Maker Pro
Maker Pro

luxeon emitters

A

andy

Jan 1, 1970
0
can you get away with running 3 luxeon emitters in series from a 12V
supply?

I know that 12V means 13-14V in practice, so in theory it should be
possible to set them up like:

12V --->|--->|--->|---[===]--- gnd

with the (variable) resistor set to whatever resistance was needed to get
the right current from whatever voltage drop was left over. But is this
likely to work OK in practice, or would it be better to use 2 in series
and a more reliable 4-5V voltage drop through the resistor?

Also, is using a variable resistor a good-enough way of dimming the
lights, or should I build a more complex circuit?

This is for lighting on a friend's canal boat.
 
B

Bob Monsen

Jan 1, 1970
0
can you get away with running 3 luxeon emitters in series from a 12V
supply?

I know that 12V means 13-14V in practice, so in theory it should be
possible to set them up like:

12V --->|--->|--->|---[===]--- gnd

with the (variable) resistor set to whatever resistance was needed to get
the right current from whatever voltage drop was left over. But is this
likely to work OK in practice, or would it be better to use 2 in series
and a more reliable 4-5V voltage drop through the resistor?

Also, is using a variable resistor a good-enough way of dimming the
lights, or should I build a more complex circuit?

This is for lighting on a friend's canal boat.

The three should work fine. However, using a simple pot means that all the
current will have to go through it. At 350mA, if your battery is 13.8V,
and the Vf is 3.42 (see http://tinyurl.com/8a9ut for datasheet) then the
pot will have to dissipate about 1.2W. Thus, one of those tiny radio
shack pots will just burn up. If you do this, make sure the one you buy
can take at least 2W.

One simple way around buying a pot like this would be to use a transistor
to set the current. One approach is to use a single NPN power
transistor (like one in a TO220 case) and use the pot to set the base
voltage. This has the disadvantage that the current (and thus the
brightness) will vary widely depending on the temperature. However, it may
be good enough. Note that the NPN will dissipate that 1.2W as above, so it
may need a heatsink, depending on location.

If more precision is required, a 1 ohm resistor between the NPN's emitter
and ground can be used to measure the current. This can be used as
negative feedback to an opamp, which then is used to control the base of
the NPN with its output. By setting the voltage on the positive input, you
can control the current with good precision, and the current won't change
with temperature. A pot can be used to set the non-inverting input of the
opamp. However, even more precision can be obtained by using a temperature
compensated voltage reference. This way, your current will no longer be
proportional to the input voltage.

However, it just occurred to me that you could use a low voltage TLV431,
which uses a 1.2V ref, and a 3.5 ohm sense resistor, like this:

-----------------o-------------- 12V
| |
| |
[led string] [330]
| |
c\| |
|-b---------o
e/| | K
| ---
o---------R-/ \ TLV431
| ---
[3R5 1W] | A
| |
----o------------o--------------- GND

When the current gets above 350mA, the TLV431 will turn on, pulling the
base of the NPN lower, thus decreasing the current. This gives you an
accurate, temp compensated current source which doesn't depend on the
input voltage. It only requires a drop of about 1.5V at 350mA, so it can
handle a voltage down to a bit lower than 12V with 3 luxeon emitters in
series.
 
A

andy

Jan 1, 1970
0
can you get away with running 3 luxeon emitters in series from a 12V
supply?

I know that 12V means 13-14V in practice, so in theory it should be
possible to set them up like:

12V --->|--->|--->|---[===]--- gnd

with the (variable) resistor set to whatever resistance was needed to get
the right current from whatever voltage drop was left over. But is this
likely to work OK in practice, or would it be better to use 2 in series
and a more reliable 4-5V voltage drop through the resistor?

Also, is using a variable resistor a good-enough way of dimming the
lights, or should I build a more complex circuit?

This is for lighting on a friend's canal boat.

The three should work fine. However, using a simple pot means that all the
current will have to go through it. At 350mA, if your battery is 13.8V,
and the Vf is 3.42 (see http://tinyurl.com/8a9ut for datasheet) then the
pot will have to dissipate about 1.2W. Thus, one of those tiny radio
shack pots will just burn up. If you do this, make sure the one you buy
can take at least 2W.

One simple way around buying a pot like this would be to use a transistor
to set the current. One approach is to use a single NPN power
transistor (like one in a TO220 case) and use the pot to set the base
voltage. This has the disadvantage that the current (and thus the
brightness) will vary widely depending on the temperature. However, it may
be good enough. Note that the NPN will dissipate that 1.2W as above, so it
may need a heatsink, depending on location.

If more precision is required, a 1 ohm resistor between the NPN's emitter
and ground can be used to measure the current. This can be used as
negative feedback to an opamp, which then is used to control the base of
the NPN with its output. By setting the voltage on the positive input, you
can control the current with good precision, and the current won't change
with temperature. A pot can be used to set the non-inverting input of the
opamp. However, even more precision can be obtained by using a temperature
compensated voltage reference. This way, your current will no longer be
proportional to the input voltage.

However, it just occurred to me that you could use a low voltage TLV431,
which uses a 1.2V ref, and a 3.5 ohm sense resistor, like this:

-----------------o-------------- 12V
| |
| |
[led string] [330]
| |
c\| |
|-b---------o
e/| | K
| ---
o---------R-/ \ TLV431
| ---
[3R5 1W] | A
| |
----o------------o--------------- GND

When the current gets above 350mA, the TLV431 will turn on, pulling the
base of the NPN lower, thus decreasing the current. This gives you an
accurate, temp compensated current source which doesn't depend on the
input voltage. It only requires a drop of about 1.5V at 350mA, so it can
handle a voltage down to a bit lower than 12V with 3 luxeon emitters in
series.

thanks - I'll try the transistor current source first, I think, and then
try your circuit if that doesn't work.
 
J

Jasen Betts

Jan 1, 1970
0
can you get away with running 3 luxeon emitters in series from a 12V
supply?

I know that 12V means 13-14V in practice, so in theory it should be
possible to set them up like:

12V --->|--->|--->|---[===]--- gnd

you didn't mention which luxeons you were planning to use

the nax forward volltagge is 4v so unless you intend using a switching
regulator two in series is the highest I'd go - otherwise use a boos
regulator and wire them all in series....
This is for lighting on a friend's canal boat.

from reading the specs it seems some of them may be effected
by humidity (they mention humidity next to the longevity chart)

Bye.
Jasen
 
J

Jasen Betts

Jan 1, 1970
0
On 2005-09-17 said:
can you get away with running 3 luxeon emitters in series from a 12
supply

I know that 12V means 13-14V in practice, so in theory it should b
possible to set them up like

12V --->|--->|--->|---[===]--- gn
you didn't mention which luxeons you were planning to us

the nax forward volltagge is 4v so unless you intend using a switchin
regulator two in series is the highest I'd go - otherwise use a boo
regulator and wire them all in series...
This is for lighting on a friend's canal boat
from reading the specs it seems some of them may be effecte
by humidity (they mention humidity next to the longevity chart

Bye
Jase
 
A

andy

Jan 1, 1970
0
can you get away with running 3 luxeon emitters in series from a 12V
supply?

I know that 12V means 13-14V in practice, so in theory it should be
possible to set them up like:

12V --->|--->|--->|---[===]--- gnd

you didn't mention which luxeons you were planning to use

luxeon star III, white, lambertian radiation pattern.
the max forward voltage is 4v so unless you intend using a switching
regulator two in series is the highest I'd go - otherwise use a boos
regulator and wire them all in series....

I've checked this since, and 4v is the max forward voltage, not the
typical forward voltage, which is more like 3.6v, so I think I can get
away with using 3 as long as the battery isn't run too low.

What I'm thinking of doing now is to make a dimmable current source with a
voltage reference, op amp, and transistor, and see if that works.
 
Top