Hello everyone, I'm hoping someone can help me with a circuit I've been looking for. I'm basically after a circuit to make an LED throb, but not turn off. Essentially I want it to go from medium to full brightness.
All the schematics I've found so far only go from full to off back to full.

I'm quite confined for space so an Arduino would be too big.

Thanks in advance


Jonny

 

I have quite a few actually. It is a common request.

LEDs, 555s, Flashers, and Light Chasers Chapter 12.

The power supply must be 3X the Vcc of the LED plus a bit more for the simplest schematic. It is a very easy project overall.

 

Hi Bill,
Great schematics, but I can't tell which one to use since I don't want the LEDs to go out, just down to 50% brightness.

 

I will get back with you. It is generally a good idea to mention the little details such as 50% brightness on the first post.

No matter, it is still easy enough. Just drop the power supply to 3X the LED voltage and subtract a volt or two. Use the first schematic, the one with the two transistors and a 555. Generally the problem is to have it go dark just at the peak of the pulse, the bottom peak. Making it work outside this range is easy.

6V would do it for the power supply voltage.

 

Actually, I got it wrong, you need to increase the power supply voltage for it to work. 12VDC would do it.

Want to go over the theory of operation?

 

I popped out to shops at lunch and got two transistors and a 555 timer. Got batteries on order, so it'll be a couple of days.

What do you mean theory of operation?

 

I forgot to pick up a 22uF capacitor, I'll try to pick one up later.
Actually I've got a 11.1v Lipo I could use in the mean time, would that be enough to test with?

 

Not tried: To give a little more flexability to ckt, add an RC ,Rx, Cx to output of 555. Cx can be same as Ct. Color of LED is needed. Pots are almost a necesity to establish opereting parameters, can be replaced with fixed Rs later,. Rx larger than Rt. R1- 250k to 1MΩ. R2 not needed or part of R1. With pin 3 low, adjust R1 for 1/2 brillance, then adj Rx for max when pin 3 is high. Lots of interaction between pots. Power supply V about 3V higher than LED.

 

If you do want to go digital, look at
http://www.adafruit.com/blog/2011/0...lace-open-source-wearable-electronic-pendant/
/mike

 

If you wish LED to pulse as specified, try this: If you do not have pots might try fixed values as shown. Half brillance by eyeball adjustment of R1

 

This is a bit simpler, I think.

The funny thing is, the 9V is perfect for just turning the LED off, which is the preferred scenario. 12V will throw off the numbers just enough to do what you want it to do. The pot can be eliminated, but it is fun to play with.

A good theory of operation is shown in this article, it explains how it works. This circuit is the basic circuit you are working with, which makes it a good start.

555 Hysteretic Oscillator

The transistor circuit taps into the triangle wave and uses it to light the LEDs. Using two transistors as a Darlington Pair creates a very high impedance tap into the triangle wave.

 

This is a bit simpler, I think.

The funny thing is, the 9V is perfect for just turning the LED off, which is the preferred scenario. 12V will throw off the numbers just enough to do what you want it to do. The pot can be eliminated, but it is fun to play with.

A good theory of operation is shown in this article, it explains how it works. This circuit is the basic circuit you are working with, which makes it a good start.

555 Hysteretic Oscillator

The transistor circuit taps into the triangle wave and uses it to light the LEDs. Using two transistors as a Darlington Pair creates a very high impedance tap into the triangle wave.

I think I'll give this one a go as I got final parts for it today. Gotta dig the breadboard out tonight and give it a go.
I'll be using bright blue LEDs, is this likely to throw off the numbers anymore?

 

This is a bit simpler, I think.

That's really elegant. Love it. I understand the reason for the darlington, but does using a normal transistor cause too much distortion of the wave? Just idly curious how bad it would be.

 

With single PNP load timing will be shortened, with NPN, may stretch timing to infinity, assuming Z to be around 5k for either emitter followers. Different colored LEDs will effect 1/2 brillance vs Vcc, guess blue might require Vcc of 14.2V??

 

Circuit works a treat, I had a 5k pot lying around so I hooked it up and coupled it with a couple of fixed resistors to get a good value of timing between pulses.

Sorry to ask this, but would it be possible to make the pulses erratic, or random? Or make it appear less so constant.

 

Depending on the resources available, you can amplify a noise signal and use it to add variability to the circuit.

 

There are ways. For example, the chapter From Four, Twenty is pseudo random, but not very good.

You could go with a PIC or some other µC, they will do random (pseudo or other) a treat.

The Darlington pair is required, people have tried the same configuration with a single transistor with poor results. You need the really high impedance a Darlington Pair offers.

Bernard has pointed me to flickering LEDs that might help.

If you want flicker, look at the Fire! chapter 12. You can use audio to drive an LED, or treat it as a pseudo noise source.

What is the effect you are wanting to accomplish? It may not be flicker you are after.

You can also use slow music for slow flicker.

 

Basically I'm making a prop from the Marvel comics and recent Marvel movies called the "Tesseract". It's basically a blue glowing cube, it pulses and has movement within the cube.
I don't have a hope reproducing the animation around the cube, but thought the random pulsing should be achievable.
Below's is a link to a Youtube video where someone has made their own animation of the cube
http://www.youtube.com/watch?v=MZlnu99ggV0

 

I've used a 556 dual timer instead of the 555, and used the second timer to slowly change the voltage on the control pin of the first timer. That gives about a 3-fold range of frequency change, for instance from 0.5 to 1.5 Hz. I used the slow timer at about 0.1Hz, so that the change is slow.

I can post the circuit if you like.

 

Bill, I've implemented your circuit but found the luminesence of the LEDs to be far below what I get when I put 3v straight across them.
I've tried stripping out R3 but it's made little difference.

I've experimented a bit with R1 which does increase the brightness but then I'm loosing the strob effect. Any tips?

I have a 12v supply from 4 x 3v (2032) coin cells, and am looking to power at least 4 x ultra bright blue 3mm LEDs.