Hey guys, in a post by dor123 about a blinking indicator lamp on his refrigerator in which he posted a video, I replied as I noticed the same phenomenon occurring in power strips with the little neon glow lamps on the "on" switch.... I had one behave normal for 5 years and flicker for the last 10 lol so I can't confidently say it is EOL lol... Well a fellow member Naomi said she has seen the same phenomenon and noticed it change depending on the load connected to the power strips....interesting!  I need some brainstorming on this as well as a VERY odd phenomen that occurred back at my parents old home in my room.  I had a ceiling fan that was ways on low.  At about 2-3am each night it would emit a pulses buzzing sound that would change pitch (and at first scared me) very very much like if you connect an audio amplifier output to an electric motor (you hear music albeit crappy), but the fan would literally eminate harmonics as if a dirty audio sine wave with harmonics were superimposed with the 60hz 120v ac. It was weird to say the least.  At this exact instant, the IR receiver on my tv would blink indicating it was receiving a signal of some sort (but the TV set would not react by any action). Every night on the dot around 3am just once (I was ways a night owl) lol.  Any brainstorming is appreciated! :-)

@Fan noise at 3AM: I would guess the origin was coming from a communication signal used for network management (the down link; e.g. high/low tariff switching control,...) - utilities use few kHz signals modulated by about 1bps, the amplitude is about few percent of the normal mains voltage. The frequency is on purpose shoosen so, it does not "divide" with the mains frequency, so it is possible to get the real carrier signal from all the mains harmonics in the receiver.

Now it may happen, than in your fan the phase shift capacitor, together with the motor leakage inductance formed a kind of resonator, what by (bad) luck happen to match the carrier frequency of the control signals. So higher amplitude of it builds up there, what then create intermodulation products with the main 60Hz (and it's harmonics; in the motor are plenty of nonlinearities for that - e.g. the play in rotor bearings) and those then create the creepy noises.
Something similar could happen in the TV - the signals arrive onto the input mains rectifier, what is highly nonlinear block, so there the intermodulation products are created. Those then modulate the stand-by supply converter, what then generate a signal disturbing the IR receiver.
I have problems with that with some HID's - some tend to flicker on that and sometimes even extinguish.


For the indicator: The discharge tend to be from nature unstable over cold electrodes of the neon indicators and mainly on lower currents, where it create rather smaller spots. These spots then tend to dance along the electrodes.
This is usually very sensitive on temperature - few degrees difference could alter the "dancing" quite a lot.
Now if the extension cord get heated by the high load connected to it, it changes the bulb temperature, so the dancing become different.

In many shelters here there is a high-RPM exhaust fan (async capacitor-run motor) of about 80w, with a switch that allows to change its direction by flipping the wires between the "main" and "capacitor" coils (actually the 2 coil and capacitor are connected together, and the switch only moves the Live wire between the 2 sides efectively swapping the roles of the coils). Many also ha a speed control which is basically looks like a standard light dimmer circuit

I liked to play with those - i let it run on max speed in  direction, and flip the direction switch to see how it changes, so it made about this sound

[full noise of air flow in one direction] - zzzzzzzzZzZzZzZz z z z zz zZZ zZZZ zZZZZZz zzZZZZZZZZZzzz zzzzzZZZZZZ[here is when the blades are stopped]ZZZZzzzzzzz zzZZZZZZZZZz zzzZZZZZz zZZZ  zZZ zz  z z z zZzZzZzZzzzzzzzzz - [full noise of air flow in other direction]

From this i think that if for some reason a "mismatch" appears between the speed of te fan and the line frequency, it will make those noises untill it reaches again the proper speed (which i not exactly a multiple of the line frequency in async motor, but looks like it still have speeds at which it is "ok with" and ones that dont)

So what might cause a change in either the fan speed ? (or the line frequency, but this is unlikely cause it would effect anyone downstream from the power plant, not even from the substation)

Maybe its power was interrupted briefly by a dip in the line power or by some electronic controller that switched itoff and on

Maybe the motor is sligtly overheating. Assume you switched it on at 00:00 and it takes 2-3 hours to reach its overheatd temperature, it will make a thermal protection cutout react at about the same time every day. The fan then begin losing speed, but quicky enough the thermal disconnect clicks back on, and you got the mismatched fan speed vs line frequency

More interesting effects can happen if the disconnect is miswired and cuts out only the run coil but not the capacitor coil - The capacitor coil passes some current that can make zzzz deending on the rotor's location (which keeps spinning with inertia), but can't do anything to actually keep up the speed - so you get te noises and possibly even loud ones

Maybe generally somthing going on with the run capacitor, but thats unlikely cause capacitor failures don occur in a constant pattern at the same time every day

Maybe three is some othre load sharing a neutral with the fan. The neutral is weak, so he other load when it switches at 2-3AM (by electronic control, thermostat etc) causes an undervoltage or overvoltage on the fan

As for the TV,i think it might actually be reacting to a glitc in the line voltage and not to the fan

@Ash: Ceiling fans run rather slow, so can not make such effects you described...

Thank you guys very much for the input thus far!  I am learning alot and appreciate it!  Medved, what kind of gear is controlled by those harmonics and what is the intent?  Ash I have messed with the reversable speed fans like that also! :-)

Thank you guys very much for the input thus far!  I am learning alot and appreciate it!  Medved, what kind of gear is controlled by those harmonics and what is the intent?  Ash I have messed with the reversable speed fans like that also! :-)

The communication signals are on purpose on different frequencies than the harmonics. The harmonics are generated by nonlinear loads (e.g. any electronic having a rectifier on it's input,...) and those would disturb the communication signals, if the carrier would match with the mains harmonics.

The signals are used for some basic functions, what are now part of what is called "smart grid".

The basic ones - controlling heavy loads on customer side during low electricity demand (heating, water heaters,...).
As an incentive to encourage customers to "hook up" their equipment to these signals is the installation of a meter with two registration dial sets, where one register consumption during peak demand, another in the valley between. And the "valley" counter is then billed at way lower rate, so during the valley periods the electricity become way cheaper for the user. And this make the investment into larger heat storage capacity heaters and control equipment (or even accept a bit of comfort loss for some) when opting for the dual tariff quite financially attractive.
The house electrical panel is equipped with the receiver unit, what read the messages and decodes, if they are intended for it and what it should do. It have a relay, what on one of it's contact control power contactors in the controlled loads and by second contact control an electromagnet in the meter, switching the gear to engage either the "high" or "low" tariff dials).
And viola, the grid operator have another load under his control to stabilize the grid...

The signals used here are in a form of one bit per second digital messages containing the address (so which receiver should respond) and the command (hi/low tariff, but it allow sixteen commands in total). The complete message takes about a minute and then another message could be send.
The rather low frequency is chosen on purpose so, it does not match the mains harmonics (so the receiver won't get disturbed so easily) and to pass through ordinary distribution transformers, as the injection is done usually at 110kV level, so there are at least two or three downstream transformers till it reaches the customer. This is actually very old system, dating somewhere to 60's till 70's (at least here).

Today, way more capable systems are being installed, mainly using high bandwidth bidirectional communication, what allow on top of the basic demand control faster and more accurate response, some diagnostics, remote meter readings. These system do not have as long range, but that is rather an advantage these days, as it is of no problem to install few repeaters, what may compact the data, generate commands on it's own, e.g. control the load based on actual transformer temperature, or so.
And the newest field in here are the "mains balance aware" controllers, intended to help to stabilize the network in very fast response by delaying/advancing e.g. the chiller/freezer thermostats. The delay/advances are in low minutes, so it mean no effect on the chiller's performance, but when aggregated among large number of such devices, it could save a lot of "fast response" generation power (power generation units run at partial load, but ready to kick in to full power within seconds, so e.g. require the boiler to be heated at full power). And because this is the most expensive (by money, but mainly by the environmental impact, as a lot of the energy is actually wasted) energy, such systems, even when controlling percents of the total load, are expected to cut the emission (all of them) by percents in total.

Thank you very much for taking the time to explain all of that for me! I truly appreciate your knowledge! :-)

I see this too.

I have a power strip made by Woods. Bought new in 1998, it's been plugged in every day since then, and the neon light started flickering last year.  It stays lit a lot of the time, but as soon as I turn the twin-blade window fan on, it flickers on and off like you wouldn't believe.  There's nothing to worry about.  It's just that the little neon light can't put itself out on it's own, so it just goes into a constant on/off flickering loop, depending on how much electricity is flowing through the power strip.

Yeah the slight voltage drop caused by the fan turning on could make the neon start flickering like crazy.
Those little neon indicator lamps can also flicker more or less depending on the ambient light in the room, it seems!

It does, if you put it in darkness it will stay off mostly

Wholly cow...so even the energy input of the ambient light from outside is enough to stimulate the neon into ionization when its on the verge like that? That's neat!

Its the most basic photoelectric effect

Metal in a lamp MAY release electrons under certain circumstances. Each metal have energy level known as binding energy. The electrons have less energy than that, therefore the metal can hold them back. IF an electron gets enough energy to overcome the binding energy, it is free to go

There are many ways to give the electron enough energy, some of which you allready know from other processes of lighting :

 - Pulling it out with force of electric field, as seen in Instant Start lamp starting. But.... The process may be destructive on the material from which the electron is pulled

 - Preheating the entire mass of metal, which gives the particles inside more kinetic energy

 - Using a material with extremely low binding energy as coating on electrode, so you have to use less energy to get electrons out of it, Those materials are known as emitters

Notice that none of the effects rely on the gass fill in the lamp. They take place as well in total vacuum, such as inside CRTs and electron valves

As long as electron is released from the electrode, low voltage is sufficient to pull it over to the other electrode - And indeed, most discharge lamps have arc voltages no more than 100-200v after they "get going", compared to KV's to pull the electron witrh force without help. Therefore, in most of the cases what we are waiting for is fo the electron to escape from the cathode somehow, and it is not because we did not supply voltage between the cathode and anode. We did, just not enough to blow up all the distance between them and make an Instant start



But sometimes none of this is enough. Glow lamps have no ignitors, no preheating and no emitter coating. Here comes the 4th way to release an electron :

 - Knocking the electron out with another particle which hits the electrode with the needed amount of energy

This other particle may be... a photon. each photon have amount of energy determined by its frequency, higher temp colors have higher energy. If the energy of a hitting photon is sufficient it will be able to release the electron it hit

In glow lamps however, manufacturers usually dont rely on external ligt but they include the source of stray particles inside the lamp - usually as a bit of radioactive gas added to the gas fill. The radioactive material emits particles all the time as it is dacaying, and those particles hit the cathode all the time

Things become interesting when the amount of radioactivity drops to critically low level as the material decayed in an old lamp. It takes a longer wait untill some more atom of the gas falls apart and emits something (multiply that by the chance that the particle emitted is even going in the direction in the cathode, as its random) so you begin to wait a while to see the neon light up, wait "ages" for old starters to even begin to glow (not the waiting untill they close), and so on

The light emitted in the lamp is also sufficient to get more electrons out, so once the lamp striked it will hold itself up... Except one thing, The neon here is working on AC. After each time the voltage drops and the lamp extinguishes, the "initial electron" must find its way out again. Sometimes it'll get hit by a particle from radioactive decay. Sometimes by a stray photon (even in "total darkness" - materials in the real world do emit tiny amounts of "ghost light" due to effects like phosphorescence, chemical and radioactive reactions, bacteria etc). Sometimes a surge in the line voltage would be just enough for a one-off Instant Start. Sometimes the lamp will wait in the darkness for next time

Ash thank you very very much for that,  you answered way more questions that I've had but forgotten about all at once.  Very much appreciated!  I now realize my flawed thinking of the gas needing the push, but it is actually the electrons from the electrode!  and you also answered how the arc voltage is so low! A question, neon tubing used in neon signs that run in the kV range....do they have any radioactive fill or emitter?  Or is a clear neon tube that glows red filled with only pure neon gas. Also is the radioactive fill of say an old 3 watt medium base GE Neon glow lamp (because large) dangerous to us?  Thanks again!

I am not sure but i think those contain no radioactivity at all, for 2 reasons i can think of :

The point of using radioactive materials is to make sure the arc will strike in complete darkness. It makes sense for a Ne indicator inside a light switch (which should glow so you can find it) or a starter (or you'll switch on the light in a dark room and nothing will happen - as it indeed does with some starters, especially old ones. Then you light a LED flashlight at the starter for a split second and it strikes). But it is never demanded from an outdoor Ne sign to strike in complete darkness - There is the street lighting, the moon, the ambient light from other sources.... Somehow the stray photon will be present, thats for sure. And thats before we take into account the phosphorescence from the phosphors present in some Ne tubes.... Actually, would be interesting experiment

The Ne sign making shops are relatively low tech industry, i dont think they can bother with all the health/safety issues in handling radioactive materials. Manufacturers like Philips are getting rid of radioactive materials this days (see all the Philips "radioactivity free" starters), and they are in way better position to get permits for all this stuff than a Ne shop



It is not dangerous. All you need is tiny amount to make 1 stray particle once in a while so the amount is tiny (the lamp may be big, but the electrodes are big too so the chance of particle to hit the electrode are high, so we are back to a tiny amount of total particles needed)

The high voltage make the Neon tubing to start without any other aid - it ignite simply on the "brute force" of the strong electrical field.

The indicator lamps could be designed to have the electrodes very close together, so the electrical field could be made strong too.

But still, the lower the striking voltage, the better, so nobody care, if some pieces exhibit unusually low striking voltages (the electrodes have some unusually sharp edge from not so proper machining,...)

the largest problem is with the glolwbottle starters: Their ignition voltage have to be quite accurately controlled: It should be always above the main lamp operating voltage even on an elevated temperature and bathing in a high energy light (otherwise it would false trigger, so flicker with the lamp) and on the other hand the mains should always trigger it in order to start the lamp (even at cold and dark winter evening).
Therefor it were more the glowbottle starters, where the radioactive aid was used - the rather high beta- or alpha particle energy was the dominant discharge trigger there, so it become rather insensitive to other mechanisms - so the trigger voltage become quite stable and well under control. But at the cost of managing the radioactive materials on the manufacturing floor...