I have an LED light that I use for jogging in the dark. It has 3 modes; fast blink, slower blink and steady. If we eliminate fast blink (because I don’t know how fast it is), which setting will maximize battery life? If it were an incandescent bulb, I would assume that the steady light would use less power due to the inrush current. With LED’s, I have no idea.
I have used LED lights on my bike for years. The batteries last far longer in flash mode.
Chances are, that LED light is going to last damn near forever regardless. Relative to incandescent bulbs, they’re fantastically efficient.
My son has a $1.50 2xAA LED flashlight that I bought him at Wal-Mart. It puts out as much or more light than the incandescents of my youth, AND puts that light out for FAR longer on 2 AA batteries than the old incandescent ones did with 2 D batteries. It puts out light for like 48 hours or something absurd like that.
Blinking, at some speed, has been used to conserve battery power with LED indicators since they were first used. Since LEDs switch on and off at essentially instantaneous speeds - getting 50-100k/s cycling out of them for remotes and stuff is trivial even with cheap units - even the earliest designers learned that applying a 10% duty cycle (on 0.01s, off 0.09s, or even faster) would produce a perfectly bright indicator at just over 10% of the current demand of a steady LED.
Point being, LED brightness is almost always achieved by this kind of “duty cycling” or switching it on and off very rapidly according to a precise time relationship, rather than by trying to limit the current through it, as must be done with incandescent bulbs that switch on and off much more slowly (nearly 1 second for some).
So you can figure out the efficiency of your flashy-light by assuming that LED on = 100% power usage and the on/off ratio in time represents something very close to the same in power savings: on a quarter second, off a second means it’s using 1/5 the power.
And as bump already said, LEDs are hugely more power efficient than incandescent, so even the steady, bright beam lasts some large multiple of what an equivalent (and probably dimmer) grain-of-rice halogen bulb does. Use a dimmer marker LED, and flash it… and you could be talking hundreds of hours operation.
All true. But many applications have used the low power consumption of LEDs to radically downsize the batteries.
The OP’s light in incandescent form might run on 2 AA batteries whereas the LED version is equipped with a single CR3032 or such.
Bottom line: It’s IMO invalid to assume power conservation is never a user consideration in LED portable lighting.
I’m surprised the multi-LED lights that are common haven’t found a way to control brightness by turning off/blinking a varying number of LEDs. I know it’s going to add cost, but a single chip could take care of that for a negligible amount. With lightweight batteries we could have phenomenal live on a single charge for flashlights. It’s going to be a big change from the old days when a couple of D cells would last almost long enough with barely enough illumination to find a new pair of batteries.
I’ve watched LED flashlights evolve on what seems like a monthly basis, but we’ve already reached the point where real improvements (such as a dedicated chip that could control brightness and focus through switching) have been back-burnered against going super-cheap (down to free, at Harbor Freight) and super-ultra-techy (the new plethora of $50-200 “tactical LED flashlights” etc.)
LED is going to change everything about lighting in the next few years, but the dull, sensible stuff like flashlights that essentially never need new batteries are slow coming against the (par’n the pun) flashier stuff.
Even Radio Shack, though, had a simple, very elderly “flasher” chip - an LM3909? - dating from the early 1980s that would flash a bulb or LED with no fourth component. So I predict that a few generic “flashlight” chips are already around and will come into more and more common use, much the way nearly all alarm clocks use one of about three $1 control chips. But it’s a matter of profitability and consumer awareness/demand… and right now taclights that can blind someone a mile off seem to be reigning in the race.
That’s worth framing. 
My suspicion is that there are really 2 sorts of flashlight users- the occasional kind, who value long battery life (and the ability for the flashlight to function well after a long storage period), and who aren’t interested in lots of light, and the “moar lumens!!” crowd, who look for ever more absurd amounts of light, at the expense of battery life, battery availability, heat, practicality, etc…
Basically the first group is happy with the cheap LEDs that burn for 48 hours and cost $1.50, and the second group wants the one that puts out 800 lumens, even if only for 45 minutes. Most of the group that would fall in the middle is probably currently satisfied by one of the previous generations of the ultra-brights where it actually put out say… 250-300 lumens for 6 hours or something like that. If users can get that for say… 40 bucks, there’s not a lot of market for improving that a lot.
I’m firmly in group 3: I want it bright, but I want the batteries to last.
Now admittedly, the application here wasn’t strictly “flashlight”. It was a headlight for my bike.
For something like 7 years I commuted to work on a bicycle, and usually I traveled one or both ways in the dark. I found that most bike headlights are more about being seen than seeing.
At 12mph, you cover about 17 feet per second. If you light only shows 25 feet in front of you, you’ve got just over a second to react to that branch that has fallen across the bike path.
On the other hand, using your light for an hour each day will use up batteries real quick.
So, I’d use a couple blinky headlights that used AAA batteries. If I set them to flash, the batteries would last a week between chargings (because anybody using batteries that fast should be using rechargeable).
And to SEE, I had a light that needed its battery charged about every 2 days. Which was a pain, because the battery was connected to the light by a long cord, and I’d have to take it off the bike to connect it to the charger.
Back to the OP: in my experience, every LED light I’ve ever had said it would last FAR longer set to blinking than constantly on.
Well, you could always use the “on-the-bike-generator” approach. The one’s in my day were huge and clunky because they had to provide 2-D cells worth of power, but I bet you can find something sleeker now. (Apparently they have generators that work off the chain rather than off the wheels.)
As an aside, I passed a group of cyclists using LED lights the other day and they were very, very bright. Unfortunately, there doesn’t seem to be the same concept of aligning the lights so as not to blind oncoming motorists that you have with car headlights. So my vision was somewhat compromised when approaching the tail end of the group, which may have been counter productive.
As a fellow cyclist, I detest your choice of brightness. My commute is 9 miles (8 on various forms of bike paths) and these new bright lights that other cyclists are using are rude as hell to other cyclists. I can’t see anything but the blinding light for the 50 m up to you and then suffer blindness for another minute afterward. I’ve been bike commuting for well over a decade and I can tell you that within a week if you don’t know every pothole and hazard on your route and can’t avoid it, you are not paying enough attention. These ridiculously bright lights don’t improve your ability to see anything on the route they take away from it by blinding yourself. Try cycling with a lower power light and you will see how much better you can see 50-100 ft in front of you rather than just your small cone of blindness. In all of my years of cycling and all of my friends on the trails, I have never met anyone who ran into “a branch across your path”, but I know several who have gone over car hoods so lights are absolutely about being seen.
My pace is a bit higher than yours, ~25 mph, and I am just going on a rant as I have ended up off trail in the grass three times this week because of these brighter lights. But for the good of everyone else’s retinas, turn off the laser!
If you are saying an LED driven at 10% duty cycle looks as bright as 100% duty cycle (at the same drive current), that’s not correct. It looks much dimmer.
It’s easier yet to just blink the whole array at some duty cycle less than 100%. Fewer wires and it’s likely that the control circuit already has a means of blinking the whole array. No difference in efficiency.
Also, it’s likely that the LEDs in an array are at least partially in series so get the voltage closer to the battery input. Your control granularity would only be on a per-string basis.
I do have a headlamp with a very bright central light and a few smaller LEDs off to the side. You can select one or the other depending on the brightness you want. But there’s a very large factor here; 20:1 or more.
FWIW, brightness and long battery life can go hand in hand. If you want it to last really long, regardless of brightness, it’s best to pick LEDs based on their luminous efficacy (lumens / watt), even if you don’t want it ridiculously bright – because that same LED can be almost always be run at a lower brightness for a much longer time.
LEDs have gotten dramatically more efficient in the last few years.
If you want an absurdly long-lasting LED light, get one of the really efficient ones from a company like Zebralight or Nitecore, preferably with a lithium-ion battery, and then put it on a really low-power setting. They’ll go for daaaaays that way… for example, this one at 10 lumens (about the brightness of a candle) can go for 192 hours (8 days), and that’s just with one 18650 battery.
Both companies also make headlights that are more comfortable than having to hold a light in your hand.
I suspect the lights you are talking about are quite a bit brighter than anything I’ve used. And my lights were always aimed decidedly down. I’m not trying to light up signs a hundred feet away, I’m trying to see what’s 30 feet in front of me on the ground.
Maybe I need to describe the worst part of my ride for you to get the problem.
The bike path is right next to the road. There is a curb on one side and privacy fences on the other. Trees hang over the path for most of it. The road is 2 lanes in each direction with a grassy median. There are streetlights on the other side.
So there is enough light in the environment to ruin my night vision, but no light at all on the path. Blessedly, the direction I was riding in the dark is the direction of traffic, so I’m not being blinded by the lights of oncoming cars. That waits for my return trip in the morning.
Every windy day brings small branches down on the path, and people seem to love to throw beer bottles onto the path from passing cars. I can memorize the obstacles, but there are new ones every day.
One thing there are blessedly few of is other cyclists.
I have encountered some speedy folks on road bikes in the morning: they ride in the road and not on the path, because the “path” is more like an extra-wide sidewalk. It is NOT smooth. The cyclists I have encountered on the path generally have no lights at all. This includes a family of 4 who were traveling the same direction as me but slower, and a guy dressed all in black on a bike with no reflectors coming at me head-on.
Nobody should be out without a light at 10:30pm.
Almost all of the headlights I ever used cost $10 or less, and ran on AAA batteries. My first had an incandescent bulb and ate AA batteries like Pez.
But after the time my bike got run-over (I thankfully hopped off in time to not go under the car with it), I covered myself with bright yellow clothes, reflective strips, and all kinds of blinky lights. Being seen was no longer an issue. But I was taking 40 minutes to cover 8 miles not because I couldn’t ride faster, but because I couldn’t see.
I used to stare longingly at the $400 lights in the bike stores, but … I’d bought an $80 light once, and it was barely better than the $10 ones.
The big rechargeable I described in my first post was $60 on sale, and it’s battery was the size of a water bottle (and designed to go in the bottle cage). It lit up about 30 feet pretty well.
I’d think a cyclist would want some lights shining downward, to cover enough of the terrain ahead to react to holes, tire treads, dead bodies, and the like just ahead. Then a flashers facing four direction to be very visible to others. I don’t know if there’s a standard to meet, but the flashers should be red so they aren’t blinding.
For continuously variable brightness there are a lot of options for flashing and turning off some LEDs, but a practical problem with continuously variable anything is the higher cost of a control that will last. Some reasonable number of pre-set intensities with individual switches will cost less, last longer, and be more easily sealed to keep out the elements.
I agree. And that’s the problem with many bike headlights - they have wide symmetric beams, throwing as much light into the sky (or at oncoming drivers’ eyes) as onto the ground. Also, they throw a lot of light at the ground immediately in front (and causing glare), but doesn’t have much reach.
What you want is a headlight with a beam pattern like car headlights: some light downwards to illuminate the ground immediately in front, more intense beam towards the distance to illuminate the ground further ahead, and very little light elsewhere. German companies like Busch & Muller make lights like this, because Germany has strict rules on bike lights.