Hey peeps I posted a question about solar lights previously and have a few questions regarding this diagram which I've now built:

1) How would you increase brightness while at the same time preventing flickering? Increasing the inductors to 330uH and decreasing capacitor to 330pF seemed to work however its still not bright enough.

2) How would I make the solar resistor less sensitive? Specifically, I want to make it so that the light only turns on in really dark conditions. Increasing the resistor to 10K seemed to have such an effect however at 20k the circuit just stopped working.. is the resistor the only thing thing that affects this?

Thanks guys

 

You can try inserting a resistor in parallel with the light sensor. Say, 2-5Kohms. That might help a little. Better to replace Q1 with a comparator. BTW, Q1 doesn't appear to switch any load. It's only purpose looks to be to drain the battery.

 

Gee, that schematic looks damn familiar: I believe I drew that! It's part of a solar led teardown thread (scopesys, post a link to it if you can).

Q1 has a vital use as this is a boost converter. Q1 turns on to build a current in L1, so when Q1 turns off L1 flyback and dumps it's current into the LED, and the LED "regulates" the output voltage.

I didn't design this circuit, I reverse engineered it from an old school light. The newer ones have a dedicated IC to do most of the functions, so it just needs a single inductor for the boost.

1) Not sure... decreasing L1 lets it build a current faster, but then it also has less energy to dump. I might guess increasing C1 would leave Q1 on longer to get more current into it (as long as L1 doesn't saturate).

2) The CDS works by shorting out Q1's drive to keep most things off. Increasing R1 should make the CDS need to go to a higher resistance to "open" but as you learned there is a limit. The parallel resistor idea may work.

 

lol yea sorry Ernie, it is indeed your diagram and I wanted to post a link however I dont have the actual link, just a saved html page

Ok increasing resistor to 30K (10K + 20K) seemed to have the desired effect (wish i looked on this thread before doing that xD), ill try the parallel idea.

Increasing inductor to 470uH increased brightness to the desired level.

I had random results with the capacitors 1500pF and 330pF both worked well, 15000pF and <100pF didn't work at all, I stuck with 330pF as it seemed a little brighter.

ahh here we go, found the link:
http://forum.allaboutcircuits.com/showthread.php?p=355913#post355913

 

No apologies needed. I'm actually pleased to see someone had read what I wrote (which is why I write things in the first place).