Homemade Lightwedge Project

[Vaporware]

Ok, I'll give it a try this week. I have the instructions from the manufacturer, but I was wondering how you use the potentiometer to vary the brightness? it explains how to use a voltage divider to vary the brightness by changing the input voltage. Is that how you used the pot?

[scotty1024]

Another option for powering white LED's.

http://www.national.com/ds.cgi/LM/LM2703.pdf

You can buy an evaluation board: http://www.national.com/pf/LM/LM2703.html

It will happily run off 2 AA/AAA's and drive up to 4 white LED's with an efficiency of around 90%.

[obelix]

Originally Posted by Vaporware

Ok, I'll give it a try this week. I have the instructions from the manufacturer, but I was wondering how you use the potentiometer to vary the brightness? it explains how to use a voltage divider to vary the brightness by changing the input voltage. Is that how you used the pot?

With Rset (I used 100k potentiometer, the central pin and one of the side pins are connected together; the 3 pin potentiometer becomes 2 pin variable resistor).
Schematic is the same to the one shown on the page 1 of the datasheet (except I used 2 LEDs, 2 other pins are unconnected), for the Rset vs output current see table on p.9).

http://www.fairchildsemi.com/ds/FA%2FFAN5607.pdf

I was considering also LM2703 (and even purchased one), but it requires more elements (including 10 uH coil). FAN5607 requires only 4 ceramic miniature capacitors.

[NatCh]

Originally Posted by Vaporware

... I was wondering how you use the potentiometer to vary the brightness?

Brightness on an LED is a function of current, not voltage -- it requires a specified voltage to work at all.

As obelix explained, you just wire up the center post (the wiper) and one of the side posts and the result varies the resistance, which for a constant voltage, varies the current (Ohm's Law, and all that ).

The only things I'd add are:

You can, if you want to, also connect the wiper post to the other side post (in addition to the center one, wire them together I mean) -- it doesn't really make any difference electrically, but it looks a bit neater to some folks.

I'd suggest you connect your power to the post clockwise from the wiper post, that way as you turn the potentiometer clockwise the resistance will decrease, increasing the current and the LED brightness, and counter-clockwise will turn the LED 'down' -- that's the way most folks expect controls to work (clockwise 'up' and counter-clockwise 'down'), so it's usually more intuitive.

If your potentiometer isn't obvious about which post is which, you can always test it before you soldier the connections.

[Vaporware]

Ok, I'll give that a try. I knew about the current thing natch, I was talking about how in the FAN's instructions they have a PWM brightness control in the chip and a voltage controlled dimmer, but I never noticed the R v I table on page 9.

I liked the explanation about pots, so thanks for that.

I'm pretty thickheaded about electrical circuits, but if I think about them long enough I can normally handle a simple one.

I used to do BEAM analog robots, but I never got those things working.

[NatCh]

Originally Posted by Vaporware

Ok, I'll give that a try. I knew about the current thing natch, I was talking about how in the FAN's instructions they have a PWM brightness control in the chip and a voltage controlled dimmer, but I never noticed the R v I table on page 9.

Ah, sorry to tell you stuff you already knew. The LED Current/brightness thing was something it took me a while to get in my own head, so I jumped to the wrong conclusion.

Glad some of it was useful anyway.

[Vaporware]

No problem Natch, I have a lot of trouble with figuring out the required resistance to get the right current, so I'm in the same boat, just a different end.

I'm going to try the FAN circuit tonight, But I really need another lightwedge panel so I don't have to rip my current one apart :|

Electronics are so fun and scary at the same time.

[NatCh]

Originally Posted by Vaporware

No problem Natch, I have a lot of trouble with figuring out the required resistance to get the right current, so I'm in the same boat, just a different end.

At the risk of telling you something else you probably already know (), just divide the voltage you have by the current you want and you'll get the resistance as the product. You just have to remember to adjust for the Current's units (i.e. if you're using mA, multiply by 1,000 to get Ω's instead of kΩ).

Ahhh! It feels good to actually use something I learned from my College Degree.

[Vaporware]

Ok guys I've started soldering this thing up. Holy moly those solder pads are incredibly hard to solder. I need to know how to wire it up. I'm connecting V+ to the Vin pin with a 4.7 uF cap going from + to grd. V- connects to Vout with a 4.7 uF cap from - to grd.

uh, what is ground in this circuit? do I just solder the GRD pin and the caps negative leg to together to get ground?

[NatCh]

Ground is gonna effectively be the negative terminal of your power supply....

[Vaporware]

So I would solder both leads of the negative filter cap to the same terminal ? this is so confusing. You can see it on the Vout terminal in the picture.

By the way, Obelix, your method of breaking up stranded wire was brilliant, and it allowed me to finish soldering to the rest of the terminals. of course it took me all night, but I got it.

[NatCh]

You're right, that doesn't sound right.

Ah, I think I see the problem, Vout isn't V-, it's the output power for the LEDs.

Here, I've marked up the schematic a bit, blue for Vout and yellow for ground/V- connections, hopefully that'll clarify it some.

[Vaporware]

Ohhhh Duh. Yeah. V- will be ground, not Vout. Oh well, that's what I get for trying to work at midnight.

[NatCh]

Yeah, my brain shuts down around 11:45pm, too.

[Vaporware]

Well, I finished, but I think I burned the chip. I get 2V across Vout, but none across the LED+. I'm going to try for a second chip.