2
\$\begingroup\$

Can this one cell LED driver work as an astable multivibrator?

one cell LED driver

if so how is it compared to the Garner's Oscillator?

Garner's Oscillator

They look very similar, are they the same? Is the inductor coil playing a important role there? Sorry my lack of knowledge.

Do you know some variations that use only discrete components (no IC) but as few as possible compared to the conventional transistor astable multivibrator given I need only one oscillating output? Would it be more efficient giving less resistance and thus less heat?

P.S.: A found this detailed explanation that answer most of my doubts, is it the same Garner's Oscillator? I also found this other one but it's clearly stated to be the Garner's Oscillator. The only difference is that sometimes it have a inductor coil and sometimes it doesn't...

P.P.S.: I let the question open without mentioning what I'm building because it's mainly for learning purposes, what I'm building is a DC buck converter from 12V to 5V, but I'm very aware I could just buy a LM2596 and it would fit very well my goal, but I'd learn nothing from that!

Not an answer maybe

Here the summary of the information collected in the commentaries, I decided to not post as answer because I’m not confident enough but at least having it here could help people with similar doubts.

Here @Finbarr confirmed the shown one cell led driver is in fact a astable multivibrator.

Whether one cell led driver and Garner's oscillator are the same thing or not I think depends on the Garner definition, it's very likely the same, I’m just not confident enough to be sure.

@G32 pointed out it's possible to have even simpler solution using a NPN as negative resistor thought breakdown giving a example here even though they made clear it has some limitations.

@D.A.S. also pointed out it may be possible to have an oscillator using a Diac, maybe it would be a similar solution to the NPN breakdown.

\$\endgroup\$
7
  • \$\begingroup\$ Asking for as few components as possible is only beneficial for cost purposes but since there's no mention of target specifications, i.e. oscillation frequency and frequency/gain response, there's no clear recommendation we can provide. \$\endgroup\$
    – Colin
    Commented Nov 28 at 17:21
  • \$\begingroup\$ @Colin you're right, having less components doesn't necessarily mean it's less complex, but I'd say that tends to be. The idea is not to be cheaper, it is just to eliminate waste, then maybe the better phrasing would be asking for efficiency comparison between the conventional transistor astable multivibrator with the garner's oscillator? \$\endgroup\$
    – Tiago Pimenta
    Commented Nov 28 at 17:47
  • \$\begingroup\$ Does this count cappels.org/dproj/simplest_LED_flasher/… Google - Single Transistor LED Flasher Circuit Using Negative Resistance \$\endgroup\$
    – G36
    Commented Nov 28 at 18:46
  • \$\begingroup\$ "Can this one cell LED driver work as an astable multivibrator?" Well, that's exactly what it is. \$\endgroup\$
    – Finbarr
    Commented Nov 28 at 18:51
  • \$\begingroup\$ simpler is not a powerful spec without a purpose or specs like impedance, voltage, risetime, frequency etc. But there are 2 pin devices that oscillate with bias and capacitance. Diac etc or Schmitt Trigger inverters. \$\endgroup\$
    – D.A.S.
    Commented Nov 28 at 19:15

1 Answer 1

Reset to default
3
\$\begingroup\$

A single unijunction transistor (UJT) or programmable unijunction (PUT) can work as a relaxation oscillator. So can a neon bulb. So can a tunnel diode, but they are weird.

The most simple 1-device oscillator with a "normal" part uses a transistor's reverse breakdown characteristic to form a relaxation oscillator.

enter image description here

The schematic is from here:

http://www.learningaboutelectronics.com/Articles/Relaxation-oscillator-circuit-with-a-transistor.php

AFAIK, Dick Cappels is the originator.

http://www.cappels.org/dproj/simplest_LED_flasher/Simplest_LED_Flasher_Circuit.html

\$\endgroup\$
3
  • \$\begingroup\$ Do you know if ngspice is capable of simulating a NPN breakdown or it would be too much? \$\endgroup\$
    – Tiago Pimenta
    Commented Nov 29 at 4:21
  • 1
    \$\begingroup\$ @TiagoPimenta The standard model structure cannot do it but Alex Bordodynov created some subcircuits that add the functionality: forum.allaboutcircuits.com/threads/… \$\endgroup\$
    – Ste Kulov
    Commented Nov 30 at 18:56
  • 2
    \$\begingroup\$ The Mad Russian strikes again. \$\endgroup\$
    – AnalogKid
    Commented Nov 30 at 22:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.