Q&A, attempt one

I decided to answer a couple of questions people asked me in the comments on a video. This might be useful to other readers too. First, let’s look at ways to ground a power supply that wasn’t originally designed for it.

The simplest option is to connect the output common to the protective earth conductor. This gives us almost complete safety during use. Any fault inside the power supply won’t “spill out” to the outside.
At the same time, we can improve the input filter’s performance and protect the mains from interference by adding a couple of capacitors at the power supply input, as shown in the diagram.
Since we’re installing the capacitors before the fuse and connecting them to the power supply case, we can only use safety-rated Y capacitors. I talked about them in a previous video.


Keep in mind that in power supplies housed in a metal case, the output common is often not grounded—or, more precisely, it has a “soft” ground.
This approach reduces the interference coupled from the power supply to its output and makes it easy to connect power supplies in series, but it does not ensure personal safety.
That doesn’t mean this approach is bad; it’s simply different and serves somewhat different purposes.
If safety is required, you can simply connect the power supply’s earth terminal to the output common, provided there is a real earth connection. If you want to ground the soldering iron tip when using this kind of power supply, you can connect it directly to the earth terminal.


When using a power supply with a soldering iron, a “hard” ground for the tip is sometimes unnecessary or even harmful, since a grounded tip must not touch points at mains voltage.
Obviously, all work should be done using an isolation transformer, but not everyone has one. Those who don’t may find themselves working on a powered device and accidentally touching its “hot” section with the soldering iron.
To get rid of static without causing a short circuit, you can connect the soldering iron tip through a resistor. The exact value isn’t critical; 100–200 kΩ is quite sufficient and relatively safe.
Still, I’d recommend full grounding and an isolation transformer.



The second question on the agenda was: why is the output filter inductor placed in the positive rail of a power supply?
In fact, the inductor isn’t placed in the positive rail specifically, but in any rail other than the common.

Let me explain.
All electronic equipment has what’s called a common, ground, or chassis ground, and so on. All voltages and signals are referenced to this conductor.
What’s more, in fairly complex mixed-signal devices, the ground may even be split into “analog” and “digital” grounds.

The common is always directly connected to the power supply output, while we filter the power rail, as shown in the first diagram.

As you can see, the power rail can be negative as well as positive. In that case, the inductor goes in the negative supply rail, even though that rail isn’t the common. I’ve shown this in the second diagram. It’s just that, in single-output power supplies, the positive potential has traditionally been the “main” one. By the way, that wasn’t always the case. In the early days of electronics, PNP transistors came first, and circuits using them are easier to build with a common positive rather than a common negative. Later, cheap NPN transistors were developed, and things changed to what you see today, with the negative rail as common. Cars went through something similar: for a while, the positive terminal was connected to the chassis, not the negative one.

There are also multi-channel power supplies. In this case, the filter inductor is likewise placed in the output rail, leaving the common conductor untouched. I’ve shown this in the third diagram.



And, of course, here’s a video on the subject.


Thanks for reading, and as always, I look forward to your questions.