Is there a contest where make a simple electronic device as complicated as poss

[Beamin]

...ible?

Kind of like the blinking light contest but say the goal is to turn on a LED but instead of doing it in as few parts as possible make a whole bunch of ridiculous electronics to do it?

What would you call that? I think that might be a future project where you just keep adding complexity to it.

[julianhigginson]

rube goldberg electronics?
https://en.wikipedia.org/wiki/Rube_Goldberg_machine

[bd139]

It's called IT.

[schmitt trigger]

A nephew once made a LED blinker out of a FPGA just for kicks.

Now seriously....you want to see a real complex yet commercially useful circuit? Look no further than an analog-TV sync generator made out of vacuum tubes.

[AndyC_772]

A nephew once made a LED blinker out of a FPGA just for kicks.

I've done exactly that in a commercial project, albeit as a way of testing a much more capable system.

With a few more lines of code, that LED could have easily been programmed to blink a pattern at several GHz.

[Brumby]

There was a suggestion a few months back about members of the EEVblog putting something like that together.  No contest - just a bit of EE whimsy.

[schmitt trigger]

A nephew once made a LED blinker out of a FPGA just for kicks.

I've done exactly that in a commercial project, albeit as a way of testing a much more capable system.

If I remember correctly, the purpose of his "experiment" was to use a complex and absurd Verilog command structure to perform a ridiculously simple task.

[NiHaoMike]

To light a LED, it (obviously) requires at least one LED. Then add as many things as you can think of to make it light up.

Here's my idea, which goes around the world:

Start with a large number of participants (in the US) playing an online "game", the more players the better. They are divided into two teams - one using AMD Ryzen CPUs and the other using Intel Coffee Lake CPUs. Both sides use their CPUs to mine earnhoney.

The mined coins are converted to Paypal and sent to Naomi Wu, who makes a short video every week announcing the percentage contribution by team. She also spins a roulette wheel every time, using the contribution percentages to define which sectors denote a win for which team. But that's not the final decision. She blows a whistle near the end of the video for the next step.

Julian Ilett preloads his breadboard shift register with a single one and lets it run on a 555 timer. The percentages define which LEDs denote a win for each team. He plays the video from Naomi Wu on his phone and when she blows the whistle, a 567 tone detector activates a latch that stops the 555, at which point the LED that's lit decides which team gets a win.

An ESP8266 in Micah Elizabeth Scott's lab reads the results and lights up a red LED if team AMD wins, a blue LED if team Intel wins, or both if there's a tie. The picture is automatically posted (possibly with Tuco in the view) and the players try to recruit members to make their color light up more.

I'm sure something similar can be implemented running in the other direction with GPUs (AMD/Nvidia) mining Curecoin/Foldingcoin, Micah Elizabeth Scott and Julian Ilett getting the coins (the players can choose who the coins go to) and deciding the winners, and Naomi Wu wearing an IoT necklace that glows red, green, or striped red/green depending on which team won and posting a picture.

[Richard Crowley]

Submit it for UL approval?   

[brucehoult]

That should nicely support Ms Wu in the style to which she would like to become accustomed.

[Gregg]

And the winner is XKCD
https://xkcd.com/730/

[brucehoult]

...ible?

Kind of like the blinking light contest but say the goal is to turn on a LED but instead of doing it in as few parts as possible make a whole bunch of ridiculous electronics to do it?

What would you call that? I think that might be a future project where you just keep adding complexity to it.

Have you read "Running Blind" by Desmond Bagley?

[brucehoult]

And the winner is XKCD
https://xkcd.com/730/

When that cartoon came out I actually worked out the resistance of the network of 1 ohm resistors. It must have been slow that afternoon...

[Red Squirrel]

Submit it for UL approval?   

Would be funny to do this for a simple led blinker project for kicks... and disposable income.     

[RobK_NL]

Blinking a string of 17 CREE XP-E's @1A from a 12V power supply requires some complexity ...

[hans]

Design a FPGA board with SDRAM/DDR that has enough memory to run a modern day operating system.

Then proceed to design a custom microprocessor and computer architecture. The ever increasing complexity of todays systems, especially a blinky program, requires that a functional programming language such as Haskell is used. Take care to use as many Monads or Applicative functors as possible, such that it is possible to eventually synthesize the processor design for a quantum computer. Since we cannot order  our quantum computer design yet, fortunately we have made the necessary preparations and prepared a FPGA board instead. For now we will need to do with one of the many FPGA hardware orientated EDSL's to generate VHDL/Verilog such that you can synthesize it for your specific FPGA.

After having finished a microprocessor that can run code and has a MMU for modern operating systems, you need a compiler for it. You could attempt to port GCC to your particular architecture, but since GCC is written in C you might as well attempt to reuse code from the Haskell design and implement a custom C compiler in a shared codebase between the processor.

Once you have finished writing this compiler, you can finally start on the real project and that is to port or implement a Java virtual machine for it. You could try to do so directly by implementing the JVM bytecode as your instruction set, but that creates a vendor lock-in which is undesirable.

Once Java is running and has relevant I/O drivers available, you can make use one of the many IOT and enterprise cloud service libraries available. The purpose is twofold; one is such that you can blink LEDs via one of the many home automation protocols, which can work from anywhere around the world. Secondly is that anyone else can also do blink the LED when they inevitably hack the crappy IOT implementation.

Please take however that this project is a short term proposal. There are rumors that running Java on quantum computer is a bad idea, as it will immediately break down quantum entanglement and crashes the machine.  /s

_{Man I do love shitposting..}

[Cyberdragon]

Blinking a string of 17 CREE XP-E's @1A from a 12V power supply requires some complexity ...

Um...use a relay?

[Beamin]

To light a LED, it (obviously) requires at least one LED. Then add as many things as you can think of to make it light up.

Here's my idea, which goes around the world:

Start with a large number of participants (in the US) playing an online "game", the more players the better. They are divided into two teams - one using AMD Ryzen CPUs and the other using Intel Coffee Lake CPUs. Both sides use their CPUs to mine earnhoney.

The mined coins are converted to Paypal and sent to Naomi Wu, who makes a short video every week announcing the percentage contribution by team. She also spins a roulette wheel every time, using the contribution percentages to define which sectors denote a win for which team. But that's not the final decision. She blows a whistle near the end of the video for the next step.

Julian Ilett preloads his breadboard shift register with a single one and lets it run on a 555 timer. The percentages define which LEDs denote a win for each team. He plays the video from Naomi Wu on his phone and when she blows the whistle, a 567 tone detector activates a latch that stops the 555, at which point the LED that's lit decides which team gets a win.

An ESP8266 in Micah Elizabeth Scott's lab reads the results and lights up a red LED if team AMD wins, a blue LED if team Intel wins, or both if there's a tie. The picture is automatically posted (possibly with Tuco in the view) and the players try to recruit members to make their color light up more.

I'm sure something similar can be implemented running in the other direction with GPUs (AMD/Nvidia) mining Curecoin/Foldingcoin, Micah Elizabeth Scott and Julian Ilett getting the coins (the players can choose who the coins go to) and deciding the winners, and Naomi Wu wearing an IoT necklace that glows red, green, or striped red/green depending on which team won and posting a picture.

Is there an intro video?

[Mr. Scram]

A nephew once made a LED blinker out of a FPGA just for kicks.

Now seriously....you want to see a real complex yet commercially useful circuit? Look no further than an analog-TV sync generator made out of vacuum tubes.

Isn't that one of the basic FPGA starter projects? I can recommend doing something similar to everyone with remotely similar interests.

[Red Squirrel]

Blinking a string of 17 CREE XP-E's @1A from a 12V power supply requires some complexity ...

Um...use a relay?

Current limited SMPS using reed relays instead of mosfets.   

I actually kinda want to try that now... 

[Beamin]

Blinking a string of 17 CREE XP-E's @1A from a 12V power supply requires some complexity ...

Um...use a relay?

Current limited SMPS using reed relays instead of mosfets.   

I actually kinda want to try that now...

Does that work because the mosfets are used as hard on/off switches? whats that called saturation mode?

[Cyberdragon]

It works, it's called a vibrator supply (used since low voltage tube equipment and some early DC to AC converters), but it would be low frequency. You would also have to have a pretty hefty subber to stop contact arcing.

[Red Squirrel]

Back EMF would probably be an issue too.     

[Cyberdragon]

Back EMF would probably be an issue too. 

There's filters on the input supply, probably acting as inrush current limiting too. Some even use a mechanical rectifier (often tied to the primary oscillator).

[NiHaoMike]

Is there an intro video?

Problem description:
http: //www.youtube.com/ watch?v=ATsIdNJ4pMc
(Link intentionally broken because it's way off topic.)

As for why only Ryzens and Coffee Lakes, earnhoney mining difficulty is quite high now and something like my old Sandy Bridge E would not be able to make enough to be worth the power it takes to run. The smartphone coins have also dramatically increased in difficulty, still yielding a good amount of profit for the power usage but it is no longer profitable to buy new phones for mining. None of the GPU coins easily convert to a form Naomi accepts. There is an ASIC coin that is currently leading the way in efficiency and does convert to a form she accepts but it requires... a practically unobtanium ASIC. FPGAs do well on certain coins except they take way too long to pay for themselves.

Trying to get back on topic, if the power issue could somehow be magically solved, it would be funny to mine coins using only 7400 chips. (Interesting mental exercise: How many chips would it take to give Naomi the salary of an average electronics engineer if you could somehow power them all for free?) And while we're at it, put a LED on every node to turn it into a light show that would be well at home in Las Vegas.

Some even use a mechanical rectifier (often tied to the primary oscillator).

I remember the time in 3rd grade when I found that it was possible to make a "DC transformer" by coupling two motors together. Or a mechanical inverter by coupling a DC motor to a stepper motor salvaged out of an old floppy drive, which would generate enough voltage to give quite a tingle. I also discovered that it was trivial to make a LED light up by connecting it to the stepper and turning it by hand, sadly before white LEDs were even commercially available...