Joule thief like boost converter / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Joule thief like boost converter

Move Thread

LAN_403

cjk2

Thu Mar 16 2006, 06:49AM

Registered Member #51 Joined: Thu Feb 09 2006, 04:17AM
Location:
Posts: 263

I got bored so I decided to build a simple boost converter to run an LED off a "dead" alkaline battery. My circuit works well but I wish there was a way to regulate the current the LED sees so its brightness stays the same over a range of input voltages. I am using a simple astable type circuit to generate a more or less square wave to drive a transistor with to chop the current into my inductor. My inductor is custom made using maybe 30 turns per layer of 30awg wire in about 5 layers on a 3/16" x 3/4" ferrite rod core. Any ideas on how to regulate to current? I was thinking it would be good if the on time of the circuit could change to compensate for different input voltages. Is this possible with my circuit?

Heres my circuit:

and here it is lighting an LED like it has been for the last 30 hours on a "dead" battery

1142491749 51 FT0 Boostconverter Running

[Edit: Fixed oversized picture]

Philcogrump

Thu Mar 16 2006, 11:42PM

Registered Member #212 Joined: Sun Feb 19 2006, 05:42PM
Location: Texas
Posts: 20

How "dead" is your dead battery? Voltages under 0.8V or something?

Would it have to be at least 0.7V for the circuit to operate because that is the common minimum voltage of solid-state devices?

cjk2

Fri Mar 17 2006, 12:11AM

Registered Member #51 Joined: Thu Feb 09 2006, 04:17AM
Location:
Posts: 263

My battery is now at .69v and the LED is still on. It has been runing for 48hours now. I dont think It will last much longer as the transistors will top switching at .65v. I consider this a great success though and I am happy I have been able to design and build a simple boost converter. I just wish I could limit current so it would last even longer.

Arjan EMM

Fri Mar 17 2006, 08:19AM

Registered Member #149 Joined: Sun Feb 12 2006, 09:11AM
Location: Netherlands
Posts: 12

You could for example play with the on and off times of the chopper transistor.
If you lower the capacitance of the left capacitor, of lower the resistance of the left 7.5k resistor, the on times get shorter. If you increase the right capacitor or resistor the off times get longer.
Both should lower the power consumption if i'm correct.

cjk2

Fri Mar 17 2006, 06:28PM

Registered Member #51 Joined: Thu Feb 09 2006, 04:17AM
Location:
Posts: 263

Yes you are correct. I have actually been using a 10k pot in the place of the right 7k5 resistor so I can adjust power. My idea was to find a way to set up some kind of a regulator that will automaticaly keep power the same no matter what input voltage is.

My LED is now getting very dim but is still on. My battery is at .66v now and the circuit has been running for arround 54hours.

kell

Sat Mar 18 2006, 04:01PM

Registered Member #142 Joined: Sat Feb 11 2006, 01:19PM
Location:
Posts: 102

..speculating and thinking out loud about a way to regulate voltage across the led so you won't have to insert a current sense resistor in the circuit...

Connect the anode of a series of 1N4148 signal diodes to the anode of the led. The cathode of the last signal diode in the series goes to the base of a npn transistor. You will need an electrolytic cap (guessing a few hundred uF) across the base-emitter junction of this transistor. Ground this transistor's emitter, and connect its collector to the base of the transistor that drives the inductor.
When the voltage across the led exceeds the series voltage drop of the signal diodes and the base-emitter of the transistor, the transistor will start to conduct and divert some of the drive current to the base of the transistor that drives the inductor, lowering peak inductor current.
Adding diodes in series with the 1N4148 will increase the voltage; you may use schottky or germanium if you can't get the exact voltage you want with ordinary silicon diodes, but I don't you'll need to do that.
I don't know if you will need one diode in the series or five, because you are driving the led with pulses and I don't know what peak voltage across the led will deliver the correct amount of average power at your frequency.
Final note, putting resistors in the base and collector of the transistor you're adding to the circuit wouldn't hurt.
All this just off the top of my head; the circuit may need some mods or tweaking. Just offering it for the basic idea.

Edit: I just ran across this Joule Thief link with a single-transistor circuit:

Conundrum

Tue Mar 21 2006, 09:32AM

Registered Member #96 Joined: Thu Feb 09 2006, 05:37PM
Location: CI, Earth
Posts: 4062

Have you considered using a TL431 as part of the feedback network?

Another well known trick is to use gate-source connected JFETs as current regulators.

-A

cjk2

Tue Mar 21 2006, 11:49PM

Registered Member #51 Joined: Thu Feb 09 2006, 04:17AM
Location:
Posts: 263

Thanks for the great ideas. I guess ill leave the circuit alone at this point though, seeing as it powered that LED for about 5 days.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob