Word of warning AND a request for help.

[BirdofPrey]

About two months ago (three max) I bought three of the "First Alert" CFL night lights with the rocker switch that they sell at Target and Bed, Bath, and Beyond (surely other stores as well but these are the two places I've seen them).

I thought they were the answer to my prayers because the LED night lights I've bought up until then just simply did NOT put off enough light to do much good. Especially in my very dark and very steep (OLD house) stair well.

Anyway, these little CFLs (single u-shape tube) are rated at I believe 1.5 watts. They put out a fantastic amount of light. I used one in the room leading to my stairs, in the plug in along the stairs (lit all of them up quite well), and in the room at the top of the stairs which then leads into my bedroom.

However, about two weeks ago, the light in the stairs practically went out. I can't say if it dimmed over time or just immediatly dimmed but when my girlfriend went to go downstairs late one night she came back with the light in hand and told me it was so dim she couldn't see the steps. I looked at the light and at the base of one side it was slightly black with mostly gray. I plugged it back in there in the bedroom. It was bright (not as bright as it should have been) for a moment and then went dim with just a little light at the other base (opposite the discoloration). I chalked it up to a bad light and moved on.

Two nights ago, the one upstairs was all but out. I turned it off and back on. It got just as bright as when new, then about 5 seconds later did the same thing the other one did. These two were bought about a month apart.

I have one left now. It was bought at the same time as the last one that died on me. I've got a feeling it will go anytime now.

So, don't buy these lights. Apparently, as much as I liked them at first and as much as I THOUGHT they were great, they suck.

I figure a decent LED night light is my only recourse. Any suggestions?

No, I do not intend to make my own. I do not have the skill, knowledge, parts, or want of risk to my home in making my own. I just want to know if anyone has a suggestion of a QUALITY night light using LEDs that can come close to the output of the little CFL lights I'm talking about.

I'm tired of nearly killing myself late at night and early in the morning when coming down or going up my stairs.

 

[SirJMD]

1.5W should translate to around 90 lumens. A good 1W should be able to output the same.

A solution could be a cheap LED-christmas light. Its low cost and a good amount of light.

 

[Apollo Cree]

Use an incandescent. Electricity cost is minimal, light output is pretty good. Replacement bulbs are cheap.

I bought a "Lights of America" "40W replacement" LED replacement bulb. It was about bright enough to be used as a night light but that was about it.

I bought some LED night lights at Walmart that I was pretty happy with. Unfortunately, I don't have any of them where I am now. It MIGHT have been this one, but I'm not sure. http://www.lampsplus.com/Products/Set-of-Two-LED-White-Night-Light__62865.html

I think it was a GE light, about $4 each.

It dimmed itself during the day, and put out what I considerable reasonable night light level light.

 

[LEDninja]

I just use a warm white 2W cree bulb as my hallway light running 24/7. The light on the left is a 3 LED nitelite.

 

[Illum]

Sounds like the electronics overheated...
This sort of things happen unfortunately way too frequently...especially lamp makers who go the easy way out and use hefty resistors to clamp voltage as compared to a standard rectifier and a cap...along with other items that retain heat. Often there will be a thermal issue and the electronics just burn themselves out.

CFL's are awesome, but...like LEDs [well, LEDs are a little more tolerant] they require special drivers. I wish I can come up with my own small ballast, but thats beyond me at the moment. Unfortunately there are no single LED AC drivers that I am aware of, minimum is for "up to 3 LED" 350ma, which you could use just one LED but the cost/benefit would be less than optimal. :shrug:

 

[panicmechanic]

Unfortunately there are no single LED AC drivers that I am aware of, minimum is for "up to 3 LED" 350ma, which you could use just one LED but the cost/benefit would be less than optimal. :shrug:

If you can live with less-than-optimal efficiency, you can use this one.
There's also a DIY housing available. The driver fits the E27 socket nicely. The bulb assy is poor quality, so be prepared to do some filing, drilling, grinding and glueing to make a long lasting bulb out of it.
I also made one with this driver and a P7, but had to make a bigger mounting plate for better heat transfer to the fins.

 

[Illum]

If you can live with less-than-optimal efficiency, you can use this one.
There's also a DIY housing available. The driver fits the E27 socket nicely. The bulb assy is poor quality, so be prepared to do some filing, drilling, grinding and glueing to make a long lasting bulb out of it.
I also made one with this driver and a P7, but had to make a bigger mounting plate for better heat transfer to the fins.

the G10 looks familiar, grab a transformer with the secondary output of about 12Vac, slap on a rectifier bridge, a couple large caps and your good to go for 3LED drivers...the only issue I can see is the non-isolated nature and the hazard for shocks...

Perhaps I was too vague, I was not able to find any single LED AC-DC encapsulated drivers

 

[jtr1962]

I figure a decent LED night light is my only recourse. Any suggestions?

No, I do not intend to make my own. I do not have the skill, knowledge, parts, or want of risk to my home in making my own. I just want to know if anyone has a suggestion of a QUALITY night light using LEDs that can come close to the output of the little CFL lights I'm talking about.

I'm tired of nearly killing myself late at night and early in the morning when coming down or going up my stairs.

So far I haven't seen any decent commercial night lights. LED night lights are obviously the most promising as they don't burn out if made properly, only get dimmer with time ( letting the end user replace them before the light becomes too little to do the job ). Unfortunately, the problem with all commercial LED night lights I've tried is either poor design and/or using cheap LEDs which dim to uselessness in a matter of months.

In the end it may come down to either making one yourself, or enlisting the services of someone who has the capability plus parts. I've tried all kinds of night lights in my search for the perfect one. Fluorescent night lights give good output, and some last a reasonably long time. However, in most cases replacement lamps aren't available, or they cost as much as simply buying a new night light. In fact, in the majority of cases the lamps aren't even user replaceable. Incandescent night lights are pointless. The lamps don't last long ( 2 - 4 weeks in my experience ), aren't the best color for night vision, and since they fail completely and suddenly you can be left in the dark. That only leaves LED. Good in theory, but all of which I've tried only lasted a few months.

Those burnt out fluorescent night lights are probably amenable to modding with decent LEDs as I've done here. If you note, I mentioned using a metal oxide resistor for safety reasons in case the current-limiting capacitor fails. For my uses I consider that enough. If I did this for anyone else I'd probably put a very low current fuse in series with both prongs of the plug for added safety. In short, my mod is better protected than many commercial products, so no worries about burning your house down if you do it the same way ( or enlist someone's help to do it for you ). And by using quality LEDs, driving them at low current, and providing adequate filtering against spikes, I can be assured that I'll likely get 50,000 hours before these fade to 70% ( according to the LED's data sheet and my estimates of junction temperature ). I've already used the same LEDs to mod Target 3-LED night lights. After about 7 months continuous use ( ~5,000 hours ) they look as bright as the day I modded them. Bottom line-don't overlook the DIY route as it's often the only way to get what you want. It seems there's a lot of commercial products out there which are "almost good". With a little modding they can be made perfect. Here is another interesting night light mod I came across.

 

[JohnR66]

1.5W should translate to around 90 lumens. A good 1W should be able to output the same.

A solution could be a cheap LED-christmas light. Its low cost and a good amount of light.

I doubt it is anywhere near this output. Short tube florescent lamps are inefficient due to a short positive column. The casing probably doesn't help either. Lifetime may be limited due to high current crest from the cheap current limiting arrangment in these things. This damages the lamp electrodes fast.

 

[photonwave]

"Short tube florescent lamps are inefficient due to a short positive column."

But that could be solved by simply ditching the electrodes and going with micro-induction coils. That would also pump the tube life up by nearly double.

 

[fyrstormer]

I like the Christmas lights idea. Staple-gun them to the underside of the banister railing and call it a night.

 

[jtr1962]

Regarding output, I'd probably say you're lucky if the tube was putting out 60 lumens. A tube that small will not do any better than about 45 lm/W. Add in ballast losses and you're lucky to manage 40 lm/W. This isn't even accounting for fixture losses. The light the tube emits towards the wall is pretty much wasted. I'll guess 45-50 lumens actually made it out the front. Still much brighter than a regular incandescent night light bulb. A 7 watt bulb is about 28 lumens, a 4-watt probably half that.

The 3.5 watt fluoro night lights I modded had their output easily matched by ~75 lumens worth of LEDs using half the power ( and this with LEDs which are nothing steller efficiency-wise, testing at roughly 50 lm/W ). In fact, I'd say the LEDs are slightly brighter than the tube.

 

[photonwave]

"A tube that small will not do any better than about 45 lm/W."

What about Luxim's 250w plasma bulb the size of a tic-tac, pushing well near 150 lumens per watt?

What a tube will output, and how long it lasts, is pretty much entirely dependent upon its construction. Small electrode-based fluorescents might not output much but induction-based ones are pushing 110lm/w and higher.

 

[BirdofPrey]

BTW folks, this is the light I bought.

Stay far away from them. One of the three is still going strong. The other two are toast.

http://www.shophometrends.com/product.asp?pn=812008&bhcd2=1266869160

 

[jtr1962]

"A tube that small will not do any better than about 45 lm/W."

What about Luxim's 250w plasma bulb the size of a tic-tac, pushing well near 150 lumens per watt?

What a tube will output, and how long it lasts, is pretty much entirely dependent upon its construction. Small electrode-based fluorescents might not output much but induction-based ones are pushing 110lm/w and higher.

True but an induction-based tube is not going to find it's way into a low-cost night light. Also, I'm not talking here so much about size as power. Those high-efficiency induction lamps are much higher power than a 1.5 watt night light. As a general rule, with any lighting technology except LED, there are usually large efficiency gains with higher lamp wattages. I doubt you could even make a 3 watt induction lamp, much less make one efficient. For night lights the best technology at this point in time is LED. It's just a pity better LEDs aren't used.

 

[jtr1962]

BTW folks, this is the light I bought.

Stay far away from them. One of the three is still going strong. The other two are toast.

http://www.shophometrends.com/product.asp?pn=812008&bhcd2=1266869160

That actually looks pretty modable if you or someone near you who's handy wants to attempt an LED mod. The enclosure and switch seem to be decently constructed. It's a shame more attention wasn't paid to the electrical aspects.

 

[photonwave]

True but an induction-based tube is not going to find it's way into a low-cost night light. Also, I'm not talking here so much about size as power. Those high-efficiency induction lamps are much higher power than a 1.5 watt night light. As a general rule, with any lighting technology except LED, there are usually large efficiency gains with higher lamp wattages. I doubt you could even make a 3 watt induction lamp, much less make one efficient. For night lights the best technology at this point in time is LED. It's just a pity better LEDs aren't used.

What would you consider efficient? 100,000 hour bulb life is something I'd consider efficient in one way. Even beating a typical ~35lm/w halogens can do would be another form of efficiency. Guess it's application-dependent, in the end.

They've got 15w induction lamps for large flashlights that accommodate medium-base sockets. I don't see why they can't use a lower wattage, as all you need to do is adjust the voltage and frequency the induction coils generate and make the tube smaller. Also, smaller tubes generally create a brighter point-source emission at equivalent wattages (40w T-12 vs 40w T-8, T-8 is brighter) so the only real hurdle I see would be building a tube that small. We've got T2 sizes, ditch the electrodes, go induction, and you can push more power through them.

It seems way more like an engineering problem and we've already got the means to do it, there just has to be a push or incentive for a company or another person to do it. If they made two-inch long T2 tubes, I'd wind my own induction coils and ballast and try it myself, I've done it with regular T-12 tubes to make them VHO for the good old days of high school physics labs.

Don't get me wrong, I'm all for LED, but I think induction might have a potential edge if fleshed out a bit more. Miniaturization accomplishes wonderful things. Induction has been around for 120 years, and mainly it has remained an industry lighting solution. I feel it would be much better out in the general public.

 

[BirdofPrey]

That actually looks pretty modable if you or someone near you who's handy wants to attempt an LED mod. The enclosure and switch seem to be decently constructed. It's a shame more attention wasn't paid to the electrical aspects.

If anyone is up to the task, I'll send the bodies to someone if they can do it for a reasonable price.

 

[jtr1962]

What would you consider efficient? 100,000 hour bulb life is something I'd consider efficient in one way. Even beating a typical ~35lm/w halogens can do would be another form of efficiency. Guess it's application-dependent, in the end.

I'm all for longevity in any application. 100,000 hours is possible with either induction lamps or LEDs. In fact, LEDs can greatly exceed 100,000 hours if cost is no object. Just use more and severe underdrive them. Probably a million hours are possible but not practical. That's beyond the lifespan of most things they might be used in. I also suppose longevity is a form of efficiency in that things must be made less often if they last longer.

They've got 15w induction lamps for large flashlights that accommodate medium-base sockets. I don't see why they can't use a lower wattage, as all you need to do is adjust the voltage and frequency the induction coils generate and make the tube smaller. Also, smaller tubes generally create a brighter point-source emission at equivalent wattages (40w T-12 vs 40w T-8, T-8 is brighter) so the only real hurdle I see would be building a tube that small. We've got T2 sizes, ditch the electrodes, go induction, and you can push more power through them.

I'm not saying 3 or even 1 watt induction lamps aren't possible. Rather, I'm saying what's the point when they'll end up costing a lot more than an LED solution offering pretty much the same things ( high efficiency and similar longevity ). I'd like to see more use of induction lighting myself but in areas where it makes economic sense. If we can cost effectively make screw-base lamp replacements 2 times more efficient ( and 10-15 times longer lasting ) than CFLs using induction lighting then that's certainly a good thing. But for night lights, I'm just not seeing the point.

It seems way more like an engineering problem and we've already got the means to do it, there just has to be a push or incentive for a company or another person to do it. If they made two-inch long T2 tubes, I'd wind my own induction coils and ballast and try it myself, I've done it with regular T-12 tubes to make them VHO for the good old days of high school physics labs.

As an engineer who deals with a lot of real world problems, I can tell you the primary driver of whether a solution is viable or not is often cost, even if better solutions exist. If someone can cost effectively make a 100,000 hour induction night light then you can bet you'll see them on store shelves. But the key here is cost. $10 seems to be the limit consumers are willing to spend on a decent night light. I honestly couldn't make a profit on a decent LED night light selling at that price point, never mind an induction one requiring very complex driver circuitry. Look at the night light in question. In all likelihood it just has the tube, a current-limiting capacitor, and some means to start the tube ( an inductor most likely ). And that's about the most you can expect to see in a $10 night light. Maybe someone can engineer an induction night light which could sell for a profit at $40. Great except almost nobody would buy it. You might buy it, I might buy it. Maybe a few hundred CPFers might buy it. Hardly enough to make a production run worthwhile.

That's the reason why induction lighting has remained an industrial solution. I think it's an interesting technology, but also one which so far hasn't been able to be made at price points suitable for the general public. Perhaps that may change in the future, particularly if LED can't live up to the promises.

Lots of interesting technologies turn out not to be viable for residential use. For example, PID control and variable speed compressors would be a GREAT thing to have in room air conditioners. It would prevent the annoying compressor cycling and too hot/too cold issues. But it just can't be done at a price point consumers are willing to pay for a room AC. Now if more people knew the advantages of paying twice as much for something which is 5 times better then things might be different.

 

[jtr1962]

If anyone is up to the task, I'll send the bodies to someone if they can do it for a reasonable price.

PM sent.