# Running LED bulb on lower voltage?

Discussion in 'LEDs and Optoelectronics' started by labougie, Sep 5, 2015.

1. ### labougie

2
0
Sep 5, 2015
If I try running a 12v LED bulb from a 9v PP3 battery, will I get 75% of the rated lumens or nothing at all? Thanks in advance

2. ### Fish4FunSo long, and Thanks for all the Fish!

469
108
Aug 27, 2013
labougie, Welcome to the forum!

While the question is reasonable, and simple enough.....the answer is not quite as straight-forward....."traditional" bulbs (ie incandescent ) have a relatively fixed impedance comprised primarily of resistance hence the power consumption from any given bulb is primarily dictated by ohm's law E=iR --> P = iE ==> P = E^2/R ....So the power increases as a function of the voltage squared....Light output on the other hand is essentially an inverse square...that is in a traditional light bulb, the "brightness" increments roughly as the square of the power.....All this is to say that the resulting relationship between voltage and output is fairly linear......In a traditional incandescent bulb......However, LEDs are NOT linear devices...they are semi-conductors.....In the case of an LED, semiconductor means that the relationship between voltage and output are highly non-linear....Take a "10W LED Chip" as an example......(as luck would have it I am currently working on a design using exactly those, so happen to have test data on them)....Following is an abbreviated list of experimental readings taken from an actual Chip:

Code:
```@6.8V current = 55mA  power = 374mW......there is the first sign of a dim glow in one or two of the chip's segments....light output is just barely above zero
@7.2V all of the segments of the LED are dimly lit
@8.0V current = 83mA (power = 664mW) the entire chip is lit and is relatively bright with a
@8.2V current = 102mA (power = 837mW)  the LEDs are considerably brighter
@8.5V current = 154mA (power = 1.309W) and the chip is fairly bright....If you leave the voltage constant @ 8.5V for several minutes the current slowly climbs to 160mA (power = 1.36W!)
**Note: The climb in current occurs because the LED is getting warmer....the warmer it gets, the more current will flow at any given voltage (more on that later) ....
@9.0V current = 260mA (power = 2.34W)  the LED is perhaps 30% of "full power brightness"

NOTE: holding the voltage @ 9V the current fairly quickly rises to the 270mA level as the LED heats up...
from 9V onward I will report both the initial current and then the current after some specified time period(s)

@9.0V current = 270mA (power = 2.43W)  time = 1 minute
@9.5V current = 375mA (power = 3.56W) time = initial         NOTE:  the LED is roughly 70% of "full brightness" and VERY WARM to the touch on a moderately sized heat sink...
@9.5V current = 430mA (power = 4.09W) time = 1 minute
NOTE: At this point Forced Air Cooling is Introduced
@9.5V current = 390mA (power = 3.71W) time = 30 seconds after forced air cooling begun
NOTE: It is important to note that the "brightness" is not perceptively effected by the change in temperature...just the power consumption.

@10.0V current = 550mA (power = 5.5W)  time = initial .. roughly 80% of full brightness
NOTE: removed the forced air cooling
@10.0V current = 575mA (power = 5.75W) time = 15 Seconds after forced air cooling removed
@10.0V current = 600mA (power = 6.00W) time = 1 minute NOTE: the heat sink is HOT to the touch and the current continues to rise
@10.0V current = 620mA (power = 6.20W) time = 3 minutes NOTE: approaching equilibrium
NOTE: replaced the forced air cooling,
@10.0V current = 550mA (power = 5.50W) NOTE: no noticeable decrease in light output

@10.5V current = 650mA (power = 6.83W)
@10.5V current = 700mA (power = 7.35W) time = 1 minute NOTE: forced air cooling still on.....brightness is roughly 90% of the brightness @ full rated power
@10.5V current = 725mA (power = 7.60W) time = 3 minutes
NOTE: Removed Forced Air Cooling
@10.5V current = 750mA (power = 7.88W) time = 15 Seconds
@10.5V current = 780mA (power = 8.20W) time = 1 minute
@10.5V current = 800mA (power = 8.40W) time = 3 minutes   NOTE: Heat Sink VERY HOT
NOTE: Forced Air cooling resumed
@10.5V current = 725mA (power = 7.60W) time = 1 minute

@11.0V current = 920mA (power = 10.12W) ***Rated Power***  NOTE: only a very slight increase in brightness from the 10.5V / 725mA

```
I should point out that identical chips taken from the same batch have some variation in the Voltage/Current .... the extreme variation in current for the 10.0V point at similar temperatures was 85mA with the majority of the chips measuring 550mA +/- 20mA...the highest was 595mA and the lowest was 510mA...there were a total of 10 chips tested....

So, to answer your question directly, NO, you cannot expect a linear light output based on a "fixed voltage", NOR should you expect an LED rated @ 12V to actually be able to handle 12V....however, in the specific case cited above it turns out that 9.0V was actually roughly 30% of full brightness light output and that by 9.5V the brightness was ~70% of full brightness...since typical batteries are nominally rated (voltage @ some specific current) their Actual measured no-load voltage is typically higher than the reference voltage specified....for instance a 12V car battery might have a measured voltage of 12.8V even though its nominal rating is 12V...to charge a 12V car battery requires ~14V, so while you might think of your car having a 12V electrical system, the typical measured voltage of a car's voltage is 14.4V...all of which is to say that just because you have a battery with a nominal voltage of 9V does not mean that the measured voltage might not be 9.5V to 10V depending on they type of battery.....

In general Hi-Power LEDs need to be driven by specially designed Constant Current drivers...the amount of light output is primarily a function of current and temperature...the voltage is constantly adjusted....

Good Luck!

Fish

3. ### Old Steve

734
169
Jul 23, 2015
Interesting data, Fish. I've never played with any of the modern LEDs, only the single ~1.8V to 2V Vf LEDs.
That was pretty hard to read in a code window, hope you don't mind that I reproduced it in 'quote' brackets for readability.

4. ### cjdelphi

1,096
104
Oct 26, 2011
I've had an RGB 10w attiny85 running for 2 years now using just 9v sourced to each rgb via simple npn , the current is 1/3rd of the power it can do...

As winter turns to summer the ambient temp increase causes a lower foward voltage drop...

No problem as current rises accordingly light output goes up but no where near it should it relies on no resistance between collector / emitter and 1 - 10k on it's base (the only thing limiting the current)

It's worked a charm, downside is i'm not using it's full potential and balast the circuit properly

5. ### labougie

2
0
Sep 5, 2015
Wow! What a comprehensive reply from Fish4Fun, for which many thanks.

The question comes from my need to set up navigation lights for my 12 ft open dinghy (Upper Thames usage) without having (and having to charge) a lead acid battery. The nav light housing I have takes a 31mm festoon bulb and these are available as 1w LED replacement car number plate or interior bulbs, hence my query about running them from a PP3. However from what's been said, it doesn't seem as though it's a workable idea - perhaps I'll have to go with the lead acid 12 sealed battery.

Thanks for the input.

6. ### Fish4FunSo long, and Thanks for all the Fish!

469
108
Aug 27, 2013
lablougie,

Ahhhhh....Those bulbs have a "driver" built into them (at least the ones I have seen do) and might or might not operate @ 9V.....regardless, I would suggest 1 of 3 things: 1) simply buy one of those 8-cell AA battery holders....the nominal voltage will be 8 * 1.5V = 12V.....a single LED bulb like the one you mentioned should operate for > 12 hours on a single set of batteries....2) Get a "3S" LiOn Battery built for the RC Hobby industry + a charger.....depending on the mAh rating you could likely run the bulb for several hours to over a week on a single charge....3) Simply buy a stern light designed to operate off 2 "D" cell batteries...they sell for ~ \$20 here in the states...meet all USCG regulations and operate for ~12 hours on a pair of batteries.

Good Luck!

Fish