Discussion on UPS and power surge protectors

[westom]

Moderator note: This thread has been split from another thread to keep it on topic.

I guess I'll just get a normal surge protector.

If your computer needs protection, then so does the furnace, dishwasher, clock radios, dimmer switches, the GFCIs so important for protecting human life, every TV, telephone, etc. What must be protected during a surge? Smoke detectors.

Only protection in every facility that cannot have damage is earthing a 'whole house' protector. Either a surge is inside hunting for earth destructively via appliances. Or one 'whole house' protector earths that energy harmlessly outside. Nothing new. This is how protection was always done; even more than 100 years ago.

Any protection adjacent to an appliance is already better inside the appliance. Your concern is a surge that can overwhelm that existing and superior protection. That was always a 'whole house' protector connected low impedance (ie 'less than 10 feet') to single point earth ground. Not safety ground (as in receptacles). Earth ground.

These superior protectors are provided by more responsible companies including ABB, Siemens, Leviton, General Electric, Intermatic, Square D, Polyphaser, or Ditek ... to name but a few. All names that any guy would know for better integrity. A Cutler-Hammer solution sold in Lowes and Home Depot for less than $50. This less expensive solution is also the superior solution. But again, first learn what all professionals require so that even direct lightning strikes cause no damage.

Protection is always about where hundreds of thousands of joules dissipate. Harmlessly outside. Any protector is only a connecting device to what does protection - single point earth ground.

BTW, read spec numbers for that UPS. Near zero. Just enough above zero to claim 100% protection ... subjectively. Better answers always include spec numbers.

 

[bud--]

Excellent information on surges and surge protection is at:
http://www.lightningsafety.com/nlsi_lhm/IEEE_Guide.pdf
- "How to protect your house and its contents from lightning: IEEE guide for surge protection of equipment connected to AC power and communication circuits" published by the IEEE in 2005 (the IEEE is a major organization of electrical and electronic engineers).
And also:
http://www.eeel.nist.gov/817/pubs/spd-anthology/files/Surges%20happen!.pdf
- "NIST recommended practice guide: Surges Happen!: how to protect the appliances in your home" published by the US National Institute of Standards and Technology in 2001

The IEEE surge guide is aimed at people with some technical background.

If your computer needs protection, then so does the furnace, dishwasher, clock radios, dimmer switches, the GFCIs so important for protecting human life, every TV, telephone, etc. What must be protected during a surge? Smoke detectors.

The NIST surge guide suggests that most equipment damage is from high voltage between power and signal wires - like power and cable at a TV.

Only protection in every facility that cannot have damage is earthing a 'whole house' protector.

Service panel protectors are a real good idea.
But from the NIST guide:
"Q - Will a surge protector installed at the service entrance be sufficient for the whole house?
A - There are two answers to than question: Yes for one-link appliances [electronic equipment], No for two-link appliances [equipment connected to power AND phone or cable or....]. Since most homes today have some kind of two-link appliances, the prudent answer to the question would be NO - but that does not mean that a surge protector installed at the service entrance is useless."

Service panel suppressors do not, by themselves, prevent high voltages from developing between power and phone/cable/... wires. (An example of where a service panel protector would provide no protection is the IEEE surge guide example starting page 30.)

The author of the NIST surge guide looked at the surge current that could come in on residential power wires. The maximum with any reasonable probability of occurring was 10,000A per wire. That is based on a 100,000A lighting strike to a utility pole adjacent to the house in typical urban overhead distribution.

Service panel protectors with far higher ratings are readily available. They are very likely to protect anything connected only to power wires from a very near very strong lightning strike. They may or may not protect equipment connected to both power and signal wires.

Any protection adjacent to an appliance is already better inside the appliance. Your concern is a surge that can overwhelm that existing and superior protection.

Nonsense.

These superior protectors are provided by more responsible companies including ABB, Siemens, Leviton, General Electric, Intermatic, Square D, Polyphaser, or Ditek ... to name but a few.

All these "responsible" manufacturers except SquareD and Polyphaser make plug-in protectors and say they are effective. Westom says plug-in protectors don't work.

SquareD says for their "best" service panel suppressor "electronic equipment may need additional protection by installing plug-in [protectors] at the point of use."

BTW, read spec numbers for that UPS. Near zero. Just enough above zero to claim 100% protection ... subjectively.

It is easier to find high ratings on plug-in protectors than UPSs.

The author of the NIST surge guide investigated how much energy might be absorbed in a MOV in a plug-in protector. Branch circuits were 10m and longer, and the surge on incoming power wires was up to 10,000A (the maximum that has any reasonable probability of occurring as above). The maximum energy at the MOV was a surprisingly small 35 joules. In 13 of 15 cases it was 1 joule or less. (Energies werewith US wiring and may be different for other service schemes.)

Plug-in protectors with much higher ratings are readily available. High ratings mean long life. A plug-in protector with high ratings, wired correctly (as below), is very likely to protect from a very near very strong lightning strike

If using a plug-in protector all interconnected equipment needs to be connected to the same protector. External connections, like coax also must go through the protector.

As explained in the IEEE surge guide (starting page 30) plug-in protectors work primarily by limiting the voltage from each wire (power and signal) to the ground at the protector. To do that all wires must go through the protector.

Warranties on protected equipment will require that all wires go through the protector.

 

[westom]

That makes sense.

To further define what others have posted. NIST is quite blunt about what an adjacent protector does.

A very important point to keep in mind is that your surge protector will work by diverting the surges to ground. The best surge protector in the world can be useless if grounding is not done properly.

Not just 'useless'. We engineers even traced surge damage BECAUSE a protector was too close to electronics and too far from earth ground. It bypassed protection inside the computer. Connected that surge destructively to earth via its motherboard and NIC interface.

So its warranty is routinely not honored; as so many cautioned. And so many others reported. It is a profit center; not surge protection.

Insurance is useful when a carrier is required by law to honor those claims. Power strip protectors have numerous exemptions. And are not subject to insurance laws. Anyone can read its numerous fine print exemptions.

If it does something useful, then a manufacturer's numeric spec number is posted. Nobody can post what does not exist. Some will assume a warranty means it does protection. Well, the best warranty in the auto industry is GM's. So that proves GM products are better than Honda, Toyota, and Hyundai? Of course not. A best warranty usually identifies lesser products - and a warranty full of exemptions.

 

[bud--]

NIST is quite blunt about what an adjacent protector does.

The NIST certainly is quite blunt.
Immediately following westom's quote from the NIST surge guide is a list of surge protectors that can be used.
Number 6 is "Plug-in...The easiest of all for anyone to do. The only question is 'Which to choose?'"

What else does the NIST surge guide says about plug-in protectors?
They are "the easiest solution".
And "one effective solution is to have the consumer install" a multiport plug-in suppressor.

The IEEE surge guide? It has 2 detailed examples of protection. Both use plug-in protectors. That is after a long discussion of them.

We engineers even traced surge damage BECAUSE a protector was too close to electronics and too far from earth ground.

"We engineers" were incompetent.

Even "engineers" can RTFM. Any competent manufacturer will tell you the same thing I said:
"If using a plug-in protector all interconnected equipment needs to be connected to the same protector. External connections, like coax also must go through the protector."

So its warranty is routinely not honored; as so many cautioned.

"So many" did not caution.

Anyone can read its numerous fine print exemptions.

My TV died yesterday. Musta been a surge.
And - all wires to a protected set of equipment MUST go through the protector.

Fine print is necessary.

If it does something useful, then a manufacturer's numeric spec number is posted. Nobody can post what does not exist.

Complete nonsense.

I have seen many specs posted for westom. He always ignores them. Apparently he knows that plug-in protectors do not work so specs can not exist.

For real science read the IEEE and NIST surge guides. Excellent information on surge protection. And both say plug-in protectors are effective.

 

[westom]

"We engineers" were incompetent.

bud is a sales promoter for power strip protectors. Posting insults is part of his game.

If honest, he could post manufacturer specifications that define protection. He cannot. He said he once did. Good. Then he can post those same numbers here. He won't. He cannot post numbers he never posted.

His products only claim to protect from surges that typically cause no damage. Then avoids hard facts in his NIST citation:

You cannot really suppress a surge altogether, nor "arrest" it (although your utility uses devices they call "surge arresters" to protect their systems). What these protective devices do is neither suppress nor arrest a surge, but simply divert it to ground, where it can do no harm. So a name that makes sense would be "surge diverter" but it was not picked

Since his products have no earth ground, then he says earthing is irrelevant. Professional organization all say different. Ironcially, his citation contradicts him

A very important point to keep in mind is that your surge protector will work by diverting the surges to ground. The best surge protector in the world can be useless if grounding is not done properly.

But then we engineers who did this stuff were careful about making earth even better. Since he never did this stuff, then he attacks with the above quoted cheapshot and other insults. Even his own citation says his products are 'useless'.

He only joined this group yesterday (as he has so many others over most of the past decade) to attack me. It is his job.

So where are these numbers that claim protection? He never posts numbers that claim protection from each type of surge. His products only claim to protect from a type of surge made irrelevant by superior protection already inside every appliance. How many more cheapshots will he post? He will not even provide that number. Even his citations show damage to appliances when a protector is not properly earthed.

 

[bud--]

bud is a sales promoter for power strip protectors.

My only association with the surge protection industry is I am using some surge protectors.
If westom had valid technical arguments he wouldn't have to lie.

Posting insults is part of his game.

Westom is insulted by the IEEE and NIST.

If honest, he could post manufacturer specifications that define protection. He cannot. He said he once did. Good. Then he can post those same numbers here. He won't. He cannot post numbers he never posted.

I have posted specs. Others have posted specs.A 10 year old can find specs. Always ignored.

Westom just repeats the same nonsense already debunked.

If plug-in protectors do not work perhaps westom can answer some simple questions:
- Why do the only 2 detailed examples of protection in the IEEE guide use plug-in protectors?
- Why does the NIST guide says plug-in protectors are "the easiest solution"?
- Why does the NIST guide say "One effective solution is to have the consumer install" a multiport plug-in protector?
- Why does the NIST guide say "Plug-in...The easiest of all for anyone to do. The only question is 'Which to choose?'"
- Why do westom's "responsible companies" make plug-in protectors?
- Why does "responsible company" SquareD says "electronic equipment may need additional protection by installing plug-in [protectors] at the point of use"?


For real science read the IEEE and NIST guides. Excellent information on surge protection. And both say plug-in protectors are effective.

Then read the sources that agree with westom that plug-in protectors do NOT work. There are none.

 

[hafa]

Westom just repeats the same nonsense already debunked.

Westom does seem to have passionate opinions on the subject of UPS units and earth grounding. See here and here. He and I went round and round on the first thread.

A good UPS serves several purposes, including voltage regulation (something that a surge protector and your PSU does not do) and transient surge suppression (especially the more common surges due to motor startups). Westom is correct in stating that a whole house protection system is the most effective for high-voltage surges from lightning strikes, but they do nothing to clean up power surges from internal sources such as motors or refrigeration.

For the OP, I'll echo the sentiments of others here: Get a decent UPS to protect your investment. It's a good ROI without the hassles and evasions of insurance company lawyers.

 

[westom]

Westom is correct in stating that a whole house protection system is the most effective for high-voltage surges from lightning strikes, but they do nothing to clean up power surges from internal sources such as motors or refrigeration.

How often do you replace dimmer switches, GFCIs, clock radios, the doorbell, and smoke detectors destroyed by surges from internal sources? Never? The refrigerator power cycles how many times a day? If those 'surges' were destrutive, then you should be replacing every digital clock multiple times daily. How often is that computer in a microwave destroyed by surges from the dishwasher?

Those surges are only noise. Made irrelevant by protection routinely found in all appliances - including cell phone chargers. A 'whole house' protector further diminishes those transients.

Anything a UPS does to 'clean' power is always done better inside electronic power supplies. That protection was superior in computers long before the IBM PC. Long ago, 120 volt electronics were required to withstand 600 volt transients without damage. Today's electronics are even more robust. Clearly, noise from appliances causes no damage.

Be concerned about a surge that actually does damage. That can overwhelm protection inside appliances. And that typically occurs about once every seven years. Informed consumers earth one 'whole house' protector. That more robust protector also makes other lesser surges (from inside or outside) irrelevant. Cost: about $1 per protected appliance.

 

[hafa]

How often do you replace dimmer switches, GFCIs, clock radios, the doorbell, and smoke detectors destroyed by surges from internal sources? Never? The refrigerator power cycles how many times a day? If those 'surges' were destrutive, then you should be replacing every digital clock multiple times daily. How often is that computer in a microwave destroyed by surges from the dishwasher?

Those surges are only noise. Made irrelevant by protection routinely found in all appliances - including cell phone chargers. A 'whole house' protector further diminishes those transients.

Anything a UPS does to 'clean' power is always done better inside electronic power supplies. That protection was superior in computers long before the IBM PC. Long ago, 120 volt electronics were required to withstand 600 volt transients without damage. Today's electronics are even more robust. Clearly, noise from appliances causes no damage.

Be concerned about a surge that actually does damage. That can overwhelm protection inside appliances. And that typically occurs about once every seven years. Informed consumers earth one 'whole house' protector. That more robust protector also makes other lesser surges (from inside or outside) irrelevant. Cost: about $1 per protected appliance.

Once again, we'll simply have to agree to disagree. As a professional power system designer and installer of 7+ years and IT professional of 16+ years with 400+ customers, I've seen more than ample evidence of the destruction wrought solely from dirty power and the increased longevity realized with the use of a good UPS. As far as I'm concerned, there's no substitute for practical field experience in this regard. BTW, here's my home power system, including suppressors. I still use quality UPS units (and yes, EVERY electronic appliance is connected to one) inside the house:

 

[westom]

I've seen more than ample evidence of the destruction wrought solely from dirty power and the increased longevity realized with the use of a good UPS.

Good. Then say specifically what parts are damaged inside electronics. And how that solution eliminated damage. Nobody cares what was observed. Observation without hard facts, numbers, and the underlying concepts is classic junk science. If you did what is required to know, then state what internal part is damaged and why. We did. For so long that 7 years experience implies a rookie.

A magic box labeled MagnaSine says nothing. If you know it does something, then posted were manufacture spec numbers or schematics. And statements that define which numbers are important. And why that number is relevent. Everyone has encountered scams by expensive magic boxes. Many even have a magic capacitor plugged in to reduce electric bills. No different than the MagnaSine recommendation. Because both are recommended by hearsay - without facts and numbers.

None of this helps the OP nor answers his questions. A UPS is installed at the service entrance that does 'cleaning' and surge protection. It costs many $1000s. That proves a $100 UPS from APC also does that? Of course not. The two devices are so electrically different that they should not even have the same name. Even the warranty for the OP considered UPS is bogus.

We often see magic boxes that claim pure sine waves. Read the numbers. Some of the dirtiest output power at 5% and 20% THD. Where is the protection? No problem. Because all electronics make that power even worse before converting high voltage (well over 300 volt) radio frequency spikes into rock sold, low voltage DC. IOW a superior AC power cleaner routinely and already is inside each supply.

So your power 'cleaner' simply provided power to something that only makes it many times 'dirtier'? Then 'cleans' that much 'dirtier' power? Why then 'clean' it with some expensive looking box? Because popular urban myths recommend it?

If you know something works better, then say what part was at risk, how often that part is damaged, what caused the damage, and why the box is a solution. Otherwise knowledge is missing.

Meanwhile, we have seen too many IT people who have no idea how electricity works; who only have, at best, an electricians grasp. An engineer would provide hard spec numbers (ie %THD); not some picture of a box. How does that picture help the OP; how does it answer his questions? %THD does. The picture does not.

Specifically what anomaly does that MagnaSine solve? A recomendation defines each anomaly. And says how each anomaly is solved. No magic box cures all 'dirty' anomalies. What specifically does a MagnaSine cure? And how does that help the OP?

 

[bud--]

Long ago, 120 volt electronics were required to withstand 600 volt transients without damage.

Required by who?

Today's electronics are even more robust.

Cite.

Informed consumers earth one 'whole house' protector.

Repeating from the NIST guide:
"Q - Will a surge protector installed at the service entrance be sufficient for the whole house?
A - There are two answers to than question: Yes for one-link appliances [electronic equipment], No for two-link appliances [equipment connected to power AND phone or cable or....]. Since most homes today have some kind of two-link appliances, the prudent answer to the question would be NO - but that does not mean that a surge protector installed at the service entrance is useless."

They may or may not provide protection for equipment connected to both power and cable/phone/dish/... wiring. One problem can be how cable/phone/dish/... wires are protected at the building entrance. And you can have direct pickup from a very near strike with wiring acting as an antenna. There are other loose ends.

UPSs are effective protecting against surges just like other plug-in protectors. It is easier to find high ratings in plug-in protectors. Any surge protector in the US should be listed under UL1449 (some UPSs don't seem to be). UPSs, of course, allow an orderly shutdown on power failure or voltage sag.

 

[hafa]

westom, you really love to blow things out of proportion. The image was simply to illustrate a full-home protection system (that happens to be part of an inverter back panel), nothing more.

I'll not waste further time arguing with you nor answering detailed rhetorical questions that are easily answered with a simple Google search. Your irrational hatred of UPS units is evident in your past posts and your obsession is self-evident in the single-minded nature of all of the 27 posts (the sum total of your posts) you've made on this one subject.

 

[Automata]

Alright then. I split the thread since we were way off topic, since we seem to be having a good and informative discussion here. I've also given it an appropriate title (let me know if anyone wants it changed) and I've moved it to the appropriate forum.

I haven't seen anything that breaks the rules yet, but let's keep this discussion civil while it is still there.

 

[Bobnova]

You're unlikely to find a PSU that can survive 600v, the primary capacitors are rated for between 400v and 450v and are not isolated from the incoming voltage. They (PSUs) generally, but not always, include a MOV (dubiously useful against transients, useless against very brief transients) or TVS diode (much more useful) for surge protection.
Against a fairly direct lightning hit both are useless, against more reasonable surges both will buy some time.
If the surge continues for long enough to charge the primary capacitor beyond it's rating it will almost certainly de-rate rapidly or fail completely. Electrolytic capacitors do not appreciate overvoltage.

As for motor surges, I've killed a number of CFL bulbs with my vacuum cleaner, they do not at all appreciate being powered off the same end of a circuit.


I'd like to see where electronics are required to withstand 600v. I can't think of anything I have opened up recently that can withstand 600v for very long at all. Capacitor charge time with wall power on tap is not exactly long.

A good APFC circuit inside a PSU can deal with up to ~380v for quite a while, above that and the voltage inside the primary capacitor starts going up.
This is true even if the computer is turned off. Any time the switch on the rear of the PSU is on, that primary capacitor is being charged to (DC rectified) line voltage.



As a (previously) outside observer, I feel obligated to point out that bud has put more hard, proven, facts and numbers on the table than westom has.

It's also worth noting that of everybody who has posted in this thread, Hafa has the best credentials. Admittedly he's the only one with credentials listed, but still.
For myself, I'm a PSU reviewer and MCU user/programmer. And of course, an end user of electronics bits. Professionally speaking, (beyond PSU reviewing) I'm a ~24 year auto mechanic.

 

[westom]

You're unlikely to find a PSU that can survive 600v, the primary capacitors are rated for between 400v and 450v and are not isolated from the incoming voltage.

Sentence after sentence is simply speculation. Only subjective. Many paragraphs with numbers follow to expose what was only a sentence of speculation.

Using the same reasoning, interface semiconductors are destroyed when voltage exceeds 40 volts. When not integrated into a system. What happens to the same semiconductor when integrated into a system? It can withstand up to 2000 or 15,000 volts without damage. That is IEC 61000-4-2. A design standard for electronics. Despite denials, design standards define existing electronic protection at thousands of volts.

An interface semiconductor from one manufacturer:
http://datasheets.maxim-ic.com/en/ds/MAX1487E-MAX491E.pdf
It says at the top: +-15,000 volts.

Absolute maximum voltage when not integrated into a system is any voltage from -0.7 to 12 volts without damage. 20 volts would destroy this semiconductor. But the same semiconductor can withstand a 15,000 volt transients. Because a design standard IEC 61000-4-2 requires it. Due to engineering reasons beyond the technical abilities of most posters.

Same applies to surges incoming to capacitors. Caps are typically rated at 200 volts continuous (not 400 volts as erroneously posted). And surges are not continuous. So another design standard IEC 61000-4-5 requires these parts to withstand a voltage transient of up to 4000 volts. As originally noted, all appliances have serious internal protection.

Apple once installed MOVs in their Apple II. Then stopped. MOVs inside an appliance did little protection for obvious reasons. MOVs are effective only when distant from electronics and as close to earth ground as possible. Apple (and so many others) stopped including MOVs, et al inside appliances long ago. Protection is created by other innovative concepts that require engineering knowledge to appreciate. This discussion has been limited to a layman's level. Informed layman routinely ignore posts that, for example, cannot provide manufacturer spec numbers for that protection. Because posts without numbers expose a poster with only hearsay knowledge.

Bottom line - all appliances even with 200 volts capacitors will also withstand thousands of volt surges.

Even transient immunity for some signal inputs on the original IBM PC met EN55024 - 500 volts. Another engineering standard that defined high voltage transients without damage. If signal inputs that long ago withstood up to 500 volts, then why does its AC power input not meet a trivial 600 volt requirement?

Well the accusation is this: "I have never seen that number so it must not exist." Electronics have long been designed to withstand up to 600 volts without damage even in the old CEBMA standard. Today's more robust electronics now meet standards at thousands of volts.

This is only the layman's version. Actual engineering behind these numbers would be many times longer. All that only answers one "speculative" sentence. Reality posted previously and repeated again. Today's appliances may even contain protection rated at thousands of volts. Because best protection at the appliance is already inside the appliance.

I have seen protection numbers for some clocks at 2000 and 6000 volts. But then I actually did this stuff for a few generations as an engineer.


PSUs makes most transients irrelevant. Apparently Bobnova did not grasp what was written. A rare surge (ie a direct lightning strike) can overwhelm that existing superior protection. Nobody said a power supply can withstand a direct strike. Why the strawman accusation? So that appliance protection is not overwhelmed, then informed homeowners earth one 'whole house' protector. This paragraph refutes another bogus accusation. Nobody said a power supply will withstand a direct lightning strike. That rare surge (maybe once every seven years) must be earthed before entering the building. So that superior protection already inside every appliance is not overwhelmed.

Where are numbers for APC protection? Normal and continuous voltage, even for laptops, is a voltage peaking at 375. What does the UPS do? Nothing. A typical UPS would be destroyed by 380 volts continuous. When were your AC mains outputting a continuous voltage that peaked at 375? If 120 volt mains rose ony to 130 volts, then light bulb are suddenly 30% brighter. How often do your lights increase intensity that much? Never. Another straw man. The UPS must protect itself from 380 volts because AC voltage never exceeds 130 volts continuous? That claim disputed even by incandescent bulb intensity.

Useful posts provide hard numbers. Bobnova complains that he cannot find numbers. That proves 600 volt internal protection does not exist? Posted are numbers that define robust protection in appliances. Same applies to Hafa who also never posted (apparently did not learn) the numbers. He presented a MagnaSine in a discussion of hardware protection and power conditioning. He now admits that product is completley irrelevant to the discussion. And still posted no numbers that would be relevant such as %THD.

My personal troll promotes power strip protectors. It is his job. But he cannot find manufacturer spec numbers for protection. He has followed me into every discussion for most of the past decade. Has been challenged repeatedly to post manufacturer spec numbers that claim protection. And never once did. Of course, he could be honest and post those numbers. And then we see why his products only claim to protect from transients made irrelevant by the so many numbers posted above.

He even misrepresents the NIST. The NIST says his products are the easiest solution. Then shows why that many times more expensive solution is also "useless". NIST's word - not mine.

Same problem applies to a UPS. Its protection numbers are near zero. A UPS typically connects AC mains directly to the appliance. How does that relay do protection? It doesn't. Just like the near zero protection numbers in its specs. UPS even has a warranty that will not be honored. Because its function: to provide temporary and dirty power during a blackout. Because electronics are so robust as to even make its 'dirty' power irrelevant.

Anyone can make subjective accusations in sentence. Reality with numbers requires many paragraphs. And this was only a summary at the layman's level. Who else has posted this much knowledge from decades of experience? Layman separate the naive from the informed by noting one hard fact. The informed post numbers - not subjective myths. Or a picture of a box that somehow proves knowledge.

 

[Bobnova]

Same applies to surges incoming to capacitors. Caps are typically rated at 200 volts continuous (not 400 volts as erroneously posted). And surges are not continuous. So another design standard IEC 61000-4-5 requires these parts to withstand a voltage transient of up to 4000 volts. As originally noted, all appliances have serious internal protection.

I'd go through the entire post and refute it, but the line about capacitors and 200v is simply too far wrong for me to take any of the rest seriously.
Every single capacitor has it's own, individual, rating.
APFC caps are, like I said, generally rated for between 400 and 450 volts. It says right on the capacitor what it's rating is.
The voltage doubling primary capacitor setup in old PSUs uses capacitors that are typically 200v, but APFC PSU do not and cannot use a cap rated that low.
As a bonus, here is the IEC 61000-4-5 guideline: http://www.sanki-e.com/uploadimg/contents/20100722110236817.pdf
Note that class two, which PSUs may fall into, requires only very minimal testing.
Note that class three, which PSUs may also fall into, is limited to 2kV. Far from the 15kV you've stated.

Furthermore, every IC has its own individual maximum rating as well. Do not confuse the ability to withstand a 15kV static charge with being able to survive a 15kV surge.
http://www.st.com/internet/com/TECH...AL_LITERATURE/APPLICATION_NOTE/DM00023467.pdf
That's the IEC 61000-4-2 testing procedure for STM bits. Note that it is all about static.
ICs have, generally speaking, anti-static protection didoes built into them. In many cases they are as simple as a zener diode for each pin to GND. This is why you see so many 5v devices with a 5.5v maximum rating, the zeners (at the low end of their expected range) can trigger at 5.5v and destroy themselves (and the chip) rapidly due to overheating.
Also on the subject of the IEC 61000-4-2 static regulations, note that in STMs example a 14kV static gun is only actually giving a 152 V spike. Static has very limited amperage, so it is very easy for even tiny capacitors to drop the voltage significantly.

In short, you're confusing static protection with surge protection as well as completely wrong about capacitor ratings, and very wrong about IC ratings.

The bottom line is, I can tell you will never be convinced that you are wrong, despite the fact that you are very wrong indeed and are spreading misinformation.
Rather than try to convince you further and frustrating myself in this thread pointing out false statement after false statement, I will simply ignore it.

To those who read this thread in the future: Do your own research, do NOT simply believe what any one individual says.

Especially when they make blanket statements (any), and doubly so when the blanket statements are completely incorrect (such as the capacitor bit).


EDIT: Oh man, this is too much. After this I'm done. I found Westom's credentials while googling for his (myriad) other forum threads hating on surge protectors:

My friend knows someone who knows this stuff.

From about two pages into this thread: http://ths.gardenweb.com/forums/load/appl/msg121610005596.html

 

[westom]

I'd go through the entire post and refute it,

Your numbers confirm what was posted. Electronics were required to withstand 600 volts without damage. Today, that number for appliances is thousands. As much as 4000. Because protection inside appliances is robust.

And again, not robust to withstand a lightning strike as you misread and misrepresented. So robust as to make all but the rare and destructive surge irrelevant. An event that might occur once every seven years - another number. Why do I have so many numbers? Oh. I did this stuff.

Why does a part destroyed by 20 volts suddenly withstand 15,000 volts? Again you intentionally misrepresented the point. Then need not admit to concept never learned. You did not even know of the 600 volt number. Or the many standards that require 15,000 volts or 4000 volts without damage. Even 200 volt capacitors can withstand a thousand volts when part of a system. All appliances already have robust protection.

Well, here is a Tivo power supply. Its capacitor is 330 uf 200 volts (not 400 volts). It is sometimes powered by a UPS that, in battery backup mode, outputs 200 volt square waves with a spike of up to 270 volts. (BTW, the UPS manufacturer calls that a sine wave output.) So why is that Tivo supply not destroyed? Industry standards. That UPS outputting square waves and a 270 volt spike is perfectly good power for electronics. But again, numbers provided that say so.

Design numbers from another standard:
For 120 volts, it must withstand 870 volt spikes. For 220 volts, withstands up to 1594 volts. And for 240 volts, withstands spikes of up to 1700 volts. Why do I post numbers - and you do not?

At least one Seasonic power supply spec claimed internal protection up to 1800 volts. That's not unusual today. But not well known especially when a UPS manufacturer is somehow a trusted information source (without numbers).

Appliances routinely contain protection. Meaning most surges are only noise. Surges that might destroy an adjacent protector may also be too small to harm the appliance. Many consumers saw that. Then credited the grossly undersized protector rather than superior protection inside the appliance.

If refrigerators and vacuum cleaners were creating destructive surges, then we all troop daily to a hardware store to replace kitchen and bathroom GFCIs, wall clocks, microwave ovens, and smoke detectors. Or did you just ignore that fact? If that appliance is creating destructive surges, then it is first destroying itself. And finally, if that motorized appliance is creating surges, then a responsible consumer trashed it as defective. Or fixed the rumored defect.

Rather than spend hours searching to attack the messenger, instead, why not learn how electricity works or fix a rumored defective vacuum? Or learn some useful numbers to contribute something positive or informative?

But then excuses for spending $hundreds on a UPS would no longer exist! Sales would be harmed.

The OP asked about a UPS for its warranty? Why? Informed consumers make destructive transients irrelevant at the service entrance. Then superior protection already inside appliances and defined with numbers is not overwhelmed. Informed consumers worry about what can actually cause damage. Not myths that promotes a $100 or $200 UPS.

 

[hokiealumnus]

westom, I can see you're very good at typing walls of text but you're providing about as little substance as I could imagine for so many words. You're saying lots of stuff, but none of it really amounts to anything other than your evident (<-link) vendetta (<-link)against (<-link) UPS'es (<-link) that seems to (<-link) go on (<-link) and on (<-link)....and that's merely seven links out of I'm sure many more. I'd refer you back to the last link at JohnnyGURU's forums, where cyberphunks seems to actually know what he's talking about and did a great job of tearing your professed arguments apart.

Please cease posting things in this thread like "Why do I post numbers - and you do not?", when clearly you have nothing to back up what you say. Or just cease posting in the thread at all without sourcing your walls of text, that'd be ok too.

If you DO have something to back it up, by all means supply links explaining precisely your point, becuase for all that text I've come up with exactly one point - that you hate Uninterruptable Power Supplies and will stop at nothing until they are eradicated from the face of this earth.

To sum up, take your vendetta elsewhere please. We do not have any interest at all in furthering your cause, especially when it is backed up with "My friend knows someone who knows this stuff."

 

[Automata]

Wow, I split the thread, walk away for a few minutes, and it explodes.

Guess I won't split threads anymore.

 

[westom]

If you DO have something to back it up, by all means supply links explaining precisely your point, because for all that text I've come up with exactly one point - that you hate Uninterruptible Power Supplies and will stop at nothing until they are eradicated from the face of this earth.

My god. I even said 'why' each point is valid. Sometimes even quoting professional sources. How many more facts, numbers, and sources should I cite?

If those facts and numbers are wrong, then show me. Identify the error. Or ask for details by defining what has you confused or unsure. Please don't vent emotions. I posted only facts with numbers. And did this stuff even decades ago. How many times have you earthed direct lightning strikes without damage to anything? How often have you traced damage by even putting a semiconductor under a microscope?

How can I hate a UPS when I even have a few? Its output was even defined with numbers. I do not use it to do what even the UPS manufacturer says it cannot do. I use to do what it is designed to do - temporary power during a blackout. If I am wrong, then simply post a UPS spec number that says where I am wrong. The challenge was repeated often. Show me. If that UPS does more than just battery backup, then please, quote the spec number. Show me the manufacturer statement that says otherwise.

Where is hate in any of this? I provided technical facts. A UPS only does what its specs say it will do. Facts and numbers do not hate. Please don't just deny because you are angry or confused. Show me what it is wrong or what has you confused. If you need further supporting facts and numbers, then identify the confusion. If you do not understand a concept, then ask. If you know the post is wrong, then show me the error. Manufacturer spec numbers are especially useful.

Why would I 'hate' a UPS and use a few? Where do I even imply 'hate'? And why are you so angry? If you don’t like it, don’t read it. If you know it is wrong, then cite the specific electrical concept or number in error. But please do not tell me it is wrong because you just know.