Source: https://fullstackelectronics.com/epilogue

# Epilogue

What the course set out to do, what it did, what is still open, and what comes next.

The [Introduction](https://fullstackelectronics.com/introduction.html.md) started with a chip that blinked an LED and a question about its insides. I could name the blocks of the 555, two comparators and a flip-flop, but I could not explain them. I wanted to know three things: what each transistor contributes, how to measure it, and how to take that knowledge to a board of my own. Twenty-one experiments later, the board of Experiment 21 blinks the LED of Experiment 1. This page checks the result against those three questions. Then it gives what the course did not do, and what to do next.

![Photo: four builds side by side on a grey surface, each under a card with a number. 1: a breadboard with the NE555, a capacitor, an LED and a few parts. 2: a breadboard with two ICs and some transistors, the timer from IC blocks. 3: a full breadboard with many transistors and wires, the timer from transistors. 4: the black board with its eight pins, the silkscreen names of the blocks, and DISCRETE 555 in the middle.](https://fullstackelectronics.com/experiments/img/21-four-555s.jpg)

The four builds of the Introduction: the NE555 (1), the timer from ICs (2), the timer from transistors (3) and the board (4).

## The four builds

The Introduction promised the same timer four ways, each with a working circuit at the end of its stage. Each build blinked the LED with 10 kΩ, 10 kΩ and 10 µF. The table gives the period of each, against the 208 ms of the calculation.

| Stage | Build                                         | Experiments | Period    |
|-------|-----------------------------------------------|-------------|-----------|
| One   | The NE555 chip                                | 1           | 200.46 ms |
| Two   | The timer from comparator and logic ICs       | 2           | 199 ms    |
| Three | The timer from 23 transistors on a breadboard | 3 to 13     | 192.59 ms |
| Four  | The same circuit as a surface-mount board     | 14 to 21    | 210.11 ms |

All four are within 8 % of the calculation. The simulation of Experiment 16 gave 209.3 ms for the schematic of the board, and the board ran 0.4 % from it.

## What each transistor contributes

The table answered the first promise of the Introduction: four builds that work. The first question was about the insides. Stage Three gave each transistor a job, and each job came from a flaw that an experiment measured first. The table gives the blocks in the order that the course built them.

| Block | Transistors | Experiments | Job |
|----|----|----|----|
| Threshold comparator | Q1 to Q7 | 3 to 7 | A differential pair compares two voltages. A current mirror sets its current, an active load gives it gain, and an output stage swings its output near ground and the supply. |
| Trigger comparator | Q8 to Q14 | 8 | The same circuit with PNP input transistors, as in the NE555. |
| Divider | None | 9 | Three equal resistors make the two references, 1/3 and 2/3 of the supply. |
| Latch | Q15 to Q18 | 10 and 11 | Two NOR gates in a loop store one bit: charge or discharge. |
| Output and discharge | Q19 and Q20 | 12 | Q19 drives the output. Q20 empties the capacitor. |
| Reset | Q21 to Q23 | 13 | A low on RESET holds the output low and empties the capacitor. |

So the answer to the first question is yes. Each of the 23 transistors on the board has a job that an experiment of this course named, built and measured.

## How to measure it

The second question was how to measure each part. Each experiment gave the value to expect and the reason for it before the measurement. Then the oscilloscope or the multimeter gave the reading, and the page compared the two. The four skills on the [front page](https://fullstackelectronics.com/index.html.md) came from that method:

- Prototype and debug. In Experiment 12, a pull-up in the wrong strip held Q̄ at 0. The LED flashed once and stayed dark. The table of symptoms in that experiment finds such a fault from the LED and one voltage.
- Measure. In Experiment 6, the active load narrowed the switching band of the comparator from a quarter of a volt to a few millivolts. The function generator and the oscilloscope showed the change.
- Design a board. KiCad took the master schematic through the electrical rules check, the simulation, the footprints and the layout, to a design rules check with no errors.
- Assemble and test. The 45 surface-mount parts went on by hand with no rework. At the first power-up, the board took 0.01 A and blinked the LED.

## A board of my own

The third question was how to take that knowledge to a board. Stage Four did it in eight steps, from the master schematic to the first power-up. Each step had its own check:

- ERC in Experiment 15 and the simulation in Experiment 16
- DRC in Experiment 18
- The Gerber viewer and the paper test in Experiment 19
- The microscope in Experiment 20 and the current limit in Experiment 21

The first board from the factory worked.

The board is a 555 for this circuit, with limits. It does not replace an NE555 in each circuit:

- Its output gives about 1 mA through R21. The NE555 gives up to 200 mA. With the LED of Experiment 1 on it, the high level of the board is 1.82 V. For a larger load, give the load its own transistor, as Experiment 2 did for the LED.
- The board differs from the breadboard in the five places that Experiment 14 gave. The simulation of Experiment 16 ran each change, and the board works with them.

## What is still open

The board answered the three questions, but two results are not complete:

- In the board captures of Experiment 21, the capacitor trace jumps by about 1 V at each switch. The NE555 capture on the same breadboard has no jumps. I have not found the cause yet.
- I checked the reset of the board by eye only, with no capture.

The course also left out some subjects on purpose. It used bipolar transistors only, as in the NE555, with no MOSFET or CMOS versions. It did not go into the speed limits of the circuits or into noise.

## What comes next

The open items are the first things to do. After them, the board and the bench give more experiments. Each one uses a method of this course:

- Change RA, RB or C. Calculate the new period with the formula of Experiment 1, then measure it on the board.
- Measure the blocks of the board. The test pads give S, R, Q, Q̄ and REF13, so the board shows its latch and its references as the breadboard did.
- Make the timer faster. Make C smaller in steps, and find where the period no longer follows the formula. At that point the delays of the transistors become important.
- Wire the board as a monostable, the other mode that Experiment 1 named. The NE555 datasheet gives the circuit.
- Build a second board. The order of Experiment 19 gave five. Compare the periods of the two boards with the same RA, RB and C.

For me, the next step is the one that the Introduction named. I wanted time with analogue electronics before I took my own projects further into digital hardware. This course was that time. The transistors that made the comparators also made the NOR gates and the latch. So the analogue part and the digital part of the 555 came from the same parts.

If something in your build does not work, you can tell me at [@dnixty](https://x.com/dnixty). Tell me also if you find the cause of the jumps on the capacitor trace. The [contents](https://fullstackelectronics.com/index.html.md#contents) list each experiment, and the front page gives further reading.

[Back to the contents](https://fullstackelectronics.com/index.html.md#contents)