How to determine how long a battery will last.

Hello,

I need a little help please. I know the basic formulas.. PIE etc. but I am not sure how to put them to use.

I have a 3 cell flashlight, 3 led's, that consumes from 172ma on the highest setting to 4.5 on the lowest. How long will 3 batteries 1.5 volts last on the highest and lowest settings.

How does the capacity of nicads, NiMH, alkaline figure in determining this value?

What is a typical alkaline AA capacity?

Thanks

Kem

AA Size:

Alkaline 2850 mAh (Duracell)
NiMh 1100-1400 mAh, depending on brand
NiCD 650 typical

If I'm not mistaken, you won't get the full rated capacity; the battery voltage will drop gradually over time to the point at which your light becomes useless. Technically, there's still power left, but it's too low of a voltage to power the LEDs. A lot depends on how low a voltage your circuit can accept and still work.

Here's a good analysis of different battery types:
http://home.att.net/~mikemelni1/battery.html

<FONT COLOR="#000000" SIZE="1" FACE="Verdana, Arial">This message has been edited by MikeB on 01-02-2001 at 09:36 PM</font>

Just as a rule of thumb I take the rated capacity of a battery and cut it in half then devide by the load. So say you had a led light pulling 100mA and you had 2 1000mA NiMH batteries. First the two batteries are in series so you still only have 1000mA. Then cut that in half, 500mA. Then devied by your load, 500/100=5 Or 5 hours of useable light.

Of course as the batteries get used the load will drop off quite a bit so in LED's they tend to last about 2/3's of the rated battery, or in the case above about 7.5 hours

Brock - http://www.uwgb.edu/nevermab/led1.htm

Thanks for great information. the analysis of different battery types was real interesting.

I guess the thing that is real puzzling is that with a meter on an Eternalight, with one light burning, I show about 5ma. I can crank down another light to 5 ma and it lasts about a week or so. The Eternalight is going on a month of 8-9 hours a night, plus intermittent, full power use, and it is still over 1.5 volts for each lithium cell.

I know that it is some kind of pulsing circuit...BUT the meter shows 4.5 ma. It must not be really drawing that much. I am stumped. I wish I could build one that would approximate the battery life of it.

The doc for the Eternalight says that with one lamp burning (position 12), (enough to read by ) will last 2,800 hours. I am almost ready to believe that. ???

Kem, Perhaps your meter is unable to keep up with the current usage as light flashes on and off many times a second, and is reporting more of a peak current rather than an average over time.

My Momma told me not to stare at the sun, but perhaps the answer to the Eternalight's battery life may be found by staring into its beam.

If you point the light at your face and wave it vigorously back and forth across your field of vision, you'll notice that the first one or two levels of dimming are achieved by actually dimming the LEDs -- the arc traced by the light on your retina will still be continuous. After that, you can see that the LEDs are pulsed on and off, leaving a trail of dashes in your eye. The dashes shorten in length relative to the gaps of darkness as you go down the dimness scale. The last few levels of dimness are achieved by turning off one, two, then three LEDs.

If we presume that the pulsed flashes are the same intensity as in the dimmest non-pulsed setting, then we could get an estimate of the current draw of a dimmed LED by measuring the total current draw in that setting, and dividing by four (the # of LEDs). If we can multiply that current by the proportion of time that the light is on in the lowest brightness setting (e.g. 0.2 if it's on 1/5 of the time), we'd have an estimate for the average current draw over time in the dimmest setting (pulsed, with a single LED).

One could estimate the proportion of time it's on by eyeballing it while waving it around (e.g. if it looks like the dashes are 4x the length of the dark parts), or perhaps waving it around while taking a time exposure with a camera.

Of course, this would still be just an estimate - it would be best to measure the current with an oscilliscope, I suppose.

<FONT COLOR="#000000" SIZE="1" FACE="Verdana, Arial">This message has been edited by mcjamison on 01-04-2001 at 11:55 AM</font>