Battery Backup or Surge Protector?

[westom]

@westom
I said "components" in the context of computers and electronics. The oxidisation of conductors is when theres not enough but it doesnt mean that oxygen provides all the energy, its just that there is a voltage but no charge to fill it so it encourages oxidisation. (show me some physics facts that say otherwise). a low voltage damages electronics when the conductors oxidise and become a resistor which changes the voltage received which in the digital world is a massive problem in communication as the components have a preset voltage range.

I am not sure what you are discussing. Since oxidation happens between dissimilar metals (ie aluminum and copper). Sometimes a voltage with minimal current reduces this. But that does not exist inside electronics. Again, if that was a problem, then numbers from a datasheet for that part defines it.

Conductors oxidize due to excessive current. For example, those Molex connectors that connect PSU to motherboard are rated for 6 amps. PSUs went from 20 to 24 pin connectors since no conductor must oxidize. Apparently discussing oxidation reduction (which is irrelevant to electronics except sensor devices such a thermal couples) or oxidation due to excessive current? These are two completely different effects. If relevant, then an appropriate (and provided) datasheet defines it.

 

[westom]

@westom
If the light dims when something power cycles it doesnt always indicate a wiring problem. For example try running a circuit from a generator instead of the mains, when you add load to it a connected light bulb dimms for less than a second as the generator increases in speed to provide the needed energy.

Tiny generator needs time to increase current. Delay due to mechanical parts; not due to electrical parts. Completely irrelevant to electricity from a multi-gigawatt generator that sees a near zero load increase. Power cycling a major appliance means no voltage reduction exists. Dimming is a classic example of a poorly wired building. Not due to a nuclear power plant responding to a sudden current demand from a washing machine.

When can dimming indicate a serious threat? If a street transformer has a defect, then that transformer (not a nuclear power plant) results in voltage variations. Another example is the classic 'open neutral'. Dimming due to a household defect. Dimming would not happen when a transformer and wiring is properly installed.

If you know better, then put forth numbers. Perspective (numbers) are essential to justify what we never see. That generator example is bogus. Power plants do not respond (with delay) because a major appliance was turned on. Generators do.

 

[westom]

@westom I remember when using faulty electrics that if i switched it on or an inrush current too high that the circuit would break which actually prevented sparking or fires

You have completely confused what a fuse does with hardware damage. A fuse does not protect hardware. A fuse blows after that damage to avert fire - to protect human life.

Maybe you are discussing an inrush current limiter? Unfortunately you do not know many functions and components that must exist in electronics. Apparently assumed an inrush current causes sparking or fires. High voltage - not high current - causes sparking. Observation without first learning these fundamentals is how urban myths get created.

Galvanic isolation must exist in that power supply. Apparently you are discussing an event that must not exist. Galvanic isolation is one of so many reasons why electronics have superior transient protection - even rated for thousands of volts. Surge protection is about protecting appliances from a rare anomaly that can overwhelm that required internal protection.

Moving on to disk drives - modern disk drives do as I posted and as disk drives did in the 1960s. Again, urban myths exist because so many do not know how drives work - all disk drives. All disk drives have plenty of time after power off starts to finish all writes - as was true when disk drives were slower and demanded hundreds of watts. (Today's drives operate on well less than 10 watts.)

File systems once had a problem. If the file write did not complete, then that file was lost AND an original file was lost. A software problem. That software problem was eliminated over 25 years ago. Drive hardware never had a power loss problem. OS software solved that other weakness well over 20 years ago.

Fears of blackouts and brownout damage exist due to speculation; without first learning how hardware really works. Often found in bogus (subjective) articles written by others trained in hearsay. If not defined with specification numbers, then it is clearly suspect - often wrong.

 

[System Error Message]

This thread http://superuser.com/questions/113113/why-are-brownouts-so-harmful explains what happens during a brown out and it will happen in places that have their power stations running to their limits.

File systems still have the problem of losing files. I have lost files in RAID arrays just because the hard drive was equalising itself (the drive is a high performance one and loses performance when restricted in mountings and intel assumed the drive was bad and dropped it). You should never assume softwares dont have problems and i have had lost files various times with a single hard drive in which checkdisk had to run just to recover them or part of them.

The reason to fear of blackouts and brownouts are real, simply do a bit of search and you will find many who have experienced it in homes and businesses. Some cities have a power problem and some dont. Again dont restrict yourself to the US only. Ask a proper electrician as well. Brownouts and blackouts are very real and many of us on the forum has experienced it. I used to have a lot of blackouts that even the UPS would fail because of the stress on it that even changing the battery would do no good. Infact intel has a specification on this which you can read from the link i provided if you say it is hearsay unless you're saying that intel is also using fear mongering and speculation or bogus claims in their specifications and products. Ofcourse i dare you to input 60V into your computer PSU if you say there is no such thing as a brown out and run your PC the PSU's rated input . I also invite you to go to countries that do not have proper infrastructures where blackouts are common and brownouts with flickering lights frequent. These cases really do exist.

Stop assuming @westrom , bring forth the evidence i asked. Where are the links/references to the scientific papers to disprove my case? If what people say is bogus surely there is a scientific paper to prove it. Electricity can be dangerous thats why protection exists and there has been injury and damage to equipment from various different electrical problems.

 

[System Error Message]

Tiny generator needs time to increase current. Delay due to mechanical parts; not due to electrical parts. Completely irrelevant to electricity from a multi-gigawatt generator that sees a near zero load increase. Power cycling a major appliance means no voltage reduction exists. Dimming is a classic example of a poorly wired building. Not due to a nuclear power plant responding to a sudden current demand from a washing machine.

When can dimming indicate a serious threat? If a street transformer has a defect, then that transformer (not a nuclear power plant) results in voltage variations. Another example is the classic 'open neutral'. Dimming due to a household defect. Dimming would not happen when a transformer and wiring is properly installed.

If you know better, then put forth numbers. Perspective (numbers) are essential to justify what we never see. That generator example is bogus. Power plants do not respond (with delay) because a major appliance was turned on. Generators do.

Incase you forgot power to the grid and mains is from generators that vary in speeds based on loads. While 1KW to 1MW may seem insignificant but the distance from the power station is what causes the delay in change. Maybe when i have time i'll show some scientific papers.

 

[sfx2000]

Not due to a nuclear power plant responding to a sudden current demand from a washing machine.

Honestly, I think this thread has jumped the shark...

Just saying..

 

[westom]

Incase you forgot power to the grid and mains is from generators that vary in speeds based on loads. While 1KW to 1MW may seem insignificant but the distance from the power station is what causes the delay in change. Maybe when i have time i'll show some scientific papers.

Please provide a scientific paper that says that.

When you power a major appliance, lights in all nearby homes dim as power plants take time to increase power to the neighborhood? That is what you now claim. Don't forget to provide these numbers since everything cited previously was classic junk science - missing reasons why with numbers.

 

[westom]

This thread http://superuser.com/questions/113113/why-are-brownouts-so-harmful explains what happens during a brown out and it will happen in places that have their power stations running to their limits.

Do you read your own citations?

... the power supply will still run normally even when a noticeable brownout occurs.

Brownouts cause no internal voltage variations. How do AC mains brownouts do damage when internal parts never see voltage variations? Please read your own citation. Brownouts do not cause hardware damage.

Voltage can drop so low that incandescent bulbs dim to 40% intensity. Voltage even that low is a perfectly normal voltage. Computers are required to be that robust.

What happens when voltage drops lower? Manufacturer datasheets demonstrate no hardware damage from any voltage (ie from 20 volts down to -0.5). No part has been identified at risk because no part is damaged by low voltage. Brownouts do not cause hardware damage. Even System Error Message concedes that.

That author quotes from ATX Standards that say what was standard long before PCs existed:

The power supply shall contain protection circuitry such that the application of an input voltage below the minimum specified in Section 3.1, Table 1, shall not cause damage to the power supply.

When voltage drops, electronics run normally. When voltage drops lower, electronics simply power off. Another standard even defined all low voltages with an expression in all capital letters: No Damage Region. All voltages down to zero cause no hardware damage.

Move on to data damage. superuser.com citation notes what happens when a computer is constructed by the technically naive. Such as computer assemblers who select hardware from a 'tier' rather than learn relevant hardware numbers.

The problem ... which can only occur if both the power supply and motherboard fail to realize the problem, and continue to attempt to operate.

Then data errors occur. This software problem - directly traceable to a naive computer assembler - may write corrupted files to a disk. Data corruption directly traceable to what a computer assembler did.

None of this is explained in internet articles that make claims devoid of any numbers. Unfortunately too many computer assemblers learn from a first thing read. Only wild speculation claims brownouts can do hardware damage. Articles without spec numbers are best called junk science.

When incandescent bulbs dim to well below 40% intensity, then hardware must identify an excessively low voltage and power off. That citation even cites a relevant wire - Power Good. Technically naive computer assemblers would ignore what he wrote. He explains why a computer powers off when it cannot maintain DC voltages to all parts. Power Good and a response from a 'power controller' is why some computers power off suddenly without even a BSOD. And without corrupted disk data.

Power Good is not about hardware protection. Because all voltages down to and even below zero cause no hardware damage. Power Good is about averting data corruption. It even existed in the original IBM PC. That basic computer function existed long before PCs existed.

System Error Message's corrupt disk data was explained by his citation. Systems constructed using parts from 'tiers' means a PSU and motherboard failed to power off when voltage dropped too low. System Error Message blames a brownout rather then the reason for corrupted disk data: a computer assembler who selected parts from 'tiers' rather than learn how electricity works.

UPS has one purpose. Temporary and 'dirty' power so that unsaved data can be saved. UPS does not protect hardware. Low voltage never damages properly constructed hardware. Low voltages might corrupt disk data when hardware is defective (ie selected using 'tiers'). If hardware does not shutdown when voltage drops too low, then software errors can occur. But again, that is not hardware damage as so many wildly speculate.

Brownouts only cause hardware damage when one believes the first thing told by articles devoid of numbers. Also called junk science. In a real world, no part was identified at risk due to brownouts. Not one. For one simple reason: brownouts do not damage electronics.

 

[coxhaus]

The problems I have seen in the past with disk RAIDs is the controller cards have cache and this cache is not flushed during a power out then the disk drive never gets a chance to write the data because the software is being fooled due to the write cache.

Some advanced controller cards actually have battery backups on the controller card to save data so in the event of a power outage the disk array will not be corrupted.

 

[westom]

Some advanced controller cards actually have battery backups on the controller card to save data so in the event of a power outage the disk array will not be corrupted.

If one drive is updated and another is in process when Power Good interrupts operation (incandescent bulb dims to less than 40% intensity), then the next paragraph applies.

If a file on one drive is not completely written, then that filesystem falls back onto a previously saved copy. What happens if one drive is still using an older copy? Data is not corrupted. Different files exist. RAID software must correct a filesystem problem - as it is designed to do.

Of course, this says nothing about hardware damage. A UPS was recommended to protect hardware. It does not. A UPS is only temporary and 'dirty' power so that even RAIDs can finish their writes. A UPS is only for protecting unsaved data. It does not protect electronic hardware as others claim and believe.

 

[System Error Message]

I put different contexts into the same post so brownouts and data loss wasnt what i was relating. What i was saying relating to data loss was software and hardware expectations are different. My RAID array doesnt use cache but it uses performance drives and when a drives slows down to accommodate the vibrations it thinks the drive has gone bad or has errors and gets dropped and i have lost unimportant files because the array got dropped which so happened to have my OS on it. Free mounting the drives solved the issue for me but what i was saying despite the robust design of hardware and software, data loss still occurs even if they have been made immune to power problems.

Brownout damage means that the hardware has suffered a brownout. i think @westom has never experience electrical problems before but a number of experts on this forums have and have clients who had as well. Brownouts are real but are dangerous to electronics which do not have a proper PSU (a lot of them really dont, dont just look at PCs). For example one reason why some routers hang is because they have a bad PSU (same for PC too).

I think @thiggins needs to summerise and lock this thread before more trolling happens here. There isnt fear mongering and electrical problems are real and not everyone lives in a city with ideal power conditions.

Simply put, some protection is better than no protection and precautions should be used. the UPS converts the battery power from DC to AC. A battery's output is stable and has a very stable voltage drop as it goes flat so the output of a UPS from the battery is actually stable and good till the battery goes flat or gets overloaded as batteries have a limited amp output.

What @westom is saying is that we do not need surge protectors, UPS, fuses, or any form of protection we can plug all appliances to a single plug because nothing bad is going to happen and that surge protectors dont protect against surges or some electrical problems or make a difference, fuses are useless, that computer PSUs and electronics will protect against all oddities. UPS provides "dirty" power (see the claims of florescent bulbs being dirty and unhealthy aside from the mercury) but how is there such a thing as clean or dirty electricity (how are electrons a pollutant?) and What he is also saying is that there are no risks from or no such thing as lightning, power station problems, overloaded infrastructures which are real problems globally and places where infrastructure is poor and that it takes no time for electrons or their potential to travel from one place to another (no electrical lag such as when you switch on a room from when the circuits connect to when the bulbs go on) so in that logic even light is instant and there is no latency when using fibre optics because it is instantaneous so technically gaming from the otherside of the world isnt laggy. I bet he wont even quote this paragraph.

I have had several high end PSUs with protective features spark and fail and in one case even caused the components to spark and blow and it blew a GPU (it blew a hole in a section of the GPU). Thankfully asus was kind enough to replace it with one of better yield. I have personally witnessed outdoor wifi APs fully protected with high end and proper grounding equipment fail in a lightning heavy area. Some areas have a lot of lightning and humidity and i invite @westom to live in these areas without using any electrical protection for his equipment, where the infrastructures are also poor too.

 

[coxhaus]

If one drive is updated and another is in process when Power Good interrupts operation (incandescent bulb dims to less than 40% intensity), then the next paragraph applies.

If a file on one drive is not completely written, then that filesystem falls back onto a previously saved copy. What happens if one drive is still using an older copy? Data is not corrupted. Different files exist. RAID software must correct a filesystem problem - as it is designed to do.

Of course, this says nothing about hardware damage. A UPS was recommended to protect hardware. It does not. A UPS is only temporary and 'dirty' power so that even RAIDs can finish their writes. A UPS is only for protecting unsaved data. It does not protect electronic hardware as others claim and believe.

I think the problem does not exist with the drive, it is the software fooling the drive with the write cache. The system thinks everything is good, the write operation only happened in cache not on the drive. This is not a default setting. Write cache can be turned on to the speed up the system. Not using write cache is called a write through operation which takes priority and will slow disk operations. If you turn on write cache you need to have an UPS or your file system will become corrupt with power outages.

 

[System Error Message]

a UPS isnt just so unsaved files dont vanish, a UPS is meant to provide power during a blackout and there are various types. A datacenter uses a combination of both battery and generator where the battery is used for a few seconds while the generators turn on to allow it to work even when there is no power. A cheaper UPS with a standard desktop (IGP, mid range CPU, etc, no dedicated GPU) would usually work for 20-30 minutes meaning you can still work during blackouts or when power comes on and off. Some people implement a UPS to give time for things to shut down while for some it is used to keep operating. For example you could use a UPS with a long enough duration to disconnect the mains during a thunderstom and just use the battery to keep everything running.

Optional extras a UPS can have is some protection and smoothing where the battery is also used as a capacitor (or it may also have lots of capacitors).

When a protector says a 3mm gap it means that when the circuit breaks and goes to ground theres the wire to ground and 3mm gap of air. While the voltage (if it goes through both paths) would stay the same (the voltage after the air gap would dop significantly because of series circuit with the wire after air) the current going through each is proportional to the difference. Houses usually have 2 layers of circuit breakers or fuses, one main and one for each room so the risk is greater reduced and im not sure if it is still done but there should also be 2 to 3 layers of fuses for houses as well, one or 2 from main and one for each room. A surge protector wouldnt say a 3mm gap rather a circuit breaker would as it is to explain the distance between the conductor of one circuit and another (similar to the distance between wires in a motherboard where the term tolerance is used).

A surge protector uses something totally different which is explain in a link in one of my posts far behind (and not an air gap) in which surge protectors were compared and their specs explained and it can be clearly seen that a surge protector does work, the problem is that people buy cheap ones that just have paintings on the case and nothing inside (this is why specs matter). The limited lifespan of a surge protector is why they are tested often because cheaper ones dont indicate when the protection is gone.

The other reasoning to protection isnt just homes but in industry and other applications which do not follow the 120V/240V standard and for other things too such as research in places that are harsh with electrical storms.

 

[stevech]

I think that here, the simple topic of small UPSes is being over-analyzed!
Value your time? Your data? Just use a UPS!

 

[westom]

Value your time? Your data? Just use a UPS!

... to save data that was not yet saved. It does nothing for saved data. It does nothing to protect hardware.

UPS also eliminates waiting for a computer to reboot (maybe once every two years?) after a blackout.

A properly earthed 'whole house' protector is essential for protecting hardware (and other household appliances) from a potentially destructive anomaly. These occur maybe once every seven years. Best protection costs about $1 per protected appliance. Is necessary to protect even plug-in protectors or a UPS.

Reams of previous posts are urban myths exposed and disposed. UPS for data. 'Whole house' protector for hardware.

 

[coxhaus]

Do all UPSs have replaceable batteries or are some them disposable?

I think I would want one with replaceable batteries.

 

[stevech]

All I've seen have replaceable.

 

[stevech]

... to save data that was not yet saved. It does nothing for saved data. It does nothing to protect hardware.

UPS also eliminates waiting for a computer to reboot (maybe once every two years?) after a blackout.

A properly earthed 'whole house' protector is essential for protecting hardware (and other household appliances) from a potentially destructive anomaly. These occur maybe once every seven years. Best protection costs about $1 per protected appliance. Is necessary to protect even plug-in protectors or a UPS.

Reams of previous posts are urban myths exposed and disposed. UPS for data. 'Whole house' protector for hardware.

Whole house surge protector is ill-advised. Arguably ineffective. High cost to purchase. High labor cost for electrician to install.

 

[CaptainSTX]

Whole house surge protector is ill-advised. Arguably ineffective. High cost to purchase. High labor cost for electrician to install.

Square D makes whole house surge protectors for their panels that use two blank slots on the main panel. Has a single connection to neutral (which at the panel is bonded to the ground ). Takes less than 15 minuted for a DYI if you have the minimum skills to replace a circuit breaker.

 

[coxhaus]

I thought I had seen a UPS that used an ION battery and I wondered whether it was replaceable. It was small but would work for 1 workstation.

I just saw Fry's has an APC 1500VA UPS for $129 with rebates. I guess Black Friday.