Mains Conditioner with Surge Protection

[Hankieroseman]

I think that is up for debate. Many, including me, contend constant banging by surges and spikes does result in cumulative effects on some of the devices, increasing their rate of aging. I had it explained to me once as being similar to loud music and jet engine noise on our ears. Eventually, the effects will catch up to us and hearing loss will result.

The facts are, surges and spikes are regular occurrences, even with modern, stable grids. But the vast majority are relatively small so I agree, with a quality design, quality parts, and relatively small anomalies, the surge and spike protector will work fine and last a long time - a lot longer than 2 years. But NOT indefinitely nor capable of enduring infinite hits, even if minor.

But here's the problem. No one (unless a tech/scientist in a proper lab environment) knows how many surges and spikes are hitting our electronics, day in and day out, or how severe those surges and spike may be. So there is no way for us "normal" people to realize the effects they are having on surge and spike protectors. So IMO, that 2 year recommendation is being conservatively safe. Too conservative? Maybe. I would say if you go with a quality Tripp-Lite or Panamax or similar protector, it will last many years under normal use.

Of course, if Mother Nature get mad and targets you directly, nothing will protect you, except a quality backup plan.

Will testify for loud music and working at the airport for 20+ years to hearing loss and the two P&W JT8D's in my ears.

 

[qxp]

In reference to that APC video, note he said that was one of many units he bought, and the only one that failed. I have tried to find other cases of problems with APC and aluminum transformers and wiring. I could not. It appears nearly all of the APC issues are based on the same, nearly 2 year old report that has gone viral. So I think it important not to blow the issue unfairly out of proportion.

Once he took it apart, he could not find any burnt or arc marks on any of the circuits and could not smell any significant amounts. Only after removing the transformer did he see some "minor" evidence of heat and a "slight" odd smell. He did notice with his meter the transformer had a "open". If a circuit is to fail, you want it to open rather than short - so that is good.

I saw this video first a year ago and the reason I found it is because I was looking at why several CyberPower units would randomly cutoff while computers without UPS would keep going. I did not take apart the CyberPower units to find the cause, just decided to go with different UPS.

I am just saying, one failure out of his "many" units does not indicate a major problem or trend.

Also, it should be noted there is nothing wrong with using aluminum "IF" done properly. Clearly, the maker of that transformer didn't do it properly. I note neither APC nor CyberPower manufacture transformers. APC "may" have been ripped off unknowingly as well.

Exactly. But what happened is that the industry switched to using cheaper material and only now starts seeing the long term effects. It would take a while for aluminum transformers to approach robustness of copper ones.

If I was doing quality assurance, I don't see me tearing apart a transformer to see how it is made. I would test it to see if it met electrical requirements, but that is probably it. I would hope today, APC has gone with a different source for their transformers, and a more robust QA inspection. Fingers crossed.

So what they clearly did not test is resistance to vibration.

Yeah, that's the issue. When dissimilar metals like aluminum and copper come into contact with each other in the presence of an electrolyte, galvanic corrosion can occur. In the case of bonded joints of these two metals, if not done properly, that corrosion can (eventually will) weaken the joint. That appears to be what happened in that transformer. I have to wonder what the electrolyte was that became the catalysts there. Air is not an electrolyte. There must have been some contaminates that got in there somehow during assembly. Another reason to go with a different source for the transformers.

In this case corrosion is not the issue. It is that copper is soft, aluminum is soft, but copper-aluminum plus whatever compounds are brittle. A little vibration makes a small crack, aluminum oxidizes and becomes more brittle, a little more vibration enlarges crack and so on.

Because of the copper shortage back in the 50s and 60s, as noted above, many homes were built using aluminum wiring. Mine is one of them. Fortunately for me, it was done right. They used proper gauge wiring, and all the outlets and breakers were designed for use with aluminum. Over the years, there have been several remodeling projects that used copper wires. Fortunately again, where new copper was bonded to old aluminum, proper bonding was done so no worries here about galvanic corrosion.

Exactly. But it took years to learn the lessons and the transformer industry needs to repeat it. I actually would have expected big industrial transformers to use aluminum, so the expertise must exist. But probably it has not spread to low cost suppliers.

I have wondered that too. Most UPSs have two sides to their output - side one is supported by battery and the other by surge and spike protection only. I suspect you are correct with cheap budget, basic battery backup UPS. But for the better UPS, both sides are protected by advanced, "active" intelligence of the AVR feature and therefor, not susceptible to such aging issues as a basic "passive" surge and spike protector is. I note if you look at essentially all my posts on this subject, I repeatedly say "good" UPS with AVR. Just like power supplies, there are good UPS, and cheap, budget UPS that need to be avoided at all costs.

Actually quite a few UPSs have lower surge rating in Joules than a dedicated surge protector. If you open one up you see a choke coil, some capacitors and MOVs and that's it.

 

[Shrek]

Seems there exists flux to solder aluminum to copper

 

[R-T-B]

Here is a takeapart video that shows APC UPS (its not just CyberPower) having a failed connection at the transformer:

As you can see in the video they did solder a copper wire to aluminum wire, only for the contact to go bad. The way they packaged it you would expect an initially good connection to go bad because joining such dissimilar metals creates a lot of brittle intermetallic compounds that are very susceptible to vibration.

This is pretty much what happened when people first tried to use aluminum wire in housing - initially good joints would go bad over time causing fires. Took a while to develop regulations that specified how exactly to make aluminum wire connections that would stay working.

Unfortunately, they don't say whether a given UPS has aluminum or copper transformer, so if you want to be safe get an old UPS or an industrial one (like meant for servers). Anything new in the $100-200 is probably using aluminum.

Damn I have at least one of those UPSes deployed right now, same revision and everything, even same receipt on the side.

I think I'll redeploy it to the most low vibration part of my workspace and replace it next time the battery set fails...

 

[Shrek]

This should go into a thread of its own before the moderators rightly tell us we are off topic.

 

[Bill_Bright]

I believe that a varistor is similar to many diodes and so might be able to take an indefinite number of small surges

"Indefinite" by definition means "undefined" and can mean anything, to include just 1 or 2, then toast! Perhaps you meant infinite, as in no limits.

Actually, it should not be difficult to make a device to monitor the surges.

I didn't say it would be difficult. I said "normal" people just don't monitor the power coming out of our outlets for such anomalies. That said, your example would not do it. Such a monitor to determine impact on surge and spike protectors would need to not only count the number of surges and spikes, but their potentials and durations. But even then, the numbers would need to somehow be qualified as to what is potentially damaging and what is not.

So what they clearly did not test is resistance to vibration.

Why would they? These UPS would not be expected to be subjected to vibrations. These UPS are designed for and marketed to be used in typical home and office environments. That is, just sit on the floor.

To me, it is not so much about testing (other than meeting electrical specs and requirements), but rather inspecting to see if the products adhered to stated physical requirements and safety and other standards in production. Using crimps to connect wires would not, to me, meet any standard that I am aware of. But again, if I were a QA inspector, I am not sure I would tear a part sample transformers to see their innards and how they were being made, and if any corners were being cut in production.

In this case corrosion is not the issue.

Sure it is. Or at least one of the issues. The problem with dissimilar metals coming in contact with each other is all about corrosion. Therefore, when bonding dissimilar metals, it is essential the bonds be of the type that ensures corrosion is not even possible - ether by sealing out the environment (with proper welds or solder designed for that) or by creating a bond that, in effect, changes the molecular structure at the point of bond. A mechanical "crimp" does neither. They might as well twisted the wires together.

This should go into a thread of its own before the moderators rightly tell us we are off topic.

I think it has played out anyway.

 

[qxp]

This should go into a thread of its own before the moderators rightly tell us we are off topic.

Good point ! Is there a way to split a thread ?

"Indefinite" by definition means "undefined" and can mean anything, to include just 1 or 2, then toast! Perhaps you meant infinite, as in no limits.

Sorry typo - I mean "infinite", which is what it says on the chart.

I didn't say it would be difficult. I said "normal" people just don't monitor the power coming out of our outlets for such anomalies. That said, your example would not do it. Such a monitor to determine impact on surge and spike protectors would need to not only count the number of surges and spikes, but their potentials and durations. But even then, the numbers would need to somehow be qualified as to what is potentially damaging and what is not.

Well, I thought just sampling voltage over time and plotting the excursions and/or their energy would be helpful. I have plots of line voltage from my UPS but they are too slow to notice surges. I would expect we would find that the power transformer and house wiring has a fair amount of inductance and bad surges aren't frequent (at least here in USA).

Why would they? These UPS would not be expected to be subjected to vibrations. These UPS are designed for and marketed to be used in typical home and office environments. That is, just sit on the floor.

You always have a little vibration, in particular from the fan. Over time this weakens the solder joint. This is true even for copper but especially aluminum. The usual solution is to make sure the joint is secured and is not subject to mechanical stresses. That's why on older transformers there is a little plate with contacts. If you have all-copper windings you can just take the copper wire out and solder it to PCB the transformer is attached to.

 

[Wirko]

Sure it is. Or at least one of the issues. The problem with dissimilar metals coming in contact with each other is all about corrosion. Therefore, when bonding dissimilar metals, it is essential the bonds be of the type that ensures corrosion is not even possible - ether by sealing out the environment (with proper welds or solder designed for that) or by creating a bond that, in effect, changes the molecular structure at the point of bond. A mechanical "crimp" does neither. They might as well twisted the wires together.

That's a particular problem with Aluminium-Copper bonds (and copper includes brass here). Not just any combination of dissimilar metals. Practical solutions to this problem include (as I mentioned before) inserting a stainless steel terminal block between both metals, or using partially copper-plated crimp terminals made of aluminium, such as these:

By extension, I assume that copper-clad aluminium (CCA) wire can also be safely soldered to copper. CCA is a common material that China uses to reduce cost. Aluminium transformer windings might be made of that, and it would look like copper until cut or scratched.

 

[qxp]

By extension, I assume that copper-clad aluminium (CCA) wire can also be safely soldered to copper. CCA is a common material that China uses to reduce cost. Aluminium transformer windings might be made of that, and it would look like copper until cut or scratched.

From what I see online copper thickness in thin wire can get very small - down to 25 microns:

Copper Clad Aluminum Wire (CCA) – 5 Key Benefits & Specs | Greenshine

Copper Clad Aluminum Wire (CCA) delivers copper-grade conductivity at 63% less weight. Ideal for grounding, lugs and cable-tray jumpers. Available in diameters 0.1–8.0 mm on custom reels. Request your CCA wire quote today!

www.cca-wire.com

That's problematic because if you solder it, you will most likely replace pure copper cladding with intermetallic compound of solder and copper, which is brittle and might just peel off. I wonder if anyone has run tests. Such wire would probably be ok to do crimp contacts.

 

[Bill_Bright]

Well, I thought just sampling voltage over time and plotting the excursions and/or their energy would be helpful.

It is but not to determine the impact of the effects of accumulative pounding on surge and spike suppression devices. My UPS software provides much of that information. And some kill-a-watt meters show the most recent peak values - though not sure they are quick enough for spikes which may be considering quicker than surges.

You always have a little vibration, in particular from the fan. Over time this weakens the solder joint. This is true even for copper but especially aluminum.

Nah! Sorry. Nice try but not buying that one.

Any proper soldering joint is more robust than that. Maybe if you lived next to busy railroad tracks in a flimsy house, then vibrations would have an effect. But then your TVs, computers, stoves, coffee pots, microwave ovens, routers, modems and countless other electronic devices would be falling apart too. Not happening.

Besides, not all UPS have fans. And many that do only run when really taxed. And I note when not on battery, UPS are not taxed at all.

 

[Wirko]

Not sure how this would help, as the surge would still pass through the transformer.

But then you could safely cut the ground wire and use the PC ungrounded. Also, what @qxp said. But forget it, it would still be unreliable (I mean for this particular purpose: keeping the voltage induced in an induction loop from doing damage). Not to mention the cost.

One really wants to avoid forming a ground loop

In my case, a closed ground loop could have prevented the induced voltage from concentrating in a small number of weak points, one of which was the insulation between my motherboard's Ethernet socket pins and PC ground.

But you probably mean audio interference here. Here's the situation you want to avoid: a loop is formed, one part of it consists of ground wire(s) carrying dirty currents from power supplies (and EMI filters!), and another part of it is supposed to be clean because incidentally it's the ground wire in an unbalanced audio cable. A small portion of dirty current will take the path that should remain clean. That's the reliable RRRR-recipe for pr-RRRR-roblem-ZZZZZZ.

 

[Shrek]

I meant a ground loop will cause current to run if near a lightning strike.

Ground Loops: What They Are and How To Avoid Them

 

[Wirko]

I meant a ground loop will cause current to run if near a lightning strike.

Ground Loops: What They Are and How To Avoid Them

Yes. But voltage is induced in a loop no matter if it's closed or open (or broken at multiple points). In a closed loop, it's not actually voltage but electromotive force or EMF (same unit: volts) and it can't be measured with a voltmeter. It generates current according to Ohm's law, which can do damage such as overheating of wires. In an open loop, there's measurable voltage that can cause damage in other ways - for example, it can irreversibly break down insulation.

 

[Bill_Bright]

and it can't be measured with a voltmeter.

Well, I think the parameters of the scenario need to be defined/refined in order for that to be considered true as an absolute. While it you may not be able to measure it with a multimeter, it often can be calculated mathematically or even measured in the secondary - again, depending on the scenario.

But we are talking computers here - and I note there are literally dozens of ground points on motherboard through the standoffs, via every data and power connector, every RAM slot, CPU socket, expansion cord slot and many dozens in the rear panel I/O area. And much of the reason for having so many is simply to mitigate potential loop problems before they can even develop.

No doubt the main PCB inside any UPS has multiple grounding points as well. Even a transformer is typically grounded at all 4 corners and at least one point on the primary and one on each secondary winding.

Of course, all bets are off if the device has been damaged, or is a fake knock-off.

 

[R-T-B]

This should go into a thread of its own before the moderators rightly tell us we are off topic.

Not much more to say so no new thread needed IMO. Back to the subject.

 

[Shrek]

Yes. But voltage is induced in a loop no matter if it's closed or open (or broken at multiple points). In a closed loop, it's not actually voltage but electromotive force or EMF (same unit: volts) and it can't be measured with a voltmeter. It generates current according to Ohm's law, which can do damage such as overheating of wires. In an open loop, there's measurable voltage that can cause damage in other ways - for example, it can irreversibly break down insulation.

That is why I emphasized current.

Not much more to say so no new thread needed IMO. Back to the subject.

Good point, I stop here, even though the aluminum windings issue seems fascinating to me.

 

[Hankieroseman]

Well I learned something. I'm sticking with UPS. Thanks for furthering my education gentlemen, I picked up a thing or two along the way. Cheers to each and every one of you. .

 

[95Viper]

Stick to the topic.
Stop the trash posting.
Follow the guidelines for posting/etc.

 

[Wirko]

Good point, I stop here, even though the aluminum windings issue seems fascinating to me.

Time to start a new thread then - but we know in advance that it will develop in several directions at once. Maybe the mods would allow a General Electric eternal thread in the lounge.

 

[qxp]

Started thread here - let me know if its ok.

UPS and surge protector info, tips and tidbits.

This is meant to be a dedicated thread for discussion of UPS and surge protectors. Models which worked fine for you over the years ? Models that should be avoided ? UPS and surge protector electronics UPS battery chemistry (do you prefer lead sulfur or lithium or something else ?) UPS type...

www.techpowerup.com