The Circuits Lab at UNLV is supposed to be a student's introduction to basic lab test equipment and the principles of experimentation. My lab partner and I were comfortably more experienced in the lab than that so we decided to challenge ourselves with our final project.
We knew that we could use a combination of 7400 series logic to create a series of ripple counters and display the current state on 7 segment displays. We began to think about using that to keep time. In the US mains power holds a pretty consistent 60 Hz, but this would mean working with 120VAC power which is usually frowned upon for an introductory lab. I have many years of experience working on AC power systems, but my partner does not. We spoke with our TA and decided to proceed with this crazy idea, but to use added safety devices like a ground isolating transformer to prevent any human body related circuits from being created.
Sixtieths of a second are not exactly a commonly used time measurement, but because this project was already significantly more complicated than necessary for the class we decided not to add the additional complexity of normalizing to a more common time measurement standard. We decided on 4 functional counter blocks that would ripple carry into each other as needed: sixtieths of a second, sixths of a second, seconds, and tens of seconds. Most of the counters could just be standard decade counters with a carry out to drive the clock of the next order of magnitude, but the sixths of a second counter would have to be designed to reset and carry out after the count reached 5. We were able to do this with a standard 7410 triple-input NAND gate.
Since we were going to use mains power to generate our clock it only made sense to design a rectification stage so our clock could operate independent of lab specific equipment. We found a transformer with a 19:1 configuration that would give us ~6.3VAC. A full-bridge rectifier made of discrete diodes and a 10,000uF smoothing capacitor rectified this to close to 9VDC. From there a 7805 linear regulator was used to make sure that the supply voltage was well within the 4.5V - 5.5V range for the TTL chips.
Before the 6.3VAC when into the rectifier we pulled a line of it off so we could create our 60Hz clock signal. Our first attempt was to connect the AC signal to the base of an NPN transistor with 5VDC connected directly to the collector and the emitter connected to ground through a current limiting resistor. We generated our clock signal off of the high side of the emitter resistor. This created what looked to the naked eye like a reasonable approximation of a square wave, but feeding that into the counter stages only caused confusion. Because our timing didn't depend on being in phase with the AC signal we ran our rough clock through an inverter with Schmitt trigger inputs and this cleaned up our clock signal into a nice and crisp square wave that made the rest of the circuit much much happier.
We added a pause button connected to the counters' enable pins and a reset button connected to the counters' clear pins and this gave us a circuit that could be used as a 60 second timer. We moved on to how we were going to display the current state of the counters. A simple CD4511BE BCD-to-7-Segment decoder would handle reading the counter state and outputting the correct 7 segment display signals, but we didn't want to use something as small the standard 7 segment display packages available to us in the lab. As a result we decided to design our own 7 segment displays. We used two 5mm LEDs per segment and put two 7 segment displays on each board. This meant that two identical display boards could display all four counter values. As luck would have it, Arrow was having a sale on some 16000 mcd Cree LEDs (part number LC503UYL1-15Q-A-00001), so not only were our 7 segment displays going to be much larger than normal ones at about 5cm tall, they would be easily visible across the room because of how bright it would be.
With all of these plans we built the schematic in Altium Circuit Maker and got to work doing the layout. We started this project fairly early in the term and this allowed us to look for slower and much cheaper PCB fabrication options. After some looking around we decided that DirtyPCBs' protopack option was going to the be best choice for us. This would give us 10 of each board (base and display) shipped from Asia on a two week turnaround for less than $50 total. Because their protopack has one price for all designs smaller that 5cm x 5cm and another price for all designs larger than that but small than 10cm x 10cm we decided to design for a base board that was 10cm x 10cm and a display board that was 5cm x 5cm. They also charge the same amount for solder mask colors other than green so we got the base in white and display in red just for fun. Below are the schematics, layout, and the gerber simulation of the base and display boards.
While we were waiting for our parts to come in we continued to prototype and began to connect the various functional sections of our design together in the lab. Overall this went fairly well with a few minor hiccups. By the time the boards got to us we were ready to test them as rapidly as possible. Below is a video of the first functional test of the base board before we made up our display boards with the high power LEDs.
We had to make a few small changes requiring us to cut traces on the board and run jumper wires, but all in all the troubleshooting process only took a few days. Happily all of the issues we did encounter were on the low voltage DC side of the design and our AC rectification circuit pretty much worked as designed first try. This project was a fun challenge and pushed our abilities as electrical designers. Sadly neither of us thought to take a video of the timer working before we disassembled the base board for parts to be used in future projects.