Surge protection strip failures

[G7PSK]

Hi. I have a number of said surge protectors in which the green protection warning light has gone out. Today I took 1 apart and found that the neon tube had gone white so out of curiosity I dismantled the other ones 3 were white as well and one was black with obvious heating of the current limit resistor in all but one the neon was directly across the lines in a Belkin model the neon was switched by a semiconductor of some type controlled by the MOV network that was one of the white neons (air in the tube I presume) In none of the strips did I see any obvious failure of the MOV's so are they safe to use as a surge protector or are the neons used in the warning lights all leaky the red power on lights in all but one unit work fine and only show a little darkening of the tube on the other one the neon is flickering. I cant but help think that these neons are deliberately made with a short life as the instructions all state replace if the light goes out, but as most of them are no more than power on lights it leaves me wondering.

[SeanB]

Probably a very overdriven neon lamp so that it will fail within a period just outside the warranty period. You can check the MOV units for breakdown voltage and see if they are still good, then simply replace the neon and the resistor, using a higher value. The green neon lamps have a phosphor coat on them to make green light from UV light inside. The Belkin one uses a thermal fuse to disconnect the neons and the MOV devices if the MOV units overheat due to either overload or EOL. If the thermal fuse operates replace the MOV units as well as the thermal fuse.

[G7PSK]

That's kind of what I thought. I have just been replacing them when the light goes out as they carry a warranty against damage to any equipment connected, I have never tried to claim as no equipment has failed due to surge whilst unseeing one of these power strips. So the units with the light failures have been relegated to plain multiway connectors.
I dont realy have any way to test MOV's could I use my insulation tester.

[T3sl4co1l]

Green "neons" aren't neon at all, or not much I think: they are actually fluorescent tubes with a green phosphor.  I don't remember if neon itself ever gives off enough UV for the job, or if they use a different fill gas; argon, xenon or mercury seem likely options.

Like neons and other gas discharge tubes, the breakdown voltage rises over time, and the light output falls due to sputtered electrode material darkening the inside (which you probably wouldn't be able to see in a phosphor-coated device).  They can be replaced if blinking or dim (assuming the surrounding circuit is still good).

Tim

[G7PSK]

Apologies for being too verbose, they say a picture paints a thousand words unfortunately due to the commonwealth games and school holidays coinciding my internet is so slow I could not upload pictures, eBay and Amazon only loaded as text.
I know that all lamps are not neon it is just easier to refer to gas discharge lams used for indicators as neon just as neon signs are not neon.
I checked the mov's with an insulation tester and at 250 and 500 volts I have around 4 M ohms at 100 volts that goes down to 3 M ohms but the power on lamp lights as well, I did not bother to take the unit I tested apart again to make the test just connected to the plug.
The green lights should be called buy another now lamps as when they go out that is what is expected by the manufacturers.   

[allikat]

The thread title looked interesting but I couldn't get through the wall of text. Can you summarize? Then you might get more responses.

In other words - the green lights that show his surge protected strips are good have failed.

2 causes for this were proposed.

1: The neon/fluorescent bulb has failed - solution - replace bulb and check for cause 2.  Possible modification - increase value of series resistor by said lamp to increase lifespan after repair.
2: The MOV has failed or got hot and triggered the thermal fuse (most likely while doing their job). Solution - replace MOV and fuse then check cause 1 again.

Not really adding anything here, just cutting to the chase.

[Furcot]

I suspect foul play here, as most of the manufacturers say something along the lines of "when the green light goes out the device must be replaced". As others have said the green neon is indeed the same as the red neon, just with green phosphor on the inside. I have seen many of these devices, and the problem is always the same, that the "the green light has gone out". In all the devices I have seen the red neon has a series resistor of 270K -this is correct and gives a service life of some 30 years. The green neon however has a series resistor of 120K and dramatically overruns the green neon, such that it fails very early indeed (the phosphor often turning black). I find it something of a coincidence that 120K is the correct value for the 115V US market, could this be an excuse if this ever came to light..

[MosherIV]

Hi

As many have said, when the MOVs overheat they trigger the thermal fuse. This should make the green 'neon' go out. Bigclive has done a very good tear down and explaination



Testing the surge protection with over voltage will shorten the life of the MOVs and shorten the life of the surge protection. MOVs can be tested by measuring their resistance, they should be in the mega ohm range, anything less means they are bad. They must be tested out of circuit, ie removed from the surge protection.

Each over voltage event will cause deteriation in the MOV. So after a number of years a surge protector will fail because there are always small surges. The green indicator is there to tell you it needs replacing (or repairing if you know how/want/can be bothered/want to take the risk/etc)

[floobydust]

Just a reminder, old recall of 15 million APC power strips for fire hazard. It was 1993-2003 models.

Incidents/Injuries:
"The firm has received 700 reports of the surge protectors overheating and melting and 55 claims of property damage from smoke and fire, including $916,000 in fire damage to a home and $750,000 in fire damage to a medical facility.  There are 13 reports of injuries, including smoke inhalation and contact burns from touching the overheated surge protectors."

https://www.cpsc.gov/Recalls/2013/schneider-electric-recalls-apc-surge-protectors
Might be this problem: http://www.kerrywong.com/2018/06/10/unusual-failure-mode-of-an-apc-surge-protector

[james_s]

Green "neons" aren't neon at all, or not much I think: they are actually fluorescent tubes with a green phosphor.  I don't remember if neon itself ever gives off enough UV for the job, or if they use a different fill gas; argon, xenon or mercury seem likely options.

Like neons and other gas discharge tubes, the breakdown voltage rises over time, and the light output falls due to sputtered electrode material darkening the inside (which you probably wouldn't be able to see in a phosphor-coated device).  They can be replaced if blinking or dim (assuming the surrounding circuit is still good).

Tim

They have a mix of neon and argon, more argon than the ordinary orange ones and it's the argon that produces UV to excite the phosphor. Technically any lamp that uses a phosphor to convert one wavelength of light to another is a fluorescent lamp although the blue and green ones do still (in most cases anyway) contain some neon, sometimes you can see a dim orange glow through a thin spot in the coating.

[tooki]

Green "neons" aren't neon at all, or not much I think: they are actually fluorescent tubes with a green phosphor.  I don't remember if neon itself ever gives off enough UV for the job, or if they use a different fill gas; argon, xenon or mercury seem likely options.

Like neons and other gas discharge tubes, the breakdown voltage rises over time, and the light output falls due to sputtered electrode material darkening the inside (which you probably wouldn't be able to see in a phosphor-coated device).  They can be replaced if blinking or dim (assuming the surrounding circuit is still good).

Tim

It’s also quite common in surge protectors to simply use a regular neon behind a green lens, despite the inherent silliness of such an arrangement. They look as crappy as one would expect, but I guess they do the job, as far as being able to refer to the “green” light in the manual.

[Zero999]

I don't see how a green lens would cause it to look green, because neon doesn't emit any light in the green region.

They have a mix of neon and argon, more argon than the ordinary orange ones and it's the argon that produces UV to excite the phosphor. Technically any lamp that uses a phosphor to convert one wavelength of light to another is a fluorescent lamp although the blue and green ones do still (in most cases anyway) contain some neon, sometimes you can see a dim orange glow through a thin spot in the coating.

I remember seeing a green neon and it had a purplish glow to the part of the lamp, without the phosphor.

Are you sure they don't have any mercury? If no, then why haven't mercury free fluorescent lamps been invented with neon and argon?

[james_s]

I don't know for certain whether they contain mercury or not but I don't believe they do. Mercury free fluorescent lamps have existed since the dawn of gas discharge lighting but they're not bright enough to be used for anything more than indicators.

I've also seen the orange neon behind a green lens a few times and yes it looks terrible, the light is still orange but you can tell the lens is green.

[Zero999]

One thing I've noticed is the green neon lamps don't last as long as the orange ones do. I always thought it was because they contained mercury which gradually absorbed into the glass and electrodes or is there another explanation?

[T3sl4co1l]

Nah, mercury doesn't absorb anywhere; also if it amalgamated with the electrodes, it would do that anyway.  But amalgam also has a vapor pressure, so it wouldn't fail outright.

AFAIK, failure is usually escape of helium, and something about polarization of electrodes that occurs after a lot of sputtering of the surfaces.  Which also deposits material on the glass, blocking some light output.

Tim

[james_s]

They were a bit before my time but I remember reading that those screw-in UV glow lamps that use a sign lamp envelope don't last as long as their neon counterparts. Now I'm going to have to find the book on glow lamps I have somewhere and see if that says why. The ordinary orange ones have a Penning mix which is neon with a bit of argon.

In Nixie tubes mercury is used in some of them to greatly extend the life. Apparently it creates s replenishable coating on the cathodes IIRC. I remember reading that despite appearances, the glow discharge is quite a violent process at a microscopic level, tearing atoms of metal off the surface of the cathodes.

[Zero999]

Nah, mercury doesn't absorb anywhere; also if it amalgamated with the electrodes, it would do that anyway.  But amalgam also has a vapor pressure, so it wouldn't fail outright.

AFAIK, failure is usually escape of helium, and something about polarization of electrodes that occurs after a lot of sputtering of the surfaces.  Which also deposits material on the glass, blocking some light output.

Tim

What about when fluorescent lamps fail due to mercury depletion? The argon starts to ionise instead, causing the lamp to dimly glow hot pink, rather than white. Where does the mercury go?
https://www.wikiwand.com/en/Fluorescent_lamp#/End_of_life
https://www.lighting-gallery.net/gallery/displayimage.php?album=4353&pos=3&pid=150288

[floobydust]

I believe mercury reacts and deposits on the phosphor, this is what makes fluorescent tubes blacken at the ends, with the corresponding increase in resistance.
Green neon lamps, Don Klipstein said the emitted spectrum has the line for mercury.

I've only seen green LED's in power strips, and they age rapidly due to the 1/2 wave pulsing ac they are fed with, and the series resistors run quite hot.

[T3sl4co1l]

That would make sense, and there's a higher surface area of phosphor and extremely little mercury used in fluorescents in the first place.  Although I still suspect it's not so much surface adsorption, because that should equilibriate quickly -- but diffusion into the phosphor crystal, and for that matter, perhaps not diffusion at all (depends on temperature) but ionically driven (depends on plasma temperature).

Hmm, I wonder if the phosphor can be annealed at temperatures below the softening point of glass.  It should release the mercury and fix whatever defects there are in the crystal structure (the electronic transitions that allow phosphors to work in the first place, can be short-circuited by other impurities, and defects, which annealing would fix in this case).  That last part may not be possible, or practical, though, as a lot of metal oxides have very high melting points.  Of course one also has to consider diffusion and desorption of air through the glass and electrodes, spoiling the vacuum before any of these benefits occur.  And, not that this does anything to help the well-worn filaments in the bulb, which cannot operate as cold cathodes for very long.

Tim

[tooki]

I don't see how a green lens would cause it to look green, because neon doesn't emit any light in the green region.

I didn’t say it looks green, I said it looks crappy.

The cheap tinted plastic lenses are nowhere close to ideal color filters, so they let through a LOT of light of other wavelengths. So they let through plenty of orange light. It’s just feeble and looks like a dim orange light glowing from a sickly greenish background.

[james_s]

The blackening at the ends of fluorescent lamps is due to tungsten sputtered off the cathodes depositing itself on the inside of the glass. This is greatly accelerated if you have an insufficiently heated cathode, for example if a rapid start tube is inserted so that one of the pins on one end is not making contact. I don't actually know offhand what happens to the mercury but it is indeed a problem in many of the low mercury tubes, they have just barely enough to last a reasonable period of time. It may well get buried under sputtered tungsten.

[floobydust]

I have seen CCFL tubes do the same, blackened ends and they have no filament - so I figured mercury buildup is the reason. Not that I'm an expert on ionized gasses and sputtering.

Interesting to see nixie's "cathode poisoning" might be similar to OP's neon lamp failures (although DC not AC) where an insulating coating of impurities forms on the cathode electrode.

I found mentioned neon lamp lifetime is inverse to the 4th power verses current, so pushing them hard for high brightness really kills their lifetime.
Every cheap power strip I own has dead flickering neon lamps .

[james_s]

CCFLs have no filament but they have metal cathodes. Since these cathodes are unheated (hence the cold cathode name) they are even more prone to sputtering.

It works the same as vacuum deposition mirrors. Heat some metal in a vacuum and the vapor will deposit on nearby cooler surfaces.