Led Power Circuit

[Rsmags]

I have an ultimate challenge for you guys. I need help designing a circuit to power 9 white leds (3.6v 20ma) using 2 AAA batteries.

Do you think any of these would work?:
White LED Drivers

Any help is appreciated.

Ryan

 

[Doug S]

[ QUOTE ]
Rsmags said:
I have an ultimate challenge for you guys. I need help designing a circuit to power 9 white leds (3.6v 20ma) using 2 AAA batteries.

Do you think any of these would work?:
White LED Drivers

Any help is appreciated.

Ryan

[/ QUOTE ]
Actually the circuit board from a $6 Dorcy 1AAA light will be close to your requirements. It is not regulated so the current will drop off as the battery discharges but the price is right. Short the 1 ohm resistor on the board and add some capacitance to the output for some efficiency improvement. 10uF or so. Run LEDs in parallel.

 

[Rsmags]

Doug S: I need something thats regulated.

What my application is:

I am going to make a mod for a 2AAA MiniMag (Not the Solitaire).
I need a circuit design that will run 1 5mm and 8 3mm Nichias @ 3.6v 20ma each regulated.

The 2AAA M@g is designed PERFECTLY in that I have been able to fit the 9 leds PERFECTLY so that 99.5% of the head is covered by a LED.

My design:

Ryan

 

[Steelwolf]

I would actually suggest the Max1674 or something with similar specs. The ICs you linked to had cutoff Vin of 2.7V. This would work better with 3 cells, since with 2 cells, you would very quickly reach 2.7V and still not have tapped even half the available power in the cells. The 1674 cuts off at 0.7V, so your cells would be properly used up.

Also, I have an idiosyncracy. I prefer using voltage regulation when running 5mm LED arrays. This is in the event that if one of the LEDs malfunctions, the circuit automatically compensates and reduces the current flowing, though the voltage remains constant. A current regulated circuit would continue to provide the same amount of current which must now be distributed among fewer LEDs. Sometimes, the increase in power-per-LED may be sufficient to blow out the other LEDs.

Best of luck with the Max1674 though. It is surface mounted and rather difficult to solder.

 

[RussH]

Nice work. I suppose the head might get a little warm but the power is low enough that I don't think heat will be a problem.
I pretty much agree with Steelwolf, especially about the voltage regulation. If you were really pushing the LEDs (50-60ma) you might want some small resistance in series with each one to make sure they share the current. But for anything around 20ma. it won't be critical, and I wouldn't bother with resistors or worry about whether it was current or voltage regulated.
Note that even with regulation I expect the 5mm would run brighter, even if it has the same current. I've not used any surface mount devices myself, but I have seen a lot more schematics using the 1600 Maxims than anything else. HTH, -Russ

 

[Rsmags]

Can any of you guys be kind enough to make a simple schematic that is compact with the Max1674 for me? I am not very good with designing circuits /ubbthreads/images/graemlins/frown.gif

How does one go about soldering Surface Mount Components? (Especially if they have shakey hands like me)

 

[springnr]

Surface mount components with lotsa leads can be a pain so I have at times superglued the part down to align the multiple leads till I get several soldered. Resistors, capacitors, an such aren't too bad though.

 

[cosco]

There was a schematic for madmax in past somewhere on net. I have got it saved in my HD. I can mail it to you if you are interested. It was designed to run single Luxeon but if I am not wrong it can work for you since it is voltage regulation an not current regulation. It is the wiring mentioned in datasheet of MAX 1674, simplified for the task to get into very small place of Minimag tube diameter.

cosco

 

[cosco]

PM sent

 

[koala]

A 1 watt Luxeon would give you a better color tint and less to think about uneven current to multiple LEDs. Sorry if I spoil the fun but IMHO luxeon would be a better choice.