I built this flashing oscillator circuit from the web, don't work for me:



Then I got to thinking about it, and moved the vcc capacitor lead to ground, increased
the 10k to 68k, and now it slowly flashes, with led remaining on for ~1 second, off for ~10sec.

What I don't understand is the RC bit. Once the capacitor soaks up enough voltage
to overcome the diode drop (I measured .950v), it turns on the base, and then,
the cap falls to zero, or nearly so, before repeating.

I would think the cap voltage would fall just below the diode drop voltage as it gives a weak
fast flash, and then flutter around that. How can the capacitor completely discharge, when
the discharge path is blocked by the diode-drop of the diode?

Am I clear?

 

Hello,

How did you make the "transformer" in this circuit?
The dots drawn at the "transformer" indicate the start of the windings.

Bertus

 

Hello,

How did you make the "transformer" in this circuit?
The dots drawn at the "transformer" indicate the start of the windings.

Bertus

I wound 20x20 turns on a toroid. Starting at the top, I put 20 turns on the left side, and then fit another on the right side, ending at the bottom. The windings are "mirrored". Here's how I modified it to work.

 

Look up Joule Thief circuit.




Reference: https://www.instructables.com/Making-A-Simple-Joule-Thief-made-easy/

 

Look up Joule Thief circuit.


View attachment 252278

Reference: https://www.instructables.com/Making-A-Simple-Joule-Thief-made-easy/

Yeah, it kinda does look like that...
But I don't care about any of that business.

Ahem, please allow me to reiterate:

"What I don't understand is... How can the capacitor completely discharge, when
the discharge path is blocked by the diode-drop of the diode?"

 

Perhaps you could enlighten us as to why you are analysing this circuit...
Are you just trying to make a circuit to flash an LED?
Are you trying to do so with just the one transistor?
Is it purely an exercise?
Is it homework?
Are you limited to just the 1.5v supply?

 

Perhaps you could enlighten us as to why you are analysing this circuit...
Are you just trying to make a circuit to flash an LED?
Are you trying to do so with just the one transistor?
Is it purely an exercise?
Is it homework?
Are you limited to just the 1.5v supply?

1st, I started experimenting with high current electronics, then high voltage, and now I am exploring the joys of understanding circuits with very low voltages, <2v, and the components that are optimized for this range (battery powered electronics, low threshold mosfets, schottkey diodes, etc.)

Yes, flashing a led was where I began in my present exploration. Yes, I wanted to limit supply to 1.5v. This is an exercise/exploration, with an interest in practical application(s). For instance, one might use this to create long-life beacons for nightime UAV operations, or as a twinkling star on top of a very tall xmas tree with limited access. To be sure, there are many ways to flash an led, such as a square-wave pattern, but if you want shape the flash at low voltage, extending flash on-time and intermittency, then things become more challenging, capacitor leakage,
high value resistances, etc.

I'm with the understanding that the circuit that is attendant to the RC component is forbidden on AAC.
Be sure, I AM ONLY CURIOUS ABOUT THE R-C BEHAVIOR & THE CONDUCTION OF THE DIODE.

After another closer measurement, I see that the capacitor does not drop to zero volts, but .4v, and yet, the diode's forward voltage is .950v, so it would seem there is some kind of avalanche that is going on that I was unaware. The diode only begins conduction at .95v, but once it starts to conduct, it will continue to pass current until the voltage falls to .4v or thereabouts. Is this correct, or is it something else perhaps?

 

How about the collapse of the magnetic field in the inductor supplies the missing voltage the diode needs to conduct?

 

How about the collapse of the magnetic field in the inductor supplies the missing voltage the diode needs to conduct?

That's good thinking! I image the current would be tiny, but the voltage may be sufficient.

 

The capacitor does no need to discharge by much. The purpose of the capacitor is to hold the base bias voltage fairly steady at about 0.52V.

When the circuit starts off, the transistor is in the off state. The capacitor is slowly charged through the pull-up resistor until the base-emitter voltage exceeds 0.52V. The transistor turns on and the circuit oscillates. The back emf from the toroid winding is high enough (2V) to activate the LED.

Negative pulses induced into the other winding discharges the capacitor. In a short time the voltage drops below 0.5V and the cycle repeats.
You don't need the diode in your original circuit.

The circuit below causes the LED to flash once every 2 seconds, approximately.

FLASHING JOULE THIEF

 

The capacitor does no need to discharge by much. The purpose of the capacitor is to hold the base bias voltage fairly steady at about 0.52V.

When the circuit starts off, the transistor is in the off state. The capacitor is slowly charged through the pull-up resistor until the base-emitter voltage exceeds 0.52V. The transistor turns on and the circuit oscillates. The back emf from the toroid winding is high enough (2V) to activate the LED.

Negative pulses induced into the other winding discharges the capacitor. In a short time the voltage drops below 0.5V and the cycle repeats.
You don't need the diode in your original circuit.

The circuit below causes the LED to flash once every 2 seconds, approximately.

FLASHING JOULE THIEF

View attachment 252373

Yes, it works with diode ommited, but by replacing 47k for 1M (leaky cap).
Many seconds, off, ~2sec. on, repeat.

I suspect the back emf from the base coil is recharging the cap a tiny bit as it cycles, slowing its discharge, and keeping the led lit longer. Also, electrolytic caps have a rebound effect when discharged, and this may also affect the circuit. Yet, I am still surprised at how much hysteresis is exhibited.

Adding the diode back in seems to separate intermittency from duration. In this way, a cap across the led
lengthens duration. Changing the value of the 1k also has same effect.