Super Cheap 1.5V DIY Driver!

[New_Reverie]

You've probably seen this, but there are some impressively small ones out there.

It would be pretty impressive if we could power maybe one of the 5.6mm reds with one of these things and have it sitting on your fingertip.

SMT Joule Thief - a set on Flickr

 

[Gryphon]

^ HOLY SHOOT!

 

[Benm]

Wow, thats a very compact version!

I've built a couple of circuits like these, just to power white/blue LEDs off 1.5 volts. The tricky part is that the inductor is a key component here, and its mechanical and electrical properties determine how well it all works.

There are versions with the bifilar coil as posted in this thread, but its also possible to make this circuit with a singly wound coil with a tap at some point... both work.

Downside of it all is that the output current is hard to predict/calculate and also depends greatly on input voltage. You can use it to power a bluray laser off a single 1.5v cell, but don't go anywhere near the diodes maximum current!

 

[Bluefan]

I would definately not trust a circuit like this unless somebody posted the output voltage measured by a good scope. Switched regulator are inherently noisy and that A Bad Thing for laser diodes. I have a fast scope but no time to make this (besides, I use linear regulators, way easier to keep clean).

 

[Benm]

If you put the diode and buffer caps behind it, thinghs should be measureable with common multimeters. Without, its more or less an oscillation of pretty random frequency, voltage and current.

These circuits are only good if you absolutely must power something from 1 volt or so - otherwise there are definitely better solutions available.

 

[Bluefan]

A multimeter can't tell if spieks are present, a long low peak in current is far less dangerous than a short high current peak, so a scope would be very recommended. A TRMS multimeter could say something about the noise, but not much, and I think not enough.

 

[drlava]

The joule thief is a nice little unregulated design, but as has been said, output varies with battery voltage somewhat and load as well. Perhaps there is a way to add some feedback with an op amp and current sensing and close the loop