Maker Pro
Maker Pro

LED Parrellel Current limiting Cap. Question........

B

Baronvonrex420

Jan 1, 1970
0
Let me start by "Yes, I know I'm supposed to have a current limiting resister
to each LED", but I don't. This project requires over 100 LED's and I need to
reduce drilling where I can.

I need to drive 5 and 6 3VDC, 20 ma LEDS in par. but I can't find a formula
for calculating the resister value. The life expectancy of this product is
less than 100 hours and current will be supplied with a 7812 regulated 12 VDC
current from a car battery.

I tried 33 ohm 1/2 watt, 100 ohm 1/4 watt, and tonight I tried a 10 ohm 1/2
watt resister. The LEDS work great but the resisters are hotter than heck!

What do I need to do to bring the resister temps down or calculate the actual
value I need? I need to encase this in resin and I can't afford the heat.

Any help would be appreciated! TIA!

Newbie
 
S

Sir Charles W. Shults III

Jan 1, 1970
0
Baronvonrex420 said:
Let me start by "Yes, I know I'm supposed to have a current limiting resister
to each LED", but I don't. This project requires over 100 LED's and I need to
reduce drilling where I can.

I need to drive 5 and 6 3VDC, 20 ma LEDS in par. but I can't find a formula
for calculating the resister value. The life expectancy of this product is
less than 100 hours and current will be supplied with a 7812 regulated 12 VDC
current from a car battery.

I tried 33 ohm 1/2 watt, 100 ohm 1/4 watt, and tonight I tried a 10 ohm 1/2
watt resister. The LEDS work great but the resisters are hotter than heck!

What do I need to do to bring the resister temps down or calculate the actual
value I need? I need to encase this in resin and I can't afford the heat.

Any help would be appreciated! TIA!

Newbie

Okay, let's take the case of 6 LEDs. I don't know why you MUST put them in
parallel, which is particularly wasteful of power (meaning lots of heat made for
nothing), but let me suggest a better strategy.
Since your supply is 12 volts, and you have LEDs that run at 3 volts, why
not put three LEDs in series with a single resistor, and do this twice? Now,
you greatly simplify the picture and you don't drop 9 volts in the resistors.
Here is how you calculate your resistance and the power dissipation.
Three LEDs in series will drop 9 volts, leaving 3 volts to drop in the
resistor. You know that in series, the current consumption remains unchanged,
so 20 mA is the figure we use. Since R = E/I, we divide 3 volts by .02 amps and
get 150 ohms. Now, since P = EI, the resistor will have to dissipate (3 volts x
..02 amps) watts as heat. This is 0.06 watts. Almost any small resistor can
handle this with no trouble, even the little 1/8 watt ones.
You will put two of these assemblies in parallel, meaning that you will have
6 LEDs and 2 resistors for one of these assemblies. Total power turned to heat
is a mere 0.12 watts per pair of LED strings.
With your 100 LEDs on at once, if you were running all of them in parallel
and dropping 9 volts through your resistor, you would have been dissipating 18
watts as heat if and only if your resistor was (9 volts / 2 amps) or 4.5 ohms.
But it is very likely that you would also have had a few dim ones, and not only
that, 3/4 of the power from your battery would be heating up a resistor instead
of supplying illumination.
If you wanted to string them in threes (as I mentioned above), then fully
3/4 of the power would go to lighting and only 1/4 would be wasted.

Cheers!

Chip Shults
My robotics, space and CGI web page - http://home.cfl.rr.com/aichip
 
J

Jonathan Kirwan

Jan 1, 1970
0
Let me start by "Yes, I know I'm supposed to have a current limiting resister
to each LED", but I don't. This project requires over 100 LED's and I need to
reduce drilling where I can.

I need to drive 5 and 6 3VDC, 20 ma LEDS in par. but I can't find a formula
for calculating the resister value. The life expectancy of this product is
less than 100 hours and current will be supplied with a 7812 regulated 12 VDC
current from a car battery.

I tried 33 ohm 1/2 watt, 100 ohm 1/4 watt, and tonight I tried a 10 ohm 1/2
watt resister. The LEDS work great but the resisters are hotter than heck!

What do I need to do to bring the resister temps down or calculate the actual
value I need? I need to encase this in resin and I can't afford the heat.

Any help would be appreciated! TIA!

Probably your best first bet is to not use a 7812. You are just
asking for dissipation trouble. 100*20mA is 2A and you have to
get rid of (12V-3V)*2A = 18 Watts one way or another. Pulse
width modulation probably wouldn't be a lot of use here, either.

Instead, use a switcher and set it to output your 3V rail. 2A
is quite reasonable. You won't be able to avoid the LED
dissipations of 6 watts or more, but your switcher will
definitely help out on the hot resistors. Can you live with
brightness variations?

Jon
 
B

Baronvonrex420

Jan 1, 1970
0
Thanks Sir Shults!
That makes a lot of sense and hopefully keeps the ambient temp down. I'll look
at a re-design and see how I can incorporate this into my design...

My company web site might be of interest to you, Robotic Technology Systems:

http://www.rts-group.com/

Thanks Again!
 
R

Robert Monsen

Jan 1, 1970
0
LEDs really don't have a simple linear resistance. They have a voltage drop
instead. Given the v drop of each LED, you can approximate the resistance
needed by (12-N*Vled) /.020, where N is the number of LEDs in each parallel
branch. This is assuming a 12V voltage across the branch.

You may also want to consider that LEDs can be flashed very quickly without
appearing to flicker, with a resulting large savings in power. I just hacked
up a circuit using a 555 and a 4017, both of which will run with Vcc=12V.
The 555 drives the clock of the 4017. You can use this to drive 10 branches,
with a correspondingly large savings of power. I'm not sure of your
application, so I don't really know if it will work for you, but I am
driving 4 red LEDs from a single branch, and they all light up nicely. If
the LEDs aren't bright enough with 1/10 of the time, you can use one of the
outputs of the 4017 to drive the reset, which will limit the number of
branches, and so give each branch a bigger 'duty cycle'.

One final note, most voltage regulators won't work well unless you have an
input voltage that is 2 or 3 volts above the output voltage. You may want to
run your system at 9 volts using a 7809, rather than 12 volts. If 12 volts
is absolutely required, there are low dropout voltage regulators that can
run with a smaller input/output ratio.

Regards
Bob Monsen
 
Top