Surge Protectors, scams and saints

[saturation]

Hello All:

I found the blogs by accident and was pleasantly surprised by a knowledgeable, enthusiastic, and honest appraisal of electronics and devices.

The review on audiophile equipment and cheap multimeters is exemplary.

One thing that is vital to review are plug in consumer surge suppressors.  Many claims are made and almost none can be tested by consumers, and even engineers can't without the right high voltage equipment.

Like fuses, you only find out how good a surge suppressor is when it burns out to protect devices plugged into it.  But the owner can't determine if a true serious overvoltage event occurred and was was thus saved by the surge suppressor, or simply the surge supressor blew on its own, from a defect or age.  Without an event recorder, you'll never know.

Thus, the only way to know a surge suppressor works is if the fundamental design works, and tested to failure.  This does not exist as I know, and neither Consumer Reports in the USA, or any other agency like UL, provide this type of testing.  UL tests for safety, not function, and CR does not do in depth engineering tear downs.

It would be good to see EEVblog do a tear down of a surge suppressor in detail.  You can make a whole series of this topic, as well as audiophile gear and multimeters, by going through each model and doing a torture test of each device, to prove their claims as EEVblog does with Fluke DMMs.

Just standardize you test suite, and go for it!

[tomlut]

I used to work as an R&D engineer with a surge suppression manufacturer and would be really interested in this.

I've seen some shockers (no pun intended) over the years.

My basic observation would be, like in all things -you get what you pay for. Buy a $20 plastic multi-outlet power board and expect your house to burn down. Buy a + $100 metal housed version and you can be reasonably sure your investments will be substantially protected.

Having said that there are some awful tricks and misinformation in this industry. Take "transient response technology" for example. It was invented so that a particular manufacturer's products would pass UL approval, at the detriment to actual equipment protection. The device would only clamp fast pulses, it would not respond to slow over voltages (e.g. +150% supply voltage @ 50Hz). So your equipment would not be protected if two of the three phases shorted out or the neutral lifted down the street, but the device has a nice confidence inducing UL logo on it.

Another easily identified trick is to list "total protection" or the sum of the surge rating for each phase, when in reality each phase is only protected by 1/3 of the rating.

I know where you could get access to a 20kA 8/20µs surge generator if required. It's not portable, you'd have to travel to it

[alm]

Just saw this topic for the first time, thanks for the informative review! It could be an interesting topic, I'm sure Dave could set up some test, possibly with the help of Doug. But it would be hard to simulate the extremely low output impedance of the mains supply.

It's hard to verify if a surge protector works without having all equipment twice, one plugged in in a surge protector and one in a normal power strip, or at least continuously monitoring the power quality. As long as no equipment dies, it might be just like the rock that keeps tigers away.

About the slow over voltages, is there anything a surge protector (without complex electronics like a UPS) can do against that? There's only so much energy that the MOV can convert to heat.

You talk about the fire risk, wouldn't (non-UL listed) equipment pose a similar risk if faced with a transient, for example from the EMI filter?

[saturation]

Hello tomlut,

This was my first eevblog post, no one but you replied or seems to think its an issue.  Over a year ago, I wrote and discussed the issue with 2 academic authorities, Martzloff formerly of NIST, and Dr. Standler, and why the industry has gone in bizarre ways.

http://www.eeel.nist.gov/817/pubs/spd-anthology/index.html

http://www.rbs2.com/fire.htm#anchor000001

http://www.amazon.com/Protection-Electronic-Circuits-Overvoltages-Standler/dp/0486425525/ref=cm_cr_pr_product_top



The issue of poor implementation is widespread for consumers, and less so in professional installations.  The 2 biggest problems are:

Joule rating hype in MOV driven suppressors

"Serial mode" suppressors

Anyway, the Consumer Product Safety Commission has extremely rare reports of fires caused by SPD devices since 2000, so from UL's viewpoint, safety has improved, but function remains, unknown.

The only maker I trust these days for home plug in suppressors is APC, but I still disassemble them and analyze the design of a particular model before I buy more.  So far, I've not had one that was badly made.

The once revered Tripp Lite, famous for its Isobar and metal enclosure, well I've had 2 that were very badly made, again the only way to find out is to open them.

Belkin, Philips, and other names not associated with electrical manufacturing, caveat emptor.  An anecdote from Amazon about Belkin versus APC:

http://www.amazon.com/Belkin-F5C895-TEL-SurgeMaster-8-Outlet-Protector/dp/B00000J1RU/ref=wl_it_dp_o?ie=UTF8&coliid=I3RZVU2J95MR80&colid=EIWMTQB99Q0R


Although crude and sensationalistic, this website covers the key issue with photos on consumer SPD fires:

http://www.sfowler.com/investigations/Surge%20Protectors.htm


The good news is that newer designs for MOV are more fire resistant.  Also the UL revision for SPD from early 2002 provide more fire retardant issues in designs, so to have UL SPD approval, you need to implement these changes.  Alas, a simple solution is to tape the MOV together so if it flames out or explodes, the hot shrapnel is contained and keeps from jumping into the power rails.  Fast blowing fuses are now built into some MOV, or discrete devices are placed inline with them, so if they conduct, the fuse will blow before the MOV does.

In APC designs made to conform to the most recent UL mandate, blast containers are built into the casing, so when the cases are joined, the blast compartment encloses each MOV set.  There are no screws inside many APC SPD, the reason being if fire or an explosion does occur, the metal screw doesn't fly inside as another piece of shrapnel.  APC PCB are snapped together inside the plastic case, and IMHO personally tracking what could happen, if a fire did occur, the whole blast casing and the body will melt down into the PCB, potentially suppressing the fire.

Alas, again its presumptive, to be sure we need a destructive test to see it actually fail.

Why fires occur isn't because the SPD becomes a hand grenade, but because most SPDs are mounted behind furniture, or on rugs, often covered with lint or dirt.  If it gets hot enough, it will ignite items around it.

I used to work as an R&D engineer with a surge suppression manufacturer and would be really interested in this.

I've seen some shockers (no pun intended) over the years.

My basic observation would be, like in all things -you get what you pay for. Buy a $20 plastic multi-outlet power board and expect your house to burn down. Buy a + $100 metal housed version and you can be reasonably sure your investments will be substantially protected.

Having said that there are some awful tricks and misinformation in this industry. Take "transient response technology" for example. It was invented so that a particular manufacturer's products would pass UL approval, at the detriment to actual equipment protection. The device would only clamp fast pulses, it would not respond to slow over voltages (e.g. +150% supply voltage @ 50Hz). So your equipment would not be protected if two of the three phases shorted out or the neutral lifted down the street, but the device has a nice confidence inducing UL logo on it.

Another easily identified trick is to list "total protection" or the sum of the surge rating for each phase, when in reality each phase is only protected by 1/3 of the rating.

I know where you could get access to a 20kA 8/20µs surge generator if required. It's not portable, you'd have to travel to it

[tomlut]

This is in reply to alm's post. Will get back to you ASAP saturation. But as usual the life of an oncall tech is never dull. Duty calls...

It's not that hard to test. The let-through voltage can be measured easily and the CBEMA curve can be used to find a pass/fail voltage (for the output pulse width).

Your point about the low impedance of the mains supply is an interesting one. It is only low impedance at low frequencies. Once you get to the end circuit of an ELV installation the inductances can be large enough to generate substantial ringing voltages even occasionally from induced currents from nearby lightning strikes.

There is absolutely something that can be done about slow over voltages. Let the protection device self destruct. It should (safely) shunt the current and trip the circuit breakers, just like a crowbar circuit. e.g. As long as the output voltage can be limited until the breakers trip your $10,000 home theatre setup is protected and it only costs you the price of some new surge protection ($100s range for domestic). A transient discriminatory protector will let your expensive equipment bare the brunt of the fault but be fine for the next fast impulse. You tell me which would you prefer?

The particular example of fire risk I was referring to was the example of a couple of unfused 8kA disk MOVs across the A-N lines of a $20 plastic power board. MOVs will flame if overloaded under certain conditions and all manufactures of MOVs recommend a metal housing and appropriate fusing as a minimum precaution. If you bought a $120 metal enclosed protector you would not be exposed to this risk. They are not EMI filters as such (though they will offer some low pass filtering). Generally the good ones have a high (10 to 40kA) main surge protection element, a big chunky low pass filter (not at all like the little silver boxes with an IEC connector I think you are referring to) and then some secondary surge protection in the order of 5 to 10kA.

The cost of UL approval means there are some excellent protection devices out there without this qualification - if you know what you are looking for. Also there were moves to close this particular loophole as I left the industry. Haven't followed it up lately.

[alm]

Your point about the low impedance of the mains supply is an interesting one. It is only low impedance at low frequencies. Once you get to the end circuit of an ELV installation the inductances can be large enough to generate substantial ringing voltages even occasionally from induced currents from nearby lightning strikes.

I didn't consider the effect of the inductance on the output impedance at higher frequencies, fair enough.

There is absolutely something that can be done about slow over voltages. Let the protection device self destruct. It should (safely) shunt the current and trip the circuit breakers, just like a crowbar circuit. e.g. As long as the output voltage can be limited until the breakers trip your $10,000 home theatre setup is protected and it only costs you the price of some new surge protection ($100s range for domestic). A transient discriminatory protector will let your expensive equipment bare the brunt of the fault but be fine for the next fast impulse. You tell me which would you prefer?

Sacrificing the life of the SPD is obviously fine, although it sounds like that method does require careful matching of the fuse and the MOV. Does the impedance of a MOV drop fast enough to blow a fuse before melting the MOV? From a quick peek at the first MOV datasheet I found, it seems that even a 110V MOV (probably too low for 115V mains) only starts conducting more than 100A (which might blow the fuse in reasonable time) at 300V or so, although no mention of the influence of increased temperature on this. A thermal fuse would obviously make this easier.

The particular example of fire risk I was referring to was the example of a couple of unfused 8kA disk MOVs across the A-N lines of a $20 plastic power board. MOVs will flame if overloaded under certain conditions and all manufactures of MOVs recommend a metal housing and appropriate fusing as a minimum precaution. If you bought a $120 metal enclosed protector you would not be exposed to this risk. They are not EMI filters as such (though they will offer some low pass filtering). Generally the good ones have a high (10 to 40kA) main surge protection element, a big chunky low pass filter (not at all like the little silver boxes with an IEC connector I think you are referring to) and then some secondary surge protection in the order of 5 to 10kA.

I wasn't suggesting that SPD's are/contain EMI filters, but that there might be plenty of other cheap $20 electronics that catch fire in case of a transient. For example, the capacitors in EMI filters in some old US-made 110V/220V equipment (thirty years old or so) can't really handle 230V, and blow up. I imagine some of the cheap crap capacitors might do about the same if a transient occurs, although I'd hope these wouldn't get UL approval. Either a cheap SPD catches fire because of a transient, or your cheap piece of electronics catches fire because of the transient, what's the difference? Is the former that much more likely?

[Simon]

I was thinking, ot would be good if dave did a blog on these devices, on the various types and what values to use for particular surges

[Kiriakos-GR]

Panasonic its the master at Surge Protectors for their fax machines in the last 20 years ..
I had repair many Panasonic fax machines , after an surge strike , that came from the telephone cables ..
The internal circuitry of the protector , took all the damage , and the fax device survived .. with no other damage ..  by replacing the broken parts at the Surge Protector with new , the fax machine was good to go. ( The connection with the telephone line was restored )   

The telephone lines are the sensitive ones and needs Surge Protectors .

The 220 Mains are usually well protected ,  and from the side of the Power company ,
and from the electrical system of our homes too. 

I do not trust at all the surge protective device with sockets,
and i believe that its just marketing to push sales ..
The only way to  discharge one surge , its to ground it ..  the thin cable of those surge protective device with sockets , are unable to do it ..   or they can partially handle only small surges .

If some one says to you ... Get it for partial protection ... will you ever pay the cash ? 

[saturation]

I think both your analysis show why, beyond destructive testing, a knowledgeable tester can peruse the guts of a surge protector and reasonably assume it will do what it intends to do.  The circuit design itself is fairly simple, you can easily trace the PCB and decipher it.  Over time, the design for most plug in SPD are pretty homogenous, yet despite that you'll still find variation in how its done and the quality of assembly, which finally determines if it addresses the issues you raise here.

The better SPDs have thermal fuses and current fuses linked to the MOV.  More recent designs use newer MOVs with thermal fuses integrated into the part. In older designs, you'll see the thermal fuse heat shrink wrapped with the MOVs.

http://www.worldproducts.com/TVZIntro.htm

To prevent overvoltage and causing the SPD to self destruct, more expensive SPD include a small voltage regulator circuit made to crowbar the SPD down until the fault terminates.  This is useful in areas where power line quality is poor, like in rural areas, or if there is a remote chance of power lines being downed by storms causing loss of neutral etc.,.  Panamax popularized this mode.

Another approach is to put a low cost AC voltage regulator before the SPD that's plugged into your device.   It does what the Panamax does for nearly the same cost, but it will continuously output and regulate AC until it reaches its shut down voltage.

UL addressed the issue of flame with the revised specs, so the metal enclosure is increasingly rare in UL certified plug in SPD, now called the Type 1.

The "whole home" surge arrester mains installed SPD has it own set of criteria.

http://lit.powerware.com/ll_download_bylitcode.asp?doc_id=13326

Beware of counterfeit parts, assemblies and UL stickers.  Improper labelling: a UL approval for power cord, with a legitimate sticker displayed by the maker, but the device is box labeled as an SPD.  Its a reason to be wary of SPD sold without any means of tracking its distribution channel, like on eBay: treat it like any quality electronics part you purchase.

Your point about the low impedance of the mains supply is an interesting one. It is only low impedance at low frequencies. Once you get to the end circuit of an ELV installation the inductances can be large enough to generate substantial ringing voltages even occasionally from induced currents from nearby lightning strikes.

I didn't consider the effect of the inductance on the output impedance at higher frequencies, fair enough.

There is absolutely something that can be done about slow over voltages. Let the protection device self destruct. It should (safely) shunt the current and trip the circuit breakers, just like a crowbar circuit. e.g. As long as the output voltage can be limited until the breakers trip your $10,000 home theatre setup is protected and it only costs you the price of some new surge protection ($100s range for domestic). A transient discriminatory protector will let your expensive equipment bare the brunt of the fault but be fine for the next fast impulse. You tell me which would you prefer?

Sacrificing the life of the SPD is obviously fine, although it sounds like that method does require careful matching of the fuse and the MOV. Does the impedance of a MOV drop fast enough to blow a fuse before melting the MOV? From a quick peek at the first MOV datasheet I found, it seems that even a 110V MOV (probably too low for 115V mains) only starts conducting more than 100A (which might blow the fuse in reasonable time) at 300V or so, although no mention of the influence of increased temperature on this. A thermal fuse would obviously make this easier.

The particular example of fire risk I was referring to was the example of a couple of unfused 8kA disk MOVs across the A-N lines of a $20 plastic power board. MOVs will flame if overloaded under certain conditions and all manufactures of MOVs recommend a metal housing and appropriate fusing as a minimum precaution. If you bought a $120 metal enclosed protector you would not be exposed to this risk. They are not EMI filters as such (though they will offer some low pass filtering). Generally the good ones have a high (10 to 40kA) main surge protection element, a big chunky low pass filter (not at all like the little silver boxes with an IEC connector I think you are referring to) and then some secondary surge protection in the order of 5 to 10kA.

I wasn't suggesting that SPD's are/contain EMI filters, but that there might be plenty of other cheap $20 electronics that catch fire in case of a transient. For example, the capacitors in EMI filters in some old US-made 110V/220V equipment (thirty years old or so) can't really handle 230V, and blow up. I imagine some of the cheap crap capacitors might do about the same if a transient occurs, although I'd hope these wouldn't get UL approval. Either a cheap SPD catches fire because of a transient, or your cheap piece of electronics catches fire because of the transient, what's the difference? Is the former that much more likely?

[kd5jha]

I use quite a few of the tripp lite isobar protection devices at my work, and I have found them to be very reliable... We typically use these three models; ULTRABLOK428, ISOBAR8ULTRA, ISOBAR12ULTRA. These are the higher end units that have the larger MOV's and Inductors...and we seem to have really good luck with them.

These units are being used in very high risk situations, where they have to protect from surges over and over, and while I lose two or three of these per year at most sites, they generally protect the equipment that was attached to them.

I used to work at a High Voltage laboratory years ago when I was in school, and we had a huge impulse test set (3000Kv,56KJ) see it here ( http://www.ece.msstate.edu/hvl/image_03.gif ) I wish I had thought then to try a little test on some of these high dollar surge protection devices, especially the ones for Home Theater installs...

In general the main point of failure I have noted on the ISOBAR series is the crappy switch that they are using. You can replace it out with a better switch and these things will last for quite a while. I use these at broadcast tower sites where they get popped all of the time, and while we implement other modes of surge protection as well, they do get tortured.

I find that your best bet is to layer protection, and also have good grounding practices in place using properly engineered star grounding techniques, and checking the quality of your connections on a regular interval. Also don't forget that you must protect all connections as others have pointed out...Phone, RF, Power, ETC... Lightening will find it all...

Obviously this is a discussion worthy of several articles, or even a book...but I would say that at a minimum the ISOBAR units are worthy of consideration.

On our phone lines we use two different measures of protection in series, i use a MOV based protector from ITWLinx seen here ( http://www.itwlinx.com/productpg/surgegate_modular_seriespg.php ) which will protect up to four lines (they make a good power strip protector too) and at particularly in jeopardy sites I use a device called an Optelator that optically isolates the phone line for almost total protection... With these two devices we pretty much have eliminated damage to modems and telephone devices.

RMG 

[Zero999]

Fire shouldn't be a problem if the surge protector is fitted with both a thermal fuse and an electrical fuse.

I had a surge protector which suddenly stopped working. I opened it up and I noticed that there was a thermal fuse next to the MOVs which had obviously done it's job. The trouble is when one buys a surge protector, there's no way of telling whether it's actually real,fake or has a thermal fuse until they get it home and can open the case.

[saturation]

Yes, assuming you know how to open it and look inside.  Last week, I just had a major mains problem protected by my low cost PER7 APC that was over 10 years old.

The mains zoomed to over 300 VAC, for whatever reason, and one fuse atomized into nothing one MOV exploded spreading metal particles all over the casing, while briefly bursting violently into flames and smoke.  I saw a short flash come out of the casing.  All my plugged in chargers survived, with the APC sacrificing itself.  For $10, it was good insurance.

Photo descriptions:

001: immediately on opening the case, not the location of the black sooth from the MOV flash, the white HRC fuse is intact

004: The MOV ruptures near the top and splits in half, fuse F5 is atomized, while the main fuse, upper left, is still visible, I haven't tested it for functionality

009: closeup of the rupture

011: singed insulation on black cord

013: focus of the blast onto the casing

014: closeup of the failure point on the MOV

015: other MOV appear intact, each protects one pair of lines: GN, NL, LG, so its easy to locate the major fault line

016: details on the APC PCB, dated 1998

018: more on the MOV failure point

020: flash and particles on the casing

Fire shouldn't be a problem if the surge protector is fitted with both a thermal fuse and an electrical fuse.

I had a surge protector which suddenly stopped working. I opened it up and I noticed that there was a thermal fuse next to the MOVs which had obviously done it's job. The trouble is when one buys a surge protector, there's no way of telling whether it's actually real,fake or has a thermal fuse until they get it home and can open the case.

[slburris]

So what do you think of this Tripp Lite strip available at amazon?

http://www.amazon.com/Tripp-Lite-ISOBAR8ULTRA-8-Outlet-Protector/dp/B0000511U7



Looks to have most of the features talked about in this thread, no?

Scott

[saturation]

Yes, it has.

Tripp lite Isobar is/was one of the best surge protectors out there; it has all the best in SPD design.  A problem I've seen has been in this one tripp Lite Model, made in China, it was so shabbily made it makes me wonder about all their products.  This model is rated 5 stars by Amazon buyers, but nothing is about its surge protection capacity, but the layout of the sockets!



Isobar is now made in USA, but if you buy one open it and check the construction as Dave does with many electronic items.  My guess is USA QC is much better given our laws.

In contrast, I inspected over 10 various APC models, and not one had a QC defect.  So, I now buy all APC, if you compare price, APC is bang for buck, but I still open the casing  

A single key feature to look for in buying any SPD is screw on covers, to allow you inspect its guts.  Many commodity SPD, usually Belkin, Phillips and other general consumer electronics names, now glue the case shut, absolutely stay away from those.

Quick summary of the components:

The thick blue colored MOV on the left end of the PCB are likely built in thermal fused MOVs, while the thinner MOV on the right end, are standard MOVs.  A typical design is to put a thermal fuse fairly close to the MOV so it will catch the heated MOV and trip.  The various inductors are just filters to cut down on various noise sources, like switching PSU, that permeate line voltage these days.  Good to have, and standard issue in $30 up SPDs.  All my APC with similar filter out the control signals of my X-10 home automation system  




The Isobar claims it has isolated filters and SPD for some banks within their SPD, those do work in suppressing noise of some value in high end audio systems.

My biggest critique is the price.  Its very close to low end Panamax SPD, which offer a voltage regulator that shuts down the entire SPD should voltages exceed parameters.  You can check it out.

the metal container used to be a massive plus, when SPDs didn't have fuses to blow too and limit the flash burn of the MOV, which could penetrate the plastic housing and potentially torch your sofa, or ignite lint and dust, and start fires.  It becomes more of an issue with higher joule ratings, because the potential energy released by a high joule rated MOV is more violent than a lesser one.


Beware Joule rating, its useless:

The MOV in the PCB appear in parallel suggests they are on the same line, but can't say for sure, paralleling MOV is a ploy used to increase the joule rating protecting a particular line, such as line-neutral, but its nearly useless electrically, its mostly for truth in marketing.  The reason is, parallel MOV do not have identical impedance, so faced with a voltage spike or overvoltage, they will not conduct at the exact same trigger voltage, so likely one work first before the others.  Since that MOV will  go into near zero impedance, this single MOV will likely blow before the others have a chance to act, its not like having parallel resistors.  The whole idea of high joules is allow the SPD to take a beating and still live for another day, rather than blowing like a fuse and dying.  Alas, every high voltage insult reduces the life of the MOV, so the same 'sensitive' MOV in the parallel set will eventually be the first to fail, in response to multiple overvoltage or transients.  

Its better to have a lifetime warranty on the whole device, a good company to honor that warranty, and a single standard MOV like the PER7 design, than pay for multiple MOV in parallel when protecting power lines and grounds.

These days, the competing APC model to the Isobar shields the MOV in a plastic blast shield inside the all plastic casing, UL testing says this is enough, but only a field trial will tell for sure.  The failure I had demonstrates in real world use, how valuable the thermal fuse is, and that plastic seems to be enough.  A competing APC model sells for $20-30, so for a single Isobar you can get 2-3 APC:



See the next post for the final post mortem on my APC PER7.

So what do you think of this Tripp Lite strip available at amazon?

http://www.amazon.com/Tripp-Lite-ISOBAR8ULTRA-8-Outlet-Protector/dp/B0000511U7


Looks to have most of the features talked about in this thread, no?

Scott

[saturation]

I did some troubleshooting, and here's my finale on the PER7.

Note, my SPD tripped because I took it to a country with questionable electricity.  I always travel International with a disposable SPD, and it paid off, after over 10 years of owning it, and over 20 years of using SPD, one has finally tripped! Now, to try using the lifetime gaurantee APC always claims.  Will post whether this is true or not, many users claim its easy to get a new replacement but harder to get reimbursed for plug-in equipment damage.

The MOV that failed was on the LG line, so the voltmeter was right, the excess line voltage did kill this PER7.

The thermal fuse connects the dead MOV to the ground line, shunting the excess voltage out of the circuit.  It too blew, causing this PER7 to have a permanent ground fault light on.

I've located bits of the fuse, see the pictures.  Also the flash blackened the casing, but did not penetrate it nor pose a hazard to chair, rugs, or dirt surrounding the SPD, at least at the time.  While the casing does not have a dedicated blast shield, when the 2 halves are joined, the electrical lines are mostly isolated from the PCB.  But, not entirely.  You can see some debris made in into the power line section, and could cause a short.  This is why the PER7 also has a separate fuse, and a circuit breaker, neither of which tripped.

PER7, adieu, a job well done.

[saturation]

APC, true to its word.

I called APC about my PER7 failure and its lifetime warranty.  They said no problem, took my word for it, and shipped me a new one!  Free S&H both ways.  Old one gets shipped back.

Way to go.  I bought this PER7 in 1998.

[saturation]

I received a new PER7 via Fedex in 3 days.  A tear down reveals the lower parts count, but overall, better design.

The unit is now made in the Philippines with a hand signed QC and testing certificate!  For a $10 item!  There is a growing trend to transfer manufacturing out of China.

#1: smaller PCB, but large ground planes covered in solder.

#2: HV gaps now built into PCB, MOV and thermal fuse are shrink wrapped together for better conduction of failed MOV heat and faster tripping of the thermal fuse

#3: fuse can be seen inside the heat shrink tube

#4: PER7 has been in production since early 1990s.  The unit above is a 2006 Philippine version, the unit below is the 1998 version.  The PER7 is a signature design for APC, its so distinctive if you see this shape, chances are its an APC SPD.  The casing is reused for the more advanced SPDs, using the free space inside the casing.

For cost reduction, the latest certification information is no longer embossed into the case, as in the 1998 version, its simply a decal.

#5: IMHO, a flaw in most APC designs.  APC advertises its 'fail safe' which will kill the power to the sockets should any of its SPDs encounter a major surge. 

However, in tracing the PCB, several of their designs only have the Line-Neutral line connected to a LINE thermal fuse and circuit breaker, so if tripped, disables the sockets and KILLS the GREEN LED that signals, PROTECTION ON.  When the LG, or NG fuse break, the RED LED, GROUND FAULT, lights, but the GREEN LED can still be lit.  So, its not really 'fail safe' but the simple design does signal something is wrong and the SPD should not be used.

[knreddy]

Mx3264 with 14mm mov 115j but specified as 2000j 

[knreddy]

Mx3264 with 5 sides aluminium body with individual illuminated switches but 14mm sigle mov of 125j wrongly rated as 2000j at flipkart and the EMI / RFI filtration is also minimal.

[knreddy]

MX3264 primitive filtration

[knreddy]

Mx3264 primitive filtration