Project 01: LED Flasher

Tags

LED, flasher, blinker, pulser, oscillator, multivibrator, relaxation, negative resistance, positive feedback, Sziklai pair, complementary Darlington.

Difficulty Rating

2 on my scale.

Purpose

This is our starting circuit. It continuously blinks the LED (light emitting diode). While it's a simple electronic circuit, it has many possible modifications or 'hacks'.

Bill of Materials

These parts are all in the basic kit, which I described earlier.

Label	Description	Image
T1	NPN Transistor: TN3019A, TN6719A, 2N3904, or similar
T2	PNP Transistor: ZTX749A, TN6726A, TN2907A, 2N3906 or similar
R1	Resistor, 2.7 megohms (= 2,700,000 ohms). Color stripes: red-purple-green. Or any value between 2 megohms and 3 megohms should work OK.
R2	Resistor, 910 ohms. Color stripes: white-brown-brown. Or 1 kilohm (1000 ohms) or 1.5 kilohm (1500 ohms) could work too.
R3	Resistor, 68 ohms. Color stripes: blue-gray-black. Or 47, 51, 56 or 62 ohms could work.
C1	Capacitor, 220 nanofarad (= 0.22 microfarad). The marking on this capacitor is typically "224". Here, I show a ceramic disc type capacitor. But any type of capacitor -- as long as it's 220 nanofarads -- will work fine.
LED	LED. Red-light-emitting. A yellow, orange, or green LED is also OK.

plus these general parts...

Solderless plug-in breadboard
AA (or AAA) carbon-zinc cells, two (2)
Battery holder for the 2 AA (or AAA) cells	OR

Assembly

Click on the image to see a larger version.

Start by inserting the two transistors as shown. Note that their 'faces' (flat sides) are shown upward. T1, the NPN type, I've painted white to be visually very different than the PNP. (You should paint a dot on all your NPN's so you won't mix them up with your PNP's).

Next, moving across the breadboard from left to right, insert R3, C1, R2, R1, and the LED. All the leads have to go in their correct rows of the breadboard.

Attach the power leads only as the last step. Before you power-up a circuit, always double-check that the parts are connected correctly. The very low voltage we're using can't hurt us, but transistors and LEDs can be damaged if overdriven or overloaded.

Success is when the LED flashes briefly, about 1 to 3 times per second.

Schematic

Click on the image to see a larger version.

I realize the schematic might mean nothing to absolute beginners. But it does quickly communicate many details to those who know how to read it. If you've never before seen a schematic, maybe this will be educational.

I labelled circuit nodes (n0 to n6) to help me explain how the circuit works.

Images

Click on the image to see a larger version.

Here are voltage waveforms of nodes n2 thru n6. (Nodes n0 and n1 waveforms are simply flat lines — not too interesting). These are taken from the circuit simulator software, and not from an oscilloscope connected to a live circuit. So they may be slightly off from reality, but still informative.

How Does It Work

Well, if you really want to know...

Generalities

I'll explain the operation by describing changes in the node voltages and device currents. Voltage level is relative, so we pick a "reference node", a node to which all other nodes voltages are referenced. You can also think of the "reference node" as the "local ground" and is at zero volts, practically speaking. I'll read all the node voltages as if we clipped a voltmeter's negative probe to the "reference node", and the positive probe to other nodes in turn. In most circuits, this one included, I chose "battery minus" as the reference node. It's labelled n0 in the schematic.

Circuit Family

This circuit is an oscillator — it generates a regular waveform (as long as it receives power). There are many kinds of oscillators. In particular, this is a multivibrator, because it generates a pulse waveform, a waveform that jumps quickly between "on" and "off" (as opposed to a linear or resonant oscillator which generates a smooth undulation like a pendulum). A pulse waveform looks like a series of rectangular or square pulses. This particular circuit produces a very narrow pulses, like sharp spikes.

To oscillate, it relies on three principles: amplification, positive feedback and relaxation.

Three Principles

Amplification, where a small change at the input node n3 is translated to a large change at the output node n6, is provided by two transistors, T1 and T2. Each transistor by itself can provide a lot of amplification, like a factor of 300:1. When we cascade two transistors, as this circuit has done, the factors multiply. So the combined factor is like 300x300 or 90,000:1. This makes the circuit very sensitive and useful for many things. Actually, there are many different ways to cascade transistors — this circuit uses the NPN-PNP combination known as a Sziklai pair or complementary Darlington. I chose this combo as it's a simple-but-powerful non-inverting amplifier — one transistor alone can't do this — and we need the "non-" to meet the next requirement...

Positive feedback is setup by how C1 is connected. We can point to it. You can see a connection between the two nodes (input n3 and output n6). If you remove C1, the circuit stops oscillating — there's no positive feedback — the LED stays lit.

Relaxation is harder to "see". It's not a circuit connection. It's an opposing reaction, triggered to occur when certain voltage is reached. You can think of it as a "trap door" for voltage — when it gets too high, the trap door opens and dumps out the electricity, and then resets itself (the trap door closes up, just as it was at the start). Transistor T1 is the "relaxer" or "trap door" in this circuit — in particular, between its pins 1 and 2.

Also, as we'll see when we hack the circuit, we can push T1's operation far enough away from its trigger point, the circuit stops oscillating — the LED stays lit or dark (depending which way we pushed T1).

So, this little circuit actually has a lot of things going on!

Analysis

Before the power is connected, all nodes are at zero (0) volts (...as we'd expect ☺). The LED is dark. The instant we connect power, node n1 jumps to 3 volts, and n2 jumps to 1.5 volts. Now, node n3 is the real interesting node — whatever happens here affects many other nodes...

Node n3 slowly rises from 0 to 0.5 volts because of the tiny trickle of current thru LED and R1 that charges up C1. T1 is also connected to n3, but T1 is "off" (not passing any current) and not affecting things yet. Same case for T2 and node n2.

When n3 reaches 0.5 volts, a "domino effect" occurs — now transistor T1 "turns on" (passes current), which quickly turns on T2, which fully turns on the LED and pushes current thru R3. The LED emits light! But it doesn't stay lit for long. The trap door has opened...

That sudden rush of current thru R3 raises the voltage of n6 from 0 to 1 volt. This fast rising pulse of voltage is fed back to n3. Here's our positive feedback! It raises n3 to an even higher voltage, from 0.5 to 0.8 volts. Transistor T1 is very much "on", still turning on T2, etc. But T1 is now sucking current off of node n3 (it's falling thru the trap door), much more than R1 provides, tugging the voltage downward. That shot of electricity that was fed back via C1 gets used up quickly by T1. T1 then shuts off (the trap door snaps shut), which then shuts off T2 and the LED.

To sum things up thus far, the LED has gone from dark, to lit up, and back to dark. But I'll continue just a little more.

Node n6 drops to 0 volts. Because of the feedback connection via C1, the drop at n6 also yanks n3 voltage down... below zero... to minus 0.4 volts (...negative volts? yes, circuits can generate negative voltages). Now we're back at the point where the cycle will actually repeat. Node n3 slowly rises from -0.4 to +0.5 volts, again, because of the tiny trickle of current thru LED and R1 that charges up C1. Then follows the "domino effect", the "trap door" opens and closes, etc.

History

I'll mention a few historical points which relate to this circuit. First, Bell Labs invented the transistor in 1947, a PNP type. PNP transistors were first sold publicly by Raytheon in 1949. It took a few years before NPN types were manufactured and sold. Cascading transistors, the Darlington configuration was patented in 1953, where Mr. Darlington described cascading transistors in many ways, including a way like this circuit uses. Schematics for oscillators with one NPN and one PNP appeared in print in late 1950's. Harold Barney of Bell Labs was issued a patent in 1963 for an "artificial larynx" that used an oscillator very similar to this circuit.

The first LED sold publicly was by Texas Instruments in 1962, but it emitted infrared light, not visible light. Visible red-light emitting LEDs were sold by Monsanto and Hewlett-Packard in 1968. Prices were way too high for hobbyists. Once Fairchild joined this market in the 1970's, their improved mass-production abilities dropped LED prices far below a dollar.

Simple Mods

Here are some pretty basic mods (modifications) we can easily do.

1. LED Color — The circuit will work with either a red, yellow, orange or green LED. What about blue or white LEDs? I discuss this farther down.
2. Flash Rate — We can change the rate or frequency of the pulses several ways.
   • R1 — Using a higher (or lower) value for R1 will slow down (or speed up) the rate. Up to a point. As I mention above, if we push T1's operation far enough away from its trigger point, the circuit stops oscillating.
   • C1 — Using a higher (or lower) value for C1 will slow down (or speed up) the rate.
   • Supply voltage — This is not a mod, just an observation. The battery voltage also has some effect on the flash rate. Over time, as the battery gets drained, its voltage will drop from 3 volts, and the LED flashes get dimmer. When it drops below 2.7 volts, the flash rate will increase. When it drops below 2.5 volts, the LED will stop flashing.
3. Adjustable Flash Rate — We could add in series with R1 a "variable resistor" (technical name: rheostat) or more practically, a potentiometer using its leads 2 and 3. But this approach has issues. First, to get a nice wide range of adjustment, the potentiometer would need to be a very high resistance value, like 5 or 10 megohms — these are not common values for a potentiometer. Second, the change in rate as we turn the potentiometer knob won't be linear. It will change faster at one end of the rotation than the other. I'll provide a better way in a follow-up article.
4. Touch Response — With your finger, touch node n3 (that sensitive node!) while the circuit is operating. In the layout diagram above, it's the top lead of R1. You may get the LED to give an extra flash. Your body is a big antenna for stray signals, and capacitor of static electricity. Next, use two fingers to touch across R1. You should see the LED flash rate increase, or even glow continuously. Your skin is a resistor, feeding extra current into node n3, just as if we used a lower value for R1. And lastly, use two fingers and touch one to the top of R1 and the other to the top of R3 (which is node n0). You should see the LED flash rate decrease, or even go completely dark.

Known Weaknesses and Their Fixes

Brightness of Flash — The flash of the LED is very brief and seems dim. Can we make it brighter or lengthier? Yes, indeed! I'll discuss this in the next article.

Blue or White? — If we want to flash a blue or white LED, we need to increase the supply voltage. Blue and white LEDs are made differently than those other types — they need a higher voltage to light up, like 3.2 to 3.6 volts (depending on the exact variety). (Red, yellow, orange, and green LEDs only need 1.5 to 2.1 volts.) A simple fix is to add another AA cell to our setup, so we have three AA's in series for 4.5 volts total. Works great. I'll discuss another approach, using just 2 AA cells, in a future article.

Adjustability — If we want a smoothly adjustable flash rate, we need add just a few more parts: a potentiometer (like I mentioned in #3 above) and another resistor. I'll delve into all the details in an upcoming article.

← Previous Article	↑	Next Article →
	Home Page