A little explanation is needed here:

I took a lab course in digital electronics last semester. As part of the curriculum, we had to conceive, design, and build a digital circuit which demonstrated our fluency in the course material. Since I've always been an electronhead (that's the electronics equivalent of a gearhead), I wanted to make something a bit more complicated than a device which counts the number of times you've pushed a button. I originally planned to make a complete, playable game, but I soon realized that it would take too long, so I decided to just build one of its key components: a variable-range random number generator. This circuit asks the user to enter a minimum and a maximum value, and then it randomly generates some number between them as many times as the user wants. The game required 7-digit numbers (i.e., a maximum range of 0 to 9,999,999) but I decided 4 digits (0 to 9,999) would be enough for a proof-of-concept.

Now, this task sounds trivial when you ask a computer to do it, requiring only a few dozen lines of assembly code, but that's because the computer has the advantage of being able to execute arbitrary, user-settable commands. The rules for our projects, on the other hand, specifically forbade any such software architecture. Instead, we had to build everything out of discrete TTL logic gates. This is like making a salad by first planting all the vegetables in your garden, watering the soil three times a day, watching them grow intently, and then, just when they ripen to perfection, ripping out all their seeds and planting those to harvest for your salad some time next year. Most people would order a cheeseburger instead.

Being a physicist, however, I don't know when to quit, so I gladly set to the task of designing the circuit, buying the parts (which, including breadboard, tools, and fast-as-hell shipping, came to about $160), and painstakingly assembling the board one wire at a time. I worked from sunrise to sunset (appropriately converted to Universal Student Time by shifting forward about six hours) for three days straight, partly because the actual wiring connections were being devised in my head as I went along. Everything was going well and working perfectly, and, at the end of the third day, I was 15 or 20 connections away from being done. (A good thing, too, because I was almost out of wires! How embarrassing.) Since I had enough hooked up to give a partial readout on the display (on a separate breadboard, not shown here), I decided to test my latest wiring progress by switching on the power and pressing a few buttons on the keypad (also not shown).

This was something I had done dozens of times throughout the assembly of this circuit, and it always behaved as expected, give or take a couple of stupid errors which I quickly fixed. So when I connected the power supply to the +5V rail, I had every intention of seeing a half-lit numeric display waiting in earnest for me to punch in my desired range and then give it the command to generate a new random number. That proud moment never came. Instead, as soon as the circuit powered up, the display started flashing gibberish at me without so much as a flinch of my fingertips toward the keypad, and nothing I did would coax it into behaving otherwise.

I will spare you the next 3 hours of debugging, mind wracking, and lengthy excursions into the unique dialect of cursing known only to frustrated engineers, and just say that I could not fix the problem before the project was due. The only cause of failure that I can give with any level of confidence is a short circuit inside the keypad encoder IC, which I've not shown in the photo in order to protect the innocent. As this is an expensive and uncommon chip, I only had one of them, so I couldn't replace it with another to see if that fixed the problem. Exhausted, at 4 am on the day the project was due, I finally threw in the towel, and subsequently slept on it because I hadn't done any laundry during the project so as not to waste valuable construction time.

Later that morning, I woke up, definitely did not shower, placed my breadboards in a box, and made my cheerless way over to campus. I explained everything in detail to my professor, and he kindly gave me a grade of 80% for my lovingly assembled electronic circuit which did absolutely nothing. One would think I'd be in high spirits, but to a budding scientist nothing can absolve the defeat of a failed invention lest it be its spontaneous repair.

This photograph stands as a monument to the trials and tribulations of those who have shaped science and engineering throughout the millennia with countless failures to mark singular successes. Well not really, I just want to have a memory of how my circuit looks before I disassemble this miserable piece of shit.