All of the designs so fr have used a rectangular grid. Since the output is
circular in shape, why not take a clue form the honey bees and set up the
LEDS in a hexagonal grid? This should result in more even coverage.
Myc
On Fri, Feb 15, 2008 at 3:48 AM, agscal -AGSCalabrese <
agscal@...>
wrote:
> Even LEDs from the same batch cannot be trusted to
> have the same brightness for the same current. After
> they age, it is even less likely that they will match. If you
> want them to stay matched you would have to have some
> kind of feedback or periodic re-calibration of each LED.
> I think that is probably overkill.
>
> I recommend that you use a current limiting resistor and
> put 10 or 20 or more in a string. This guarantees that every LED
> in that string will get the same current.
>
> For example, if you had 20 LEDs and each one had 1.7V
> forward voltage drop, you would have a total forward voltage
> drop of 34 volts. If you use a 36 VDC supply you can tweak the
> resistance
> to get the current you want. To get 20ma you would divide 2 volts
> by .02 A to get a resistance of 100 ohms. The 2 volts comes from
> subtracting 34V from 36V.
>
> This approach can be modified to match whatever power supply
> you have laying around to the LEDs you have.
>
> Gus
>
> On Feb 14, 2008, at 10:19 PM, javaguy11111 wrote:
>
> > I went ahead and placed an order for some as well.
> >
> > I am wondering if just doing simple current limiting resistors is
> > sufficient or if a proper LED driver chip would be the better way to
> > go. From what I have read, at least for visible LEDs, you want to
> > match the current in the LEDs to ensure that they all shine evenly.
>
> [Non-text portions of this message have been removed]
>
>
>
[Non-text portions of this message have been removed]