solar blinking LED HELP

[team_nightstalker]

i give up for now. i have done everything. redid the ferrite, re did the whole thing from scratch and nothing works. still only getting 1.5 volts at the LED.

 

[ericgibbs]

i give up for now. i have done everything. redid the ferrite, re did the whole thing from scratch and nothing works. still only getting 1.5 volts at the LED.

hi,
Sometimes its better to walk away from a project for a few hours, come back with a clear head.
Its probably something so simple you are overlooking it.

Lets know when you come back on stream. EDIT: remember give us info on the solar cell.

As a last resort, we could sacrifice a chicken or something in order to placate the project gremlin gods!

 

[audioguru]

Did you use a flashing LED with the voltage step-up circuit?
When the LED turns itself off then there is no load on the voltage stepup circuit so the voltage skyrockets to a very high voltage that destroys the transistor and the LED.

I hope you know that the two coils must have the proper phase for the oscillator to work.

 

[Hero999]

Did you check the phasing on the coil?

Try swapping either the feedback or driver connections round and see if that works.

 

[audioguru]

Make a center-tapped coil which is easier to wind because it has only 20 turns of doubled folded wire, then its center-tap is obvious, the two ends have the correct phasing and it doesn't need a pcb.

 

[Hero999]

You mean a bifilar winding?

That won't help because the middle tap will need to be broken and the oposite ends of the each coil will need to be connected to each other.

 

[audioguru]

You mean a bifilar winding?

That won't help because the middle tap will need to be broken and the oposite ends of the each coil will need to be connected to each other.

I see what you mean. The two coils must be in series for them to be a single coil with a center-tap.

 

[chipi]

why not use the good old LM3909 CHIP TO BOOST THE VOLTAGE it only needs a small cap to boost the v output to a led and it produces good flashers for 3mth from a AA batt maybe longer. and it will fit in a 35mm can .

 

[audioguru]

The old LM3909 is obsolete and has not been made for many years.
Its datasheet shows a pretty big 300uf timing capacitor.

 

[chipi]

na 100uf even 50uf made very good pulses yes maybe its atleast 20+ years ago . dunt time fly

 

[iONic]

Maybe a different circuit design would help inspire some new energy.

You can find the info at this site: https://www.nifty-stuff.com/LED-boost-circuit-joule-thief.php

Warning: It will suck your batteries to the grave if not charged soon enough.

 

[Hero999]

That won't flash, it'll produce a continious light.

You could have a go at reducing the frequency so you can see it flashing but you'd probably need some pretty large components.

Try changing the inductor to between 10mH and 100mH and the capacitor to something like 10:mu:F to 100:mu:F. Failing that try the existing circuit using a flashing LED with a 100:mu:F capacitor in parrallel with it but you'll have to be careful or it'll cook the LED.

 

[iONic]

That won't flash, it'll produce a continuous light.

Never aid it would flash. It was a much better alternative than the rusty nail Joule Thief circuit he was working with.
He wans to combine the LED flasher circuit previously posted with something like this circuit.

 

[audioguru]

I was going to add Bill Bowden's LED blinker's slow oscillator to my solar garden light circuit but I don't think it works without an additional resistor.

When the transistors finish conducting and then try to turn off, the capacitor doesn't help the transistors turn off and there is no timing. So I don't think it will oscillate.

My modified circuit uses a resistor to ground for the capacitor to discharge then charge in reverse before the transistors begin to conduct.
Since the timing capacitor sees both polarities then it should be non-polar.