I very much like the circuit by Israel Schleicher in EDN Jun 12 2003 for its
ability to drive a 3.6 volt Superbright LED off a 1.2 to 1.5 volt cell with
very minimal and inexpensive components. Since it pulses the LED, greater
perceived brightness can be acheived at great efficiency. See
http://www.reed-electronics.com/ednm...e=6%2F12%2F200
3&spacedesc=designideas

According to the formula, changes to R, L or C could thoretically provide
more current to the bright white LED - I am looking for 30ma average, but in
practice I have not been able to acheive this.

Can anyone tell me where I am going wrong? I would be happy to use three
cells if it helped.

--

On Wed, 16 Jun 2004 21:47:09 +1200, Simon Dyer wrote:

I very much like the circuit by Israel Schleicher in EDN Jun 12 2003 for its
ability to drive a 3.6 volt Superbright LED off a 1.2 to 1.5 volt cell with
very minimal and inexpensive components. Since it pulses the LED, greater
perceived brightness can be acheived at great efficiency. See
http://www.reed-electronics.com/ednm...e=6%2F12%2F200
3&spacedesc=designideas

According to the formula, changes to R, L or C could thoretically provide
more current to the bright white LED - I am looking for 30ma average, but in
practice I have not been able to acheive this.

Can anyone tell me where I am going wrong? I would be happy to use three
cells if it helped.

By believing that pulsing an LED produces greater perceived
brightness.

--
Best Regards,
Mike

Simon Dyer wrote:
I very much like the circuit by Israel Schleicher in EDN Jun 12 2003
for its ability to drive a 3.6 volt Superbright LED off a 1.2 to 1.5
volt cell with very minimal and inexpensive components. Since it
pulses the LED, greater perceived brightness can be acheived at great
efficiency. See

http://www.reed-electronics.com/ednm...e=6%2F12%2F200
3&spacedesc=designideas

According to the formula, changes to R, L or C could thoretically
provide more current to the bright white LED - I am looking for 30ma
average, but in practice I have not been able to acheive this.

Can anyone tell me where I am going wrong? I would be happy to use
three cells if it helped.

Simon,
if the LED current should be 30mA average, then it has to be with 20% duty
cycle(which is varying with battery voltage) 150mA. For this kind of current
the right transistor should be something like a BC337 with more current gain
at high currents. The inductor probably needs to be of bigger size and less
inductivity too and you can try to solve the equation given, which IMHO is
too simplified. The best would be to simulate the circuit with the free
LT-spice.
I think the first article gives a better circuit, in the pdf the connections
to the driver transistor are not drawn or got lost in some pdf-conversion,
but it is clear where they go.
--
ciao Ban
Bordighera, Italy

Simon Dyer wrote:
Since it pulses the LED, greater
perceived brightness can be acheived at great efficiency.

This is not true of the high brightness white LED types. The original
LEDs were more efficient when pulsed because the luminous efficiency
peaked at high currents, therefore constant current pulses at this level
produce an integrated light using less current than a DC current for
that same light level. This is not the case for the white LEDs. The
luminous efficiency peaks at low current levels like 5mA- so you
actually make the circuit LESS EFFICIENT by pulsing at high current
levels. The switching circuit referenced exists solely for the purpose
of voltage boost from low cell voltage battery technology to the higher
voltages required for the white LED.

Bonjorno
Thanks for the practical response Ban.
I'm not up with Spice models, but will try swapping out Q2.
Can you suggest values for the inductor?
Simon


"Ban" wrote in message
...
Simon Dyer wrote:
I very much like the circuit by Israel Schleicher in EDN Jun 12 2003
for its ability to drive a 3.6 volt Superbright LED off a 1.2 to 1.5
volt cell with very minimal and inexpensive components. Since it
pulses the LED, greater perceived brightness can be acheived at great
efficiency. See


http://www.reed-electronics.com/ednm...e=6%2F12%2F200
3&spacedesc=designideas

According to the formula, changes to R, L or C could thoretically
provide more current to the bright white LED - I am looking for 30ma
average, but in practice I have not been able to acheive this.

Can anyone tell me where I am going wrong? I would be happy to use
three cells if it helped.

Simon,
if the LED current should be 30mA average, then it has to be with 20% duty
cycle(which is varying with battery voltage) 150mA. For this kind of
current
the right transistor should be something like a BC337 with more current
gain
at high currents. The inductor probably needs to be of bigger size and
less
inductivity too and you can try to solve the equation given, which IMHO is
too simplified. The best would be to simulate the circuit with the free
LT-spice.
I think the first article gives a better circuit, in the pdf the
connections
to the driver transistor are not drawn or got lost in some pdf-conversion,
but it is clear where they go.
--
ciao Ban
Bordighera, Italy

Simon,

If all else fails, try this: 3 AA batteries in series with 10 Ohm resistor
(for current limiting) and 50 Ohm potentiometer (connect to 2 terminal only)
to LED. 30 ma corresponds to the potentiometer at about the center of its
range.

Tam