, cold frequency increases in winter because those viruses survive longer, for example, on cold door handles. In warm weather, those viruses die (degrade) faster.

 

[GearUp]

My UPS insures damage ($75,000) resulting to components in the event of failure. The 354 joule rating is probably good enough for typical power fluctuations or distant lightning strikes.
My best surge suppressor (about $45) rates at 3570 joules and also insures ($250,000). Power strips can go much lower and the other specifications are mostly Greek to me. The actual valuable value appears to be 100 joules required.
While numerous people report the benefit of surge protection I doubt many collect on either type of unit. If the internals of a power strip are at all similar then I expect age will exclude many.
Like a power strip, a UPS provides more outlets but I don't use the master control except for shut down of the computer and monitor. What I like about that is when closely spaced power disruptions occur. I know this has been more hazardous to equipment.

The dirty power seems to be handled quite well by my power supplies but failure of the UPS is a complication.

I will continue to use the batteries until they no longer recharge and evaluate the pollution contribution at that time. I'm not impressed with the quality control. One UPS will document the outages and the other provides the beeps while both should do both.

While people would vaguely acknowledge "summer colds" they would insist that cold does not cause "colds". It is not simply a correlation, but a significant contributor and not all people have the same immune system.

 

[westom]

My UPS insures damage ($75,000) resulting to components in the event of failure. The 354 joule rating is probably good enough for typical power fluctuations or distant lightning strikes.
My best surge suppressor (about $45) rates at 3570 joules and also insures ($250,000).

If a warranty says anything useful, then GM cars are superior to Honda and Toyota. Reality - that warranty has multiple exemptions. Most learn the hard way rather than read its fine print. Warranty says nothing useful. Manufacturer specification numbers say much.

A 345 joule surge is so tiny that computer power supplies convert tiny energy into low voltage DC to safety power its semiconductors. Destructive surges (that can overwhelm robust protection inside all appliances) can even be hundreds of thousands of joules. Power strips with better numbers do not claim to protect from a type of surge that typically does damage. But 345 joules in a UPS (which means 115 joules and never more than 230 joules) really could not be smaller. It lists enough joules to claim 'surge protection' subjectively in sales brochures.. Those joules are virtually zero.

Many UPSes have receptacles described as surge protection only. Protection on those receptacles is similar to protection also provided to all other wall receptacles on that same circuit (ring). IOW that surge protection circuit is same whether connected to non-battery backup receptacles on a UPS or to other wall receptacle on the same house circuit breaker.

Surge protection from surges that typically do damage means one 'whole house' protector, properly earthed and rated at least 50,000 amps. A typical lightning strike is 20,000 amps. No protector should fail even on direct lightning strikes. 50,000 amps defines protector life expectancy over many surges.

Protection for each surge is defined by an item that harmlessly absorbs hundreds of thousands of joules. That is never the protector. That is what a protector connects a surge to - single point earth ground.

If any appliance needs protection, then all appliances also need that protection. Facilities that cannot have damage properly earth one 'whole house' protector. So that all surges - including direct lightning strikes - do not overwhelm robust protection already inside every appliance. This proven solution also protects that 345 joules UPS and that 3570 joules strip. Protection necessary because even fire, in rare cases, is another problem with plug-in protectors.

Again, protectors are simple science - a connecting device to what harmlessly absorbs surge. The 'art' of appliance protection is single point earth ground. All four words have electrical significance. A protector is only as effective as its earth ground.

This proven solution typically costs a homeowner about a pound per protected appliance. This is the only solution always implemented in every facility that cannot have damage. Some protection systems do not even have protectors. But every protection layer always features a low impedance connection to earth. Not wall receptacle safety ground - earth ground.

That is appliance protection not provided by a UPS. UPS is about temporary power. Protection, that even a UPS needs, is provided by that low impedance (ie less than 3 meter) connection to what defines surge protection - earth. Plug-in UPS has no such earthing connection.

 

[GearUp]

The $250,000 insurance for a surge protector has the same grounding as the UPS so I'm confused about the thousands of joules protection. It also mentions protection of microwave which falls outside of your previous definition of susceptible appliances. Is a surge protector device worth anything at all?
I doubt you respect Wikipedia but industrial protection is what was referenced regarding the 100 joule rating.
I've noted the lack of true grounding many times and in computers that only adds a case. That seems little more protection from the power supply. I must be missing the simple science of this and the preference for those power strips.
You are now referencing cumulative damage which is what I and others have tried to impress upon you regarding the UPS.

 

[westom]

I've noted the lack of true grounding many times and in computers that only adds a case. That seems little more protection from the power supply.

First, grounding a computer case says nothing about earth ground. As made obvious in an electrical concept that says why the expression reads 'less than 3 meters'. Safety ground was specifically noted as completely different from earth ground. An adjacent UPS has no earth ground. Earthing a computer can even make computer damage easier. Earth a potentially destructive surge; not its victim.

Second, what cumulative damage? Where did that come from? Why would a UPS do anything to avert cumulative damage?

If cumulative damage is problematic, well, a UPS output can be so 'dirty' as to degrade a connected power strip protector. UPS and protector manufacturers quietly recommend no power strip protector on a UPS output. Dirty power in battery backup mode can degrade that protector. Is that your cumulative damage? Because that same 'dirty' UPS output does not do cumulative damage to attached electronic appliances.

UPS does not protect hardware. A myth - cumulative damage - is not averted by a UPS. Where is any specification that defines cumulative damage? Numbers in datasheets from MOV manufacturers define cumulative damage by dirty UPS power. What cumulative damage are you discussing?.

Third, a $250,000 warranty, safety ground on a UPS, and what makes thousands of joules irrelevant have nothing in common. Effective protection for a computer, microwave and other household appliances is all done by the same solution proven over 100 years ago.

Fourth, surge protector devices that actually do protection are completely different from surge protector devices exposed by spec numbers as ineffective. Confusion abounds where two completely different devices share this common name. A superior and effective solution not only costs tens of times less money. It also does not have that big buck warranty - that exists to promote what is really tiny protection.

UPS has one function - temporary and 'dirty' power during a blackout. UPS does not protect hardware - as demonstrates by no specification numbers to justify that speculation.

 

[Panos Beltis]

So many confusing and overanalyzed answers, I can't even think what to do, I asked a simple question, do I need it, or not? Please, I'm not a native speaker and my english are bad, I don't know anything - as you already might have guessed - but I need to know...do I really need a UPS? Or not? I'm going to upgrade my rig soon with a gtx970 4gb edition, I want to be sure that my computer won't get destroyed by a power outbreak/smth like that anyways. I read everything here and I still don't get what you mean, they are over analyzed and confusing for me.

 

[westom]

... I asked a simple question, do I need it, or not?

A UPS to do what? Stop buglars? Protect from tornadoes? Long before anyone can answer, first define what you must protect from.

Written so many times: A UPS is temporary and 'dirty' power to protect unsaved data from a blackout. That simple. Its specifications do not claim to do anything else.

Other anomalies include frequency variation, reverse polarity, RFI, floating neutral, open safety ground, surges, harmonics. EMC/EMI, and power factor. A typical UPS also does not protect from or avert those anomalies. Which anomaly do you fear?

Described was another and much less expensive solution from destructive transients. Do you want protection from something that typically damages hardware - a surge? That proven solution was provided. But first, define which anomaly causes you fear. A solution cannot be recommended until the problem is first defined.

Not existing is an anomaly called a power outbreak.

 

[jsmithepa]

I'm not a native speaker and my english are bad

Well then we have a communication problem don't we.

I don't know anything

Unfortunately if you don't know anything, and there are no forums with your language to help you, WE ARE STUCK.

So, BUY IT, since you don't understand the risk/reward/technical.

 

[GearUp]

I've noted the lack of true grounding many times and in computers that only adds a case. That seems little more protection from the power supply.

First, grounding a computer case says nothing about earth ground. As made obvious in an electrical concept that says why the expression reads 'less than 3 meters'. Safety ground was specifically noted as completely different from earth ground. An adjacent UPS has no earth ground. Earthing a computer can even make computer damage easier. Earth a potentially destructive surge; not its victim.

Second, what cumulative damage? Where did that come from? Why would a UPS do anything to avert cumulative damage?

If cumulative damage is problematic, well, a UPS output can be so 'dirty' as to degrade a connected power strip protector. UPS and protector manufacturers quietly recommend no power strip protector on a UPS output. Dirty power in battery backup mode can degrade that protector. Is that your cumulative damage? Because that same 'dirty' UPS output does not do cumulative damage to attached electronic appliances.

UPS does not protect hardware. A myth - cumulative damage - is not averted by a UPS. Where is any specification that defines cumulative damage? Numbers in datasheets from MOV manufacturers define cumulative damage by dirty UPS power. What cumulative damage are you discussing?.

Third, a $250,000 warranty, safety ground on a UPS, and what makes thousands of joules irrelevant have nothing in common. Effective protection for a computer, microwave and other household appliances is all done by the same solution proven over 100 years ago.

Fourth, surge protector devices that actually do protection are completely different from surge protector devices exposed by spec numbers as ineffective. Confusion abounds where two completely different devices share this common name. A superior and effective solution not only costs tens of times less money. It also does not have that big buck warranty - that exists to promote what is really tiny protection.

UPS has one function - temporary and 'dirty' power during a blackout. UPS does not protect hardware - as demonstrates by no specification numbers to justify that speculation.

At this point I believe you must be stating that the safety ground leads to the house circuit breaker which would trip before the UPS or power supply circuit breakers for high surges. I had expected not actual redundancy but lesser irregular power to be .moderated by the UPS and Power Supply. The protections and power management features in modern power supplies are so extensive I forgotten what they are. I've had little success with much of the management for a desktop. Even a $10 power strip will as you indicate provide more information, but likely different specifications from other units.

Devices such as telephone/modem specify plugging directly into the wall. People will in this day usually need some form of power strip to attach multiple devices but the draw is less than it had been. Nevertheless they always discourage daisy-chaining of power plugs. Therefore it wouldn't occur to me to use a surge protector strip before or after a UPS. My experience in using an incandescent lamp on the surge side of the UPS was a disappointment but I only use items that are always on for any plugs on either side except for the computer and monitor.
My surge protector strip actually has a very long thick wire, possibly 3 m, that I would never then use as the source for additional splitting. The proliferation of wall-warts, in lieu of an in-line device is unfortunate and makes the issue worse.

It is sad to see the UPS box include a printer as an attachable and controllable device. Multipurpose printers and certainly laser printers have no business being on them. The "Up to 70 minutes" battery time could be discretely placed in the manual instead of prominently on the box. It provides for comparison of their offerings but is likely to result in an undersized unit purchase. 15-20 minutes is about it for a 65W computer and 100W monitor and I'd say too long to exploit in practice. The 750 model number of my unit refers to 750VA and 450 watts.

Without the UPS the power supply will always have to handle the power event itself, sustaining whatever cumulative damage there might be. With a UPS it only has to handle power that isn't true sinusoidal for a few minutes a year. If there are other actual benefits it should be, as you state, clear what they are. The conversion from AC to DC to AC observed in monitors I expect is at least as dirty. I'm a fan of DC myself, which says something of my lack of knowledge. The rudimentary Windows battery software can always serve in place of the UPS software, which does not reflect well on the design. It does generally work and a replacement will likely be similar.
Edit: The self-test done by the UPS only checks the battery backup circuitry as far as I know at this point.

 

[westom]

Without the UPS the power supply will always have to handle the power event itself, sustaining whatever cumulative damage there might be. With a UPS it only has to handle power that isn't true sinusoidal for a few minutes a year. If there are other actual benefits it should be, as you state, clear what they are. The conversion from AC to DC to AC observed in monitors I expect is at least as dirty.

Destructive transient is called a current source. That means a current is incoming to a UPS at the same time that current is incoming to and outgoing of attached appliances. Voltage only exists when something foolishly tries to block or stop that current. Voltage increases as necessary to blow through anything that might try to stop it. An IEEE brochure shows a surge incoming to a plug-in surge protector while also outgoing into a nearby TV. Therefore (specifically for the OP) an 8000 volt surge is destructively inside that TV because a protector device is too close to the TV and too far from earth ground.

That current was not earthed BEFORE entering a building. An adjacent protector simply gave it more paths to find earth destructively via a nearby TV. Once inside, then nothing (no power strip or UPS) will stop that hunt for earth ground destructively via appliances. If that current is incoming to a UPS, that same current is outgoing via attached appliances - simultaneously. It is electricity. Both an incoming and outgoing path must exist simultaneously. That current does not enter a UPS and just stop.

Cumulative damage occurs on parts as defined in datasheets. MOVs suffer cumulative damage as defined by manufacturer charts. Nothing in an electronic power supply suffers cumulative damage. Either parts remain intact or fail catastrophically. Designs are so refined that some parts intentionally fail shorted while others fail open. But in every case, those parts do not suffer cumulative damage. Either parts remain fine or fail catastrophically.

UPS does not avert a surge. Best circuits already inside electronic appliances make tiny surges irrelevant. Circuits convert that surge into well regulated DC electricity to harmlessly power semiconductors. Damage only occurs when a surge current is so large as to even blow through more robust protection inside appliances.

Specifically for the OP: well proven solution that averts that transient (that may occur once every seven years) is proper earthing. With a low impedance (ie less than 3 meter) connection to earth; either via a hardwire (best protection) or via a 'whole house' protector (next best protection).

Specifically for the OP: a UPS does not have that low impedance (less than 3 meter) connection to earth and does ot even claim to protect hardware from that destructive transient. Even a UPS must be protected by the above and properly earthed 'whole house' protector.