Ode to Bodacious Breadboards, Part 6Ode to Bodacious Breadboards, Part 6

I think this is poised to be the last of our breadboard-related columns (at least for a little while). I also think it’s going to be a goodie because all sorts of design tips and tricks keep popping into (what I laughingly call) my mind. If you’ve just joined our party, you may want to take a moment to peruse and ponder Part 1, Part 2, Part 3, Part 4, and Part 5.

As a starting point, let’s remind ourselves that a lot of the breadboard-based prototypes I create are powered by Arduino Uno microcontroller development boards (I like the challenge of working with a 16 MHz clock and only 2 kB of SRAM), so that’s the basis for these columns.

We’ll commence by conversing about capacitors. As a reminder, capacitance values are measured in units of farads (F), and we typically deal with capacitors having small values measured in millifarads (mF, meaning 10^–3), microfarads (µF, meaning 10^–6), nanofarads (nF, meaning 10^–9), and picofarads (pF, meaning 10^–12).

Let’s begin by adding a 16 V, 100 µF electrolytic capacitor straddling the power and ground rails as close as possible to where the power and ground wires arrive from the Arduino.

Adding a 16V 100 µF electrolytic capacitor. CLIVE "MAX" MAXFIELD

This type of capacitor looks like a small coke can. It’s a polarized component, which means it can be connected only one way. Hmmm, that’s not strictly true. In the real world it can obviously be physically inserted either way. One way will work as desired. The other way will degrade performance or—in a worst-case scenario—the capacitor will fail in a catastrophic manner (which might start with it whistling and end with it exploding).

Related:Arduino’s Latest Board Targets Cloud Applications

If we assume you prefer your components to restrain themselves from whistling and exploding, then insert this capacitor with its anode (the longer lead) connected to the power rail and its cathode (the shorter lead) connected to the ground rail.

We are using this electrolytic capacitor to (a) filter out any electrical noise from the power source and to (b) help smooth out any power dips caused by components on the breadboard drawing extra current while switching.

The reason for using an electrolytic capacitor rather than the ceramic capacitors we are about to discuss is that electrolytics are available in larger values, such as the 100 µF we are using here. Also, the reason for using a 16 V part is that this is a common value (it’s OK to use a higher rated value like 25 V or 50 V if that’s what you happen to have).

The next step is to add any necessary 0.1 µF (100 nF) ceramic capacitors. Ceramic capacitors aren’t polarized, so we can insert them either way round. A good “rule of thumb” is to add one for each integrated circuit (IC) between the power and ground rails as close as possible to the wire powering the chip. 

Related:Meet the Arduino Uno R4 Minima and WiFi, Part 1

Adding 0.1 µF (100 nF) ceramic capacitors. CLIVE "MAX" MAXFIELD

Some people call these “bypass capacitors” because they provide a low-impedance path to steer voltage spikes and high-frequency noise to ground, thereby “bypassing” the IC. Other people call them “decoupling capacitors” because they act like small, localized energy reservoirs (miniature batteries, if you will), mitigating glitches in the power by keeping the voltage steady long enough for the main supply to catch up. When an IC switches, it draws power, which can cause a momentary voltage transient in the supply. This can affect nearby ICs that are “coupled” by the power and ground rails, hence the “decoupling capacitor” moniker.

Did you observe the “104” annotation on the capacitor in the illustration above? Were you wondering what this means? I was hoping you wouldn’t ask. We understand this to mean 10 x 10 to the fourth power (10 x 10⁴) = 100,000 measured in picofarads (pF), which equates to 100 nF or 0.1 µF.

Another question is how we deal with unused gates. For example, an SN7421 IC contains two 4-input AND gates. Suppose we are using only one of these gates (the one accessed via pins 1, 2, 4, 5, and 6). What are we going to do with the other gate?

Related:Meet the Arduino Uno R4 Minima and WiFi, Part 2

One option is to do nothing at all—just leave all the pins unattached. This will work 99% of the time. It’s the remaining 1% that can bite you in the nether regions if you aren’t careful. In a worst-case scenario, the gate could start oscillating and radiating unwanted electromagnetic interference (EMI).

Dealing with unused gates.  CLIVE "MAX" MAXFIELD

To be safe, we really want to disable the unused gate. One approach is to use a pull-up resistor to connect all the unused gate’s inputs to +ve (logic 1) as shown in (a) above. Another approach would be to use a pull-down resistor to connect one (or more) of the unused gate’s inputs to ground (0V, logic 0) as shown in (b) above.

Actually, as simple as this seems, it’s part of a much larger discussion, which deserves to be treated in its own right, so we will kick this can (topic) down the road to a future column.

Last, but certainly not least, do you recall my mentioning my current breadboard-based DIY clock project in our previous breadboard-based column? The point of interest to us here occurred when I added a 2:4 decoder implemented using two jelly-bean ICS—an SN7404 containing six NOT gates (I’m using only two of them) and an SN7408 containing four 2-input AND gates (I’m using all of them) as shown below. (Observe the electrolytic capacitor, the two ceramic capacitors, and the four pull-down resistors on the inputs to the unused NOT gates.)

My current breadboard-based project. CLIVE "MAX" MAXFIELD

As I mentioned in my last column, everything else was already working as planned, so adding the 2:4 decoder should have been a doddle. “This is going to be easy peasy,” I thought. “There are only two 14-pin chips,” I thought. “I’m using only six primitive logic gates,” I thought. “What could possibly go wrong?” I thought.

The answer turned out to be “Lots!” I was seeing weird voltages on the ICs and strange intermittent effects on the 7-segment displays. Eventually, it became obvious that there was something untoward in the heart of the breadboard itself. I moved the 2:4 decoder portion of the design to another board, and everything worked as planned. At some stage in the future, when I decommission this project, I’m going to dismantle the lower breadboard to see what the problem is, but that’s an activity for another day.

Meanwhile, as always, I welcome your insightful comments, penetrating questions, and sagacious suggestions. Also, if you have any breadboarding tips and tricks you’d care to share, please feel free to email me at [email protected].