The Flashing Light Prize

[mikeselectricstuff]

Inspired by the 555 timer contest...
Find an unusual, interesting or crazy way to flash an incandescent lamp?
Then enter the Flashing Light Prize :

http://www.flashinglightprize.com/

The Flashing Light Prize  is an informal & fun contest to find the most unusual way of flashing an incandescent light bulb. Build a system in line with the criteria set out below and you could win £200 !
?
Your creation could be something ridiculous and zany using bits of string and magnets or it could be a brilliantly elegant circuit topology that's never been tried before. It could be electronic, electromechanical, electrochemical, anything legal. All it's got to do is flash an incandescent bulb at between 0.5Hz and 2Hz. You even get to choose what kind of incandescent bulb to use.

[@rt]

I can think of a winner that would cost more than the 200 euro
... kind of a Rube Goldberg approach to the ring oscillator.

[mmagin]

Having never done it before (but having had enough experience to make me cautious), I'm really tempted to use this as an excuse to build a non-isolated line-voltage lamp flasher, 555 timer, "capacitive dropper" and shunt regulator for the low voltage stuff, sensitive gate triac.

I think I have everything I'd need, except for the triac.  Hmm.

[BU508A]

Just bring this up as a reminder. They have there some fancy, cool, weird and crazy stuff. And Mike's lightbulb is really impressive. 

[Howardlong]

Anyone else put in an entry here? My attention was drawn to the competition by Mike's entry.

This was mine, gawd knows WTF I was thinking.

[mikeselectricstuff]

Anyone else put in an entry here? My attention was drawn to the competition by Mike's entry.

Yeah, I think that was what kicked it off afetr not having much publicity previously - that was the main reason I did it.

[pknoe3lh]

This video makes me think I don't want to enter ...
https://youtu.be/ouAXAD5GxCs

[FrankBuss]

So this is my entry



For detailed description, source code etc. see here:

https://hackaday.io/project/26126-touch-lamp

[zackattack303]

This is really cool FrankBuss! Well done.

[FrankBuss]

Thanks. The judges will have a hard time, there are many entries with unique ideas. I couldn't decide a winner.

[Howardlong]

My favourite so far is "Four Candles", a Peltier based device powered by candles.

https://youtu.be/7TBxQuIb-Ek

[Howardlong]

Here's a second one of mine, "Pulse-o-tron".

https://youtu.be/yWravW6Uu5Y

[MK14]

I've NO connection with the following video.

It seems rather good. A Neon flasher, uses a valve/tube to switch a relay to control the lamp.

[jaromir]

And here is my entry, with no timing capacitor - using the lightbulb itself as main timing device.

pdf with some more explanation https://drive.google.com/file/d/0B7bWXJ-9TfE5WjNhVFBEa0hHZm8/view?pageId=109602901961672216857

[ivaylo]

+1 for using a TESLA transistor 

[RoGeorge]

Here are my entries:

First one was made mainly to learn how to make and edit a video with Blender.

Also, it was quite a surprise to find such a rare type of light bulb in one of my own junk boxes. I didn't knew such a thing even exists, never mind that I was having one:

https://youtu.be/p2XPMuDMxwo



The video for the real entry was only half-edited, and sent in the last minute before the deadline:

https://youtu.be/V7gQh1KFjqc

There was no time to add explanations in the second video, so it's probably dead-boring to watch.

It is made by 2 identical motors, each with 4 magnets on top, and a reed switch nearby. The 2 magnetic reed switches are in series with a light bulb and a battery. The motors rotate the magnets they have on top of their axle producing a rotating magnetic field.

When the motors are at the right distance from the switches, the reed contacts start to vibrate in the rhythm of the rotating magnetic field. Motors rotate very fast, probably around 2000 RPM.

The trick is that each motor has a slightly different RPM. The reed switches are in series, so they emulate a logic AND for the bulb. Boolean AND is the same as multiplication. Multiplying RPM1 and RPM2 corresponding frequencies will produce two spectral components with the frequency RPM1 + RPM2 and RPM1 - RPM2, because cos(x)cos(y)=1/2[cos(x?y)+cos(x+y)]

As a result, the bulb filament will see two frequencies, RPM1 + RPM2, and another frequency RPM1 - RPM2. Because the filament has thermal inertia, it will act as a low pass filter, and it will attenuate the component of frequency RPM1 + RPM2.

On the contrary, RPM1 - RPM2 will be in the range of about 1 Hz, and the light bulb will flash in the rhythm of RPM1 - RPM2.

In the end, we managed to produce a 1 Hz oscillation started from 2 motors at about 2000 RPM.

Same technique as used in telecommunication for frequency shifting, frequency modulation, software define radios, etc., but applied to an incandescent light bulb with the help of 2 reed switches.

[chefkoch84]

Here is our entry:

The Flashing Light Prize 2017 - Crookes radiometer


We use a Croocks Radiometer for frequency control, which is powered by the blinking light source itself.
We measure the speed by projecting a shadow with a red LED - (this is not powerful enough to power the radiometer itself).
We measure the changes in brightness using a photo-resistor. A little smoothing is done by an ESP8266 and the light is finally switched by a simple relay.

What perhaps does not really come out in that video:
The blinking frequency is influence obviously by the rate of the radiometers spinning; the average ingress power, so the wattage of the lamp, the distance, the style of reflector ... etc.
There is also a lower limit to the rotation ...the gas-friction of the wheel in there will cause the speed decrease in every dark-phase ... if it cannot reach the next bright-phase by momentum, the system will stay off forever.
(We demonstrate this at the very end of the video)

The resulting situation is a little bit an unstable equilibrium.
All this is surprisingly sensitive to little changes. Either getting too fast or stopping eventually ("too fast” for the price is faster than 2Hz)

One could ask if this is possible without the probe-light LED- so the only light source in the system is the light that also gets blinked
--->  Yes this is possible, but here everything is extremely sensitive to small changes.
Probably a bigger lightbulb with huge thermal mass in the filament would overcome this problem a bit.
Also for this setup, one has to be aware that the radiometer is still controlling the flash rate. A direct back coupling of the blinking light itself could happen, and that is not easy to overcome. 

[Brumby]

Also, it was quite a surprise to find such a rare type of light bulb in one of my own junk boxes. I didn't knew such a thing even exists, never mind that I was having one:

https://youtu.be/p2XPMuDMxwo

Very common globe for Christmas lights.  You would have a string of twenty 12V globes in series (for 240V mains) and with 20 normal globes, they would all stay on.  Replace one of those with the flasher globe and the whole string would flash on and off as one.

My mother hated the Bang - on and Bang - off starkness of this arrangement.  We also had an aluminium tree at the time, which wasn't the sort of place to put a mains powered string of Christmas lights - so I came up with a solution.  I cut off the mains plug and wove a slim 2 conductor figure 8 cable along the length together with a strand of tinsel for camouflage.  I then linked a point between every second globe to one of these conductors - alternating between them as I went along the length.  This gave me 10 sets of 24V lighting in parallel.  I then scrounged all the flashing globes I could find (3 of them) and put one in each of 3 sets.  Fed from a 24V supply, I now had 7 pairs of fixed lights and 3 pairs of flashing ones - and since the flashing ones were independent, their differing time constants gave a randomness to the display.

[FrankBuss]

Winner for 2017:

[Brumby]

That is a worthy winner.

Never seen a coil used in reverse before.   

[kalel]

Yes, it's a great one.

Just curious, was there something about 5 minutes running time on video? Some people mentioned in their videos that it should run for 5 minutes (unless I mixed it up with something else).

[ivaylo]

Your setup should be able to run for 5 minutes. If there is no doubt that it can, no need to shoot all 5 min.

[Howardlong]

Your setup should be able to run for 5 minutes. If there is no doubt that it can, no need to shoot all 5 min.

That was pretty much my understanding too, my first vid had it running for five minutes in real time, but my second didn't bother, you wouldn't believe how boring it is waiting, and frankly not many are going to watch the whole thing. Others did timelapse, which is a good compromise IMHO.

[kalel]

Your setup should be able to run for 5 minutes. If there is no doubt that it can, no need to shoot all 5 min.

That was pretty much my understanding too, my first vid had it running for five minutes in real time, but my second didn't bother, you wouldn't believe how boring it is waiting, and frankly not many are going to watch the whole thing. Others did timelapse, which is a good compromise IMHO.

Yes, that makes perfect sense. It's more interesting to see the original contraptions than watching the flashing itself.

[mikeselectricstuff]

2018 contest announced
https://www.flashinglightprize.com/