AC Soft Start Incandescent Lamp switch (*FINAL EXPERIMENTAL RESULTS*)

[hgg]

You should also increase the on time.  Right now the protected bulb spends 30% of its on time at reduced wattage which is not isolating the soft start mechanism.

I also believe you meant to say lux rather than lumens.   Lumens is not a distant dependent unit.

The turn on time is actually around 30 seconds or less.  It was like a capacitor charging curve.
90s passed when I could not measure any further increase in brightness, so its about 10%
at reduced wattage.  But yes, a longer time would be better.    I indeed meant to say Lux...
I've corrected it.  Thanks.

I will increase the on time to 10 minutes and I will also add a resistor to the unprotected bulb
equal to the hot resistance of the thermistors.  Thank you for your feedback and suggestions!

[SeanB]

True a resistor will drop voltage a little, and drop inrush surge a lot. The 15 minute cycle is close to what is used to stress test lamps, though most manufacturers use a 15min off 2 hours on cycle for life testing purposes, and test a batch of 100 or so each time, stopping at around 50% failure rate for lifetime. Longest lasting lamps will do the 1000 hour rating with under 10% failure, but the cheap ones will have 50% at this point, with probably 10% failing in the first 100 hours from poor manufacturing.

I bought ( not by choice) a batch of Chinese made downlighter MR16 lamps, and they are very poor, having a running life of less than rated ( 2000H) by far, most only lasting a month in service. Ironically other lamps on the same transformer, next to them are probably 2 years in the same service, and are still working perfectly, only the odd random failure. Some have done 5 years or more. The difference between Osram, Phillips as the good ones, and GE and Radiant as the poor ones, in order of lifetime. The supplier blames the fitting for the poor lamp life.....

[hgg]

Now that you mention it I don't really want to use a resistor on the unprotected bulb on start up.
I would have to use a relay and connect the equivalent resistor after the unprotected lamp has
been turned on because I don't want to protect it from inrush current.  I just want it to have the
same dimming with the protected one just in case that even such a small dimming of the protected
bulb increases its life by a large amount.  Alternatively I can use the same relay to disconnect the
thermistors after 30 seconds so that both lamps would glow with exactly the same brightness.

I wish I had your Chinese lamps...   I need them to fail fast so that the experiment will not take so
long to show any results.  The only ones I've found were these Osram 40W bulbs that were failing
faster than others I've used.

[johnwa]

I wish I had your Chinese lamps...   I need them to fail fast so that the experiment will not take so
long to show any results.  The only ones I've found were these Osram 40W bulbs that were failing
faster than others I've used.

Perhaps you could just crank the voltage up a bit, to encourage them to fail a bit sooner?
If you don't have a variac, get a 15V or so transformer, and connect its primary to the supply, and the secondary in series with the lamps. (make sure you get the phasing right).

You would need to oversize the transformer a fair bit though, to make sure its impedance doesn't influence the startup surge.

[hgg]

-- FINAL UPDATE.

Hello everybody,

This has been a very very very long experiment.  Almost boring..  But,
the protected bulb has already outlasted 4 other bulbs from the same batch!
The only sign of deterioration of the protected bulb is a slight lower brightness
than the new bulb.  Barely noticeable though.

And this time the circuit was upgraded to switch off the bulb protection after 30 seconds,
in order for both lamps to work with the same current and intensity.  I have already
installed this very small and simple addition to every lamp in my house.  The only thing you
have to be careful about is to carefully size the thermistors for the specific bulb and  that the
thermistors do not touch anything because you might start a fire...




I will wait for the fifth lamp just to confirm the NASA claims and then I will terminate the experiment.
Clearly filament evaporation is not the main factor for the life of the incandescent bulbs.
The main factor is extreme thermal stress from the initial current surge.

George.

[Zero999]

Why should I replace inexpensive incandescent bulbs with expensive CFLs?

Because you'll save more money on your energy bill in the long run?

I have not yet met a fluorescent lamp with a pleasant colour. (Also dimmed incans have a warm yellow glow similar to firelight.)

Try some of the modern LED lamps which can be found with very low colour temperatures, down to 2100k which mimics candle light very effectively.

http://www.ledplanet.co.uk/52243.html


Running incandescents at much lower than their rated power is even more inefficient because the spectrum shifts even more to the infrared, meaning even less of the energy is converted to visible light than it was run at full power. The temperature of the filament in an incandescent is such that it gives optimum cost per lumen over the lamps life. The lamp may last longer at a lower temperature but it's a false economy, since it results in higher energy bills.

[SeanB]

Funny enough yesterday I changed a failed Phillips R63 lamp, made in Poland. It is the outlier, having lasted at least 3 years in operation 8 hours a day 5 days a week 50 weeks a year as room lighting in a downlighter. I know it is 3 years old as I got the current box of 144 R80 lamps 3 years ago when I ran out of R63 lamps. 60W lamps all, and they do last at least more than double the rated 2000 hours.

Some of the 20W 50mm 12V downlighters also did more than a decade in service, as they were still open lamps, while I changed to fully enclosed lamps a decade ago, after a cheap needed it now Eurolux blew up in a shower of hot glass. The old made in EU or German lamps were well made, lasting longer than the Cheaper USA and Mexican made lamps, and the Chinese ones did not even make 200 hours.

[eetech00]

Hi

The easiest way to make an incandescent bulb last longer is....well...by never turning it off.

The idea is to keep the filament warm by allowing a small amount current to flow in the filament
when not in use. When the bulb is off, the filament is cold, its resistance is lower, so there is a greater amount of inrush current and thermal stress, when powered on. If the filiament is kept warm, resistance is higher and there will be less thermal stress on the filament when full power is applied.

eT

[Rufus]

No point trying to argue with hgg, but, for anyone else reading this thread just know it is a stupid idea, stupid experiment, and stupid conclusion drawn from the results.

[hgg]

@eetech00 this is a very good idea actually, but then it gets complicated.

Hero999:

Running incandescents at much lower than their rated power is even more inefficient because the spectrum shifts even more to the infrared, meaning even less of the energy is converted to visible light than it was run at full power

Indeed, but in my experiment the bulbs were running at full power after the 30 second protection
interval from the thermistors.  The protected bulb was completely disconnected from the the
thermistor and connected directly to mains power.

Rufus:

No point trying to argue with hgg, but, for anyone else reading this thread just know it is a stupid idea, stupid experiment, and stupid conclusion drawn from the results.

Rufus I don't know what arguments you have against this experiment but have a look at the following:

- http://en.wikipedia.org/wiki/Phoebus_cartel
- http://spectrum.ieee.org/geek-life/history/the-great-lightbulb-conspiracy

I don't think that this can be called "ethical science".

George.

[DavidDLC]

No point trying to argue with hgg, but, for anyone else reading this thread just know it is a stupid idea, stupid experiment, and stupid conclusion drawn from the results.

Then why are you in this threat and participating, just get out of here

[IanB]

Hero999:

Running incandescents at much lower than their rated power is even more inefficient because the spectrum shifts even more to the infrared, meaning even less of the energy is converted to visible light than it was run at full power

Indeed, but in my experiment the bulbs were running at full power after the 30 second protection
interval from the thermistors.  The protected bulb was completely disconnected from the the
thermistor and connected directly to mains power.

That comment about running bulbs dimmed was in response to me. However, if I take a given bulb and run it at half power it will cost half as much to run compared to running it at full power. Since the running cost is reduced the loss in efficiency is not relevant to the calculation.

[Rufus]

That comment about running bulbs dimmed was in response to me. However, if I take a given bulb and run it at half power it will cost half as much to run compared to running it at full power. Since the running cost is reduced the loss in efficiency is not relevant to the calculation.

If you run a bulb at zero power it costs nothing at all to run and you don't even have to buy a bulb.

[Zero999]

Running incandescents at much lower than their rated power is even more inefficient because the spectrum shifts even more to the infrared, meaning even less of the energy is converted to visible light than it was run at full power. The temperature of the filament in an incandescent is such that it gives optimum cost per lumen over the lamps life. The lamp may last longer at a lower temperature but it's a false economy, since it results in higher energy bills.

That comment about running bulbs dimmed was in response to me. However, if I take a given bulb and run it at half power it will cost half as much to run compared to running it at full power. Since the running cost is reduced the loss in efficiency is not relevant to the calculation.

The loss of efficiency is very relevant to the calculation because it would be much more efficient to just use a lower wattage bulb, than overrate the bulb and under power it. If you take a 60W bulb and run it at half the power, it won't produce half the light but a less than a quarter, so it would be much cheaper in the long run to just swap it for a 15W bulb run at full power. I understand you dislike white light and prefer yellow which is why I recommended a low colour temperature LED lamp, which uses a tiny fraction of the power: the energy savings pay of the high lamp cost fairly quickly.

[IanB]

But a 15 W bulb also loses the convenience of dimming. With a dimmed bulb I can continuously vary the light according to needs. I can make it brighter for a short while if I want good lighting for a task, and I can make it dimmer to watch television or a movie.

Maybe with a new living space I could design a mix of lighting and controls on separate circuits to do what I wish, but it is not practical to retrofit that to an existing space. Even with a new space it is expensive to change what the builders have previously installed (which is likely a low cost minimum).

This is not even to consider the false claims about efficiency circulated in Europe. The most need for electric lighting occurs during the winter months with long dark evenings (it gets dark at 4pm in the UK). This is also the time when houses have the central heating turned on. Bulbs operated in centrally heated houses are 100% efficient. There is no energy saving to be made by replacing them.

It's quite simple. Home lighting makes almost no contribution to energy demand and global warming and the drive to new lighting technologies like CFL and LED is driven purely by the desire of manufacturers and marketers to have new expensive products to sell (and a captive market forced to buy them). As always, when you see a drive for change, follow the money.

[edavid]

This is not even to consider the false claims about efficiency circulated in Europe. The most need for electric lighting occurs during the winter months with long dark evenings (it gets dark at 4pm in the UK). This is also the time when houses have the central heating turned on. Bulbs operated in centrally heated houses are 100% efficient. There is no energy saving to be made by replacing them.

Do people in Europe really use primarily resistive heating?  If so, you're right, that should be fixed first before worrying about lighting.

It's quite simple. Home lighting makes almost no contribution to energy demand and global warming and the drive to new lighting technologies like CFL and LED is driven purely by the desire of manufacturers and marketers to have new expensive products to sell (and a captive market forced to buy them). As always, when you see a drive for change, follow the money.

http://www.eia.gov/tools/faqs/faq.cfm?id=99&t=3

"EIA estimates that in 2012, ... residential lighting consumption was about 186 billion kWh or about 14% of all residential electricity consumption."

Is 14% the same as "almost no contribution"?

[Zero999]

But a 15 W bulb also loses the convenience of dimming. With a dimmed bulb I can continuously vary the light according to needs. I can make it brighter for a short while if I want good lighting for a task, and I can make it dimmer to watch television or a movie.

Maybe with a new living space I could design a mix of lighting and controls on separate circuits to do what I wish, but it is not practical to retrofit that to an existing space. Even with a new space it is expensive to change what the builders have previously installed (which is likely a low cost minimum).

There are already plenty of solutions to that, such as smart lighting, which enables you to control but the colour and brightness. It is expensive but very easy to retrofit, often just as easy as changing a light bulb and it's still cheaper to run.

This is not even to consider the false claims about efficiency circulated in Europe. The most need for electric lighting occurs during the winter months with long dark evenings (it gets dark at 4pm in the UK).

What about the four or so hours a night when the lights are on in the summer when the extra heat is a nuisance and will increase bills further if you have air conditioning?

This is also the time when houses have the central heating turned on. Bulbs operated in centrally heated houses are 100% efficient. There is no energy saving to be made by replacing them.

That's only the case if your home is heated with electricity which is a very expensive and inefficient form of heating. If the home is heated by natural gas, as it is in most cases in the UK, it's much more efficient to use electricity for lighting.

[PauloTiti]

I have a device that fits into the light socket (B22) and sits between the socket and the bulb that does a soft start.
It is only for a single bulb though. Now that incandescent bulbs are getting less common I don't know how readily available they are. Ebay the source of all things cheap adapter related turned up nothing.

You now have the russian soft starter device mentionned in the first post available on Ebay.

[hgg]

Hi, do you have a link?

[PauloTiti]

Hi, here is the link I found: http://www.ebay.fr/itm/321733084301

[hgg]

Thank you for the link.

From the description here : http://provod-nik.ru/p56333418-blok-zaschity-gal.html it looks like that
it uses an integrated circuit that provides the soft start.  It does not use the simple solution of the
cheap thermistor.  I think there are quite a lot of circuits on the market like that.  I though you
found one like the original.

This is the original product : http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20070018922.pdf

George.

[PauloTiti]

I didn't know about this one. But looking at some litterature they say you have to wait about 45 seconds to get the light on... Jesus ! You have to be patient ! 

[PostalPreacher]

Forgive my jumping in here at such a late date. I just purchased a refrigerator for my wife. I got a great price, but it was missing one feature of the higher priced models. The deluxe models have interior lights inside that ramp up over about a second or two. I'd like to be able to have my lights do the same. I'm not concerned about life of the bulb or anything else...I just want to make it do that for my wife. Will a simple (properly chosen) thermistor in line work? The bulbs are pretty standard (appliance) 40W 120V bulbs.

[edavid]

Forgive my jumping in here at such a late date. I just purchased a refrigerator for my wife. I got a great price, but it was missing one feature of the higher priced models. The deluxe models have interior lights inside that ramp up over about a second or two. I'd like to be able to have my lights do the same. I'm not concerned about life of the bulb or anything else...I just want to make it do that for my wife. Will a simple (properly chosen) thermistor in line work? The bulbs are pretty standard (appliance) 40W 120V bulbs.

With all respect, that is a stupid feature... why would you want to have to wait around to see what's in the refrigerator?

If I had a refrigerator like that, I would hate it.

[Audioguru]

I have been using White 3300k and 3500k CFL light bulbs for maybe 5 to 10 years. They are not expensive, about the same per year as hot, power wasting incandescents and I save time because they last such a long time that I do not need to replace them so often. Daylight colored bubs produce Blue light, not white.

My city replaced all the incandescent traffic lights with LED ones and all streetlights are also LEDs with a modem and a little antenna on top.

There was (or maybe it is still operating) an incandescent light bulb in Alaska that has been turned on only one time and it has been shining for well over 100 years. It lasts a long time because it has never been turned off and it has missed many on-surges.