Maker Pro
Maker Pro

Novel LED regulator.

C

Clive Mitchell

Jan 1, 1970
0
I just got a cheap Chinese light that operates on three AA cells, an
external adapter or a USB feed. It has eight LEDs and I initially
assumed that it would use either a single resistor for a parallel LED
array or individual resistors per LED.

I popped the lid and noted that the eight white LEDs were in parallel,
so I popped the base and examined the circuit board expecting to see a
big resistor. Nope. It was a small surface mount device with three
pins and a tab, with a capacitor next to it. I noted the device number
and Googled it.

It's a 3.3v regulator with very low drop-out (from about 0.12v to 0.5v
depending on load), built in current limiting (0.6A) and thermal
shutdown.

As many of you will know, white LEDs are usually quite happy to draw
10-20mA at 3.3V so it seems a clever way to drive a parallel array.
Particularly given the generous 0.5W rating of the regulator package.

The regulator is an FS8854.
 
M

Mike Meakin

Jan 1, 1970
0
This is not an un-known method of driving leds and is used in the automotive
world for LED 'arrays' formed from series/parallel matrices. The benefit is
a sort of automatic current sharing in that as forward voltage of an
individual LED increases (due to temperature or initial current 'hogging')
it's current then reduces. It does however require that the Vf of the LEDs
are well matched and are preferably from the same batch. I have used this
technique and noticed that Led to Led brightness variations were noticable.
Some LED manufacturers suggest a small resitor in series with each LED in
the array.

This solution is 'neat' and allows operation of multiple white leds from 3
cells. The alternative (more costly) solution is a step up switcher and
running all LEDs in series (at a constant 'shared' current).

Yes please, keep taking things to pieces - to see how they work !
 
M

Mike Meakin

Jan 1, 1970
0
It has been pointed out to me that my description of the 'current sharing'
in LED arrays was somewhat poor - and I agree !

A more comprehensive (and accurate) description is provided in an 'Elektor'
article on this topic in the April 2003 issue entitled 'LED Arrays'. For
anyone interested I can email them a pdf copy.
 
C

Clive Mitchell

Jan 1, 1970
0
Mike Meakin said:
It has been pointed out to me that my description of the 'current
sharing' in LED arrays was somewhat poor - and I agree !

Given the modest variation in forward voltage with respect to current
that the gallium nitride LEDs exhibit, I guess that as long as close
matching forward voltages are chosen the LEDs will find their own level.
Not ideal, but cheap and easy for low cost goods..
 
Top