It is a pretty common first project to use an Arduino (or similar) to blink an LED. Which, of course, brings taunts of: you could have used a 555! You can, of course, also use any sort of oscillator, but [Mustafa] has a different approach. Blinking an LED with three resistors and a capacitor. Ok, ok… one of the resistors is a light-dependent resistor, but still.
In reality, this is a classic relaxation oscillator. The capacitor charges until the LED lights. This, however, causes the capacitor to discharge, which eventually turns off the LED, and the process starts again.
There is one wrinkle that could be considered a feature. In daylight, the capacitor will stay in the off state, so the blinking only occurs in darkness. Of course, the resistor also has to have a sufficient view of the LED. You might use this as a safety light that only works in the dark.
A simple circuit, but it just goes to show that we tend to forget the simple solutions in a world where a computer costs less than a dollar.
Of course, you can get a chip whose sole purpose is to blink LEDs. We always like examples of doing more with less.
LED is an active component.
it’s not, it has two legs, it’s passive.
Wrong definition. I have 2 legs, but I’m clearly not passive.
There’s a useful definition and a pedantic-but-useless definition in play here. By the latter, surely the only passive component is a superconductor. Maybe let’s use the former?
The definition to which you refer may be: passive components can only dissipate energy (convert electrical energy to another form, and maybe back, such as capacitors and inductors). Standard def, and an LED is passive in all senses
But if there’s any sort of modulation of that by another input, then you can always use it as an active circuit element, because energy dissipation always reflects back to the source of said energy and can be used to amplify or switch things, or do something else like computation or modulation. It’s the interplay of more than one input variable that makes it “active”.
I.e. a bog standard transistor with a fixed base current is technically a resistor that saturates at some point. It has fixed behavior and turns current to heat, but if you change the saturation point by varying the base current then it’s active, because the amount that it resists current changes the outcome of the rest of the circuit.
The main difference between a passive and an active part is whether there’s an extra input that can change the behavior on the go. If the input is electrical, then it can be properly called an active component, and if it’s something else like pressure or light, then it’s an active component in the broader sense, just not in the electronics sense.
Lamps were passive components. Traditionally, at least.
The little incandescent lamps for 6v, I mean, as used by vintage bicycles who had a dynamo. They worked with both AC and DC.
They also existed as a blinking type, which had a bi-metal inside.
One difference was, however, that the blinking was a bit softer in comparison to an LED.
The wire inside did heat up/cool down each time which caused an afterglow.
As a modern replacement, halogen lamps can be used.
They don’t blink, however. Or maybe they do? I don’t know.
Old Lamps were simply resistors. I don’t know a right definition for passive components that’d work everywhere, maybe because there’s none. Even capacitor aren’t passive since you can make a amplifier from them by using the electric field they are producing. Same for inductors (like in a relay). That’s why there’s no meaning to this “passive” / “active” fight, IMHO. This would have had more sense if the post was titled “Blinking a LED with the minimal number of components” or “The cheapest way to blink a LED” or something more measurable.
A passive component has one input and one output with a knowable mapping from one to to the other. By “input” we don’t count the “legs” of the component but the physical phenomenon of what goes in and out.
If it has two legs for current to go in and out, and Kirchoff’s rules are maintained, that’s one input roughly corresponding to “power”. The output is then heat, motion, light… there might be effects like hysteresis that means the output cannot be directly predicted by the input, but the cause of that is a property of the component itself and not another input. It’s an internal variable, making it a knowable mapping if you know the inputs that have already happened. A battery for example: is it full or empty?
The LDR has two physical inputs: one for current, one for light. One modulates the other, which means there isn’t a knowable mapping from input to output – you have variables that cannot be predicted by the properties of the part alone – making it an active component in this scenario.
That means a more universal definition of an active component is that there is at least one unknowable physical input that changes the behavior of the part in a meaningful way. An electronically active component is when another electrical variable effects the change, and this is what we’re usually talking about when we differentiate between passive and active electronic circuit elements. In this sense, the LDR is passive.
Or, instead of “unkowable”, we might say “undefined and variable”. It’s just saying that it isn’t asserted so we don’t have enough information to predict what the part will actually do.
@Dude, LEDs sense light too, it’s just their light production is normally the dominant function. Temperature is also a factor.
Hence why “in a meaningful way”. If the rest of the circuit isn’t built to do anything with it, then it is used as a passive component. You could, if you would, make a thermometer or a light detector out of it, possibly even a magnetic field or radio wave detector, and it would be an active component in the physical sense, albeit not electronically since the modulating input isn’t voltage or current or charge coming in from the rest of the circuit.
@Dude: Maybe I’m wrong, but I think if a component can ever be used in an active way, then it can only approach passiveness in normal usages without actually reaching it. But on the other hand no matter what you consider active, there will be some circumstance you could install it into in which it fails to demonstrate active behavior. So I can appreciate communicating whether parts are used actively or passively, but (outside of a diagram) I don’t consider it an ideal innate property by which to categorize types of part. Especially with unclear criteria, but even without that.
Bimetallic blinking bulbs are still available. I saw them used in a YT video restoring pinball machines.
Current regulation diodes are a thing….
Fun topic…
Most two-legged devices, like resistors, caps, coils, light bulbs, relays, etc, are certainly passive. But two-legged-ness (if I might coin the phrase) does not, by itself guarantee passivity.
I’d argue that a tunnel diode has two legs, but exhibits negative resistance and can be biased to amplify. It functions in that state as an active device.
I believe the ICs used in RFID tags are also just two-pin, harnessing power from the coil and transmitting data by changing their impedance (basically shorting out the coil)
Two legs, and passive in the sense it works as “just a piece of paper”, but arguably not a passive component
Anything that can sustain oscillation or amplify (which is the same as exhibiting negative resistance) is “active”, no matter how many legs or ports it has. A nullode has no legs at all and can still be considered active (hint: it has one port).
In this circuit, the magic isn’t hidden in the LED alone, it’s the LED+LDR combination which amplifies – and that has four legs (or two ports).
It’s in the same vein as the usual “It’s forbidden to forbid.” Notice that the voltage source is also using active components. You can probably blink a led by using a capacitor + resistor and the LED and a photodiode to charge the capacitor, but you’ll still need semiconductors in the circuit. It’s simply not possible to have something intermittent without some kind of threshold somewhere, and this threshold will always be from an active component.
I agree. If it’s a semiconductor, it’s probably not fair to call it a resistor. Unless maybe you call it a semiresistor, but then it’s not really a resistor, is it? Is a diode an active component? I think so.
You can easily blink a light with a capacitor, resistor and a relay. Like how old car blinkers work.
So…you don’t need semiconductors.
The mechanical resistance of the relay blade provides the threshold. Or the shape memory alloy that’s deforming under its own power dissipation heat. Or by accumulating water on a bistable cup with a push button on the lever. In all cases, at one point you must interrupt a circuit and this can’t be made passively. But, yes, you can change the technology.
Old car blinkers were heat-driven bimetallic clickers… I never saw a relay/capacitor arrangement in ant turn signal circuit.
My electronics instructor in the 90s is the one who told me that when teaching the circuit…so I could very easily be misinformed…or maybe they did it that way at one point and abandoned it.
Probably just what they used for us as an example when teaching us the circuit in electronics.
The title isn’t ‘a blinking led circuit using only passives’. It’s ‘blinking an led passively’. And passives are indeed blinking the led. So the title is fine.
Grow up.
I think the right word is nonlinear. The “detector” is active and uses light as a 3rd input.
For a flasher I think you need two things…a non- linearity and a phase shift. As you [correctly, I think] argue, if everything was linear then the system would eventually just settle to some lowest-energy state and park there. The phase shift in this circuit might be inherent in the response time of the LDR which, compared with photo diodes or phototransistors can be very slow.
/\ This. Notice that you can replace the LED by a resistive lamp as well (since until there’s enough current to produce the heat required to emit enough light, the LDR will charge the capacitor slowly up to the voltage of the divider on the left, then as soon as the lamp produce light, the LDR will reduce, discharging the capacitor and the cycle continue).
First time I see this in 3 decades!
Gosh now I wonder if you could make this work with lightbulbs — and if you could turn these into NAND gates.
That has been done:
https://hackaday.io/project/172413-lltp-light-logic-transistorless-processor
https://hackaday.io/project/161669-light-logic-hic-svnt-dracones
Thanks! This is wild.
I remember my first electronics project, an LED blinker from EKI… a green PCB with a silkscreen of R2D2 (or similar robot) on it. Parts list was dirt simple: 4x LEDs, 2x caps, 2x transistors, 4x resistors. No 555, no CPUs, just a 9V battery. And I learned how to solder putting it together.
I found something like that almost 40 years ago in a truck stop parking lot, I mean it wasn’t a kit, just a circuit. Still have it (somewhere).
And here I was wondering how you can blink an LED without a power source
That just requires a simple yet convoluted over-unity device, which is on a different stretch of youtube.
Why must everyone assume perpetual-motion or over-unity, (or, in this thread’s overwhelming-majority “NOT PASSIVE!”) and be so vocal against their own presumptions? So Much Irrelevant-Commentary and flat-out arguing to weed-through just to find a few gems (and there are a few, here, thankfully!). It’s like a bunch of basketball enthusiasts at a football game yelling “Travelling!” Friggin’ exhausting.
Free-energy is all around us… The sun, the wind, the tides, radio waves, even heat can be turned into electrical energy with zero claim of over-unity.
I’ve zero doubt it’d be possible to blink an LED by harvesting the ambient light that same LED collects, converts into a current, and stores as a voltage in a circuit with no other powersupply.
I’d sure be curious to see how simple that circuit could be. (my overcomplicated design would probably involve numerous LEDs in series to generate enough voltage to power a microcontroller in sleep mode, until some voltage threshold was charged-up via something like a joule-thief, then discharged quickly through that same string of LEDs. Wash Rinse Repeat.)
Curiously, I have a dollar-store LED lamp that has been glowing dimly for many years… I’m afraid to take it apart to see what types of batteries are in there, lest I disturb its glow. But I do recall it was 3 AAA’s, and The cheapness of the circuit may’ve involved nary a series-resistor.
This circuit, in this video, intrigues me… I personally, had no idea why it would flash, as opposed to finding some dim steady-state, until one of the deeply-burried gems of a comment, in here, suggested maybe the LDR has a bit of a delayed-resistance-change…
And that thought inspired the idea of possibly focussing an LED, via a magnifying-glass, onto an incandescent filament (which changes resistance with heat rather than light), for some groovy circuit.
Hahaha look at kindish discussion! I’m the “Mustafa” Dr. Peker. Ok I give up, and renaming the circuit as; ” Blinking a LED without transistor or IC” to let everybody happy.
You’re fine, my friend.
People can’t be complete pendants sometimes 😄
Nice simple circuit. Great to cover the basics now and again 👍
i’m not really qualified for the ‘active vs passive’ debate though “oscillator implies threshhold implies active” seems pretty compelling to me. i do imagine there’s probably a way to use the unusual characteristics of an LED itself to avoid using the “light-sensitive resistor”, but maybe not…after all, a joule thief, simple as it is, does have a transistor..
but i did make an led blink once using just a battery, an led, and a resistor. i got it from some like super bright leds dot com sort of website and the ad said “ultra bright red, 2.4V 100mA”. and the label on the bag said the same thing. so i assembled it with the resistor for 100mA (verified with an ammeter). it worked well for a couple weeks but it was soon exhibiting all manner of odd intermittent behavior. in its most exciting phase, it would blink at a reasonable speed, a few seconds on and off. best guess is it suffered thermal damage in such a way that it would get hot and turn off until it cooled off. not sure why it would work that way, just guessing from the outside. the effect was never quite ‘reliable’ but it also was unambiguously flashing…
eventually got the datasheet for it and found out it’s 2.4V 20mA! whoever scraped the datasheet for the product listing text had copied the “peak, max 100ms out of 1s” PWM figure instead! now that i’m driving them at 20mA instead, they last for years without doing anything weird. go figure
It’s more interesting in its interpretation, I think; I covered this in a video a long time ago:
https://www.youtube.com/watch?v=Iscf6UJF3MY
In short, you’re basically crafting a UJT from elementary components. Instead of direct carrier injection (i.e. from a PN junction), you’re using light — but that has the identical (conductivity modulation) effect.
LM3909 anyone?