Ups Only Protecting Mains-Failures?

[mistitikimikis]

Two weeks ago all of a sudden a noise like somebody threw a handful of gravel against the window and the computer went down.

I am using a quit expensive PCM-UPS (Powercom) which served me for a couple of years perfect with one repair under guarantee.

Opening the UPS to try to stop the still continuously running ventilator which I could do by removing a build-in fuse, learned me that one of the big main electrolytic capacitors of the power-supply / rectifier was damaged in other words it's top was round-formed.

After (paid) repair by PCM everything seemed to be fine again until to-day a fuse (10 Amp.) of an extension-lead at the outgoing (computer)-side of the UPS blew with a loud noise. Again a heavy failure in the UPS and I praised myself lucky that I used by coincident a fused extension-lead. I didn't open the UPS now, in spite of the 24-hours at the batteries running ventilator being afraid to lose the guarantee at the repair.

The last failure made me afraid as one trusts a UPS against power-failures from the mains BUT what should has happened now when there was no fuse outside of the UPS??? The external UPS-fuse didn't trip and measuring the contacts of the UPS learned me that both poles (so computer-side) showed electricity.

In the case of an internal UPS-failure there is a possibility that the computer including other hardware can be damaged or am I thinking too negative?

[kamalabob2]

I've posted some "electric problems" our home experienced last year on Thai Visa. Any UPS other than APC brand sold in Thailand is generally not up to the standard you receive from an APC brand UPS. However the UPS is NOT a sure way to prevent damage from electrical surges. I now have my expensive electrical A/V items and the desktop computers plugged into power via either a LEONICS brand TPS series surge suppressor or a "SurgeGuard" brand TWSS series surge suppressor.

Either brand of the "surge suppressor" is readily available in any IT Mall in Thailand. I still use a UPS for computers and monitors, but the UPS is plugged into the "Surge Suppressor", which is plugged into the wall socket. Leonics and Surgeguard are not "Monster or even Belkin" priced per joule capacity.

[mistitikimikis]

For the time being I'll use this one

http://www.antoni-uni-photography.com/THAILAND/TECHNICAL-STUFF/Various/16175222_p7gJ73#!i=1771580132&k=4gSLV3p&lb=1&s=A

I brought once from Europe. Anyway stupid to think concerning the UPS only the of the mains-side and not of the outgoing side, expecting they have enough protection and never a failure.............

Thanks for your advice!

[rakman]

Another "trick" you can do for those serious spikes like lightning, is to put a couple of knots in the power cords at the "output" end. This, believe it or not, make an inductive shunt to help attenuate those hearty spikes. Of course a massive spike will overcome everything you throw at it and only the outcome will determine what's left working.

Good surge supressors have various size, ie freqency related, suppressors that cut down the spike.

Also, after a big spike, check your surge suppessors or you may find the next one is the fatal one for your gear.

[mistitikimikis]

Thanks! I'll follow your suggestion, which is explainable!

http://www.dansdata.com/sbs9.htm

Anyway, again, never ever thought about the real possibility that the "protective" UPS can cause proper damage when crashing!

[westom]

Thanks! I'll follow your suggestion, which is explainable!

That dansdata and other suggestions are myth after myth. Let's start with an easy one. A knot in a wire will stop transients. OK. Then the knot on a TV's coax cable will also stop TV signals. Try it. Put a few knots in the TV cable. Does a knot stop microamp currents? Of course not. Then how was it going to stop tens or hundreds of amps destructive currents?

That knot is an example of how easy urban myths get repeated like 'gods living on tall mountains'. After all, everyone also knows that is where gods live.

Another example. Grossly undersize a protector. A transient current too tiny to overwhelm protection already inside all appliances will, instead, destroy the protector. Then wild speculation says, "My protector sacrificed itself to save ...". Nonsense. Superior protection already inside appliances did the protection. A grossly undersized protector (grossly undersized MOVs) only created a potential house fire.

Or you could read the numbers. How does that adjacent protector, rated at hundreds of joules, absorb destructive anomalies that are hundreds of thousands of joules? Why so many recommendations without numbers? Then you might discover how easily they too were manipulated and lied to.

To increase profits. And to get the most naive to recommend that protector. Grossly undersize it. When it fails, a majority will credit the grossly undersized protector. And recommend it using subjective reasoning - no numbers.

Where is protection in a UPS? If the recommendation comes without facts and numbers, then the recommendation is probably a lie. If a UPS does that protection, then specifications numbers are also provided that define that protection.

I spit on a fire and put it out. Does that prove I can also extinguish a house fire out by spitting? Dansdata makes such claims ... subjectively. Perspective. Honesty requires numbers. Best protection for electronics costs tens or hundreds of times less money. Is found in every facility that can never have damage from lightning. Cannot be located adjacent to electronics. And remains functional after every direct lightning strike.

And yes, I said it costs that much less money. Because the well proven solution, from more responsible companies, puts your money into protection. Not into promoting urban myths. If the knot did protection, then the same knot stops TV signals. Try it yourself. Asking damning questions and always demanding the underlying science exposes lies and advertising myths. None of previous recommendations were honest or accurate.

[mistitikimikis]

Another "trick" you can do for those serious spikes like lightning, is to put a couple of knots in the power cords at the "output" end. This, believe it or not, make an inductive shunt to help attenuate those hearty spikes. Of course a massive spike will overcome everything you throw at it and only the outcome will determine what's left working.

Good surge supressors have various size, ie freqency related, suppressors that cut down the spike.

Also, after a big spike, check your surge suppessors or you may find the next one is the fatal one for your gear.

I got the UPS repaired back and installed with a protector at the entrance and one at the outlet on top of that the outlet protected with a fuse.

Amazingly I found in the mean time in a box by coincidence a short LAN-connection cable with your suggestion already from the manufacturer!

Although I know that the protection I use is weak it is at least something because, again, I don't trust now the out-going side of a UPS not anymore!

The protection I use was the last of a 3-step protection we were used to install in Europe in a domestic installation: first a very robust installation of Phoenix Contact on the street-side, than a second step in the general domestic breaker-box and the last one is the one I'm using now.

All prevention is better than cure afterwards..............

[rakman]

...

All prevention is better than cure afterwards..............

+1

[westom]

Although I know that the protection I use is weak it is at least something because, again, I don't trust now the out-going side of a UPS not anymore!

You might first learn protection as defined by numbers. For example, all ethernet interfaces already have protection rated at least 2000 volts. Yes, that is thousands. Any protection adjacent to that port is usually inferior. Promoted mostly by advertising. Not by science or numbers.

So that superior protection already inside every ethernet port is not overwhelmed, you must earth destructive transients BEFORE they enter the building. Then superior protection already inside every appliance is not overwhelmed.

View numbers for that UPS. It only claims to protect from transient that are typically not destructive. That are made irrelevant by protection that already exists. Ie, 2000 volts for ethernet ports.

Protection is never about some magic box. Protection is always about whether the typically destructive transient is earthed BEFORE it can enter a building. Then 2000+ volts inside every ethernet port is not overwhelmed.

BTW, other electronics interfaces are even designed to withstand 15,000 volts. Those who recommend bogus solutions will not post numbers. Numbers from a datasheet of a semiconductor interface chip. View the first page. Protection is rates at 15,000 volts ... because the best protection at an appliance is already inside the appliance:

http://datasheets.maxim-ic.com/en/ds/MAX1487E-MAX491E.pdf

Find the expression that says "+-15kV ESD-Protected"

But then only some posters have engineering training and experience. Are not educated by advertising. Only know because they did not demand the numbers.