2
\$\begingroup\$

I'm writing a book about repairing electronics. The audience level for the book is "very little knowledge about electronics" and it's mostly about the mindset and mindtools for finding problem sources and how to correct them. It's not buried in maths or anything like that, as it's not a electronics book per se.

That being said, I'm currently a bit stuck on the current chapter about power supplies. I've been describing the very basics of linear supplies (fuse, graetz-bridge, filter cap, regulator, output cap), where error sources most commonly appears, et cetera. I think I've managed to capture the spirit of the linear supplies, but I have a very hard time describe SMPS on a very basic level and this is what I'm asking for input of.

I could slam the reader with block diagrams, schematics and describe the basics of this very complex topic, but I'm afraid I will miss the audience the way I've been trying to pinpoint it all. There's an awful lot of maths and electrical engineering behind SMPS but I want to avoid all that and just show-and-tell the very spirit. Showing a schematic like the one below (not mine) will only boggle the reader, I think, but somehow I have to highlight at least one schematic and point at which components fail most, etc.

How can I express myself about this topic, on a basic level?

enter image description here

\$\endgroup\$
3
  • 3
    \$\begingroup\$ I don't have an answer for you off the top of my head, but worth noting: if this book is to be published in english, it's probably better you call it a "bridge rectifier", not a "graetz-bridge"; the latter term is wholly unfamiliar to me and probably most other English-speaking engineers, let alone non-engineers. \$\endgroup\$ – Hearth Aug 14 '19 at 20:43
  • 1
    \$\begingroup\$ I'm actually not sure what you're having trouble describing when you can just describe a buck or flyback converter in a few sentences. I think you just have to make sure they have a decent grasp of understand how inductors and capacitors work, especially inductors when the current is collapsing. I think it's not so bad after that. But without that, it's tough. For a buck, a mechanical flywheel smoothing out RPM from pulsating inputs is perhaps the easiest to grasp but you'll still need to move beyond that into EE a bit to explain a boost converter. \$\endgroup\$ – DKNguyen Aug 14 '19 at 20:48
  • \$\begingroup\$ The way I have learnt this stuff, the underlying principle of an SMPS is the average voltage across the inductor being zero. \$\endgroup\$ – Bart Aug 15 '19 at 10:44
4
\$\begingroup\$

I'd recommend trying to describe a simpler circuit than the one you posted in your question.

A buck converter is a much simpler SMPS. It is also DC-DC, so you don't need to worry about AC rectification:

buckconverter(source: Wikipedia)

In a buck converter, the output voltage is less than the input voltage. I might describe it this way:

This type of SMPS works by pulsing current from the source to the load. If the pulses are long, the output voltage can reach the same level as the source voltage. But if the pulses are short, the output voltage will be smaller. You can control the output voltage by adjusting the pulse timing.

You could get more detailed if they already know about inductors:

When the switch closes, current starts to flow through the inductor to the load. Eventually, the voltage on the load will reach the supply voltage. But, say you open the switch before that happens?

Well, if there was no inductor, the voltage at the load would simply disappear whenever the switch as open. But the inductor makes it so that the current flow tapers off instead of abruptly stopping. Where does this current come from? It's "pulled" through the diode!

Obviously these are simplifications (perhaps over simplifications) of what's happening. But you could build up from there.

Another refinement is to explain how the inductor stores energy in its magnetic field, which is what keeps the current flowing when the switch is opened.

An additional approach is to use the water analogy. It isn't perfect, and the analogy normally fails when describing a SMPS, because electrical current doesn't have momentum as does water flowing in a pipe. But the inductor creates something very much like electrical momentum! (which I think is very cool, by the way.)

Good luck :)

\$\endgroup\$
1
  • \$\begingroup\$ Agreed. Describing the current behavior or the transfer of energy process is the way to go. The water analogy is always useful as well, specially for those with zero electrical knowledge. I would add the output capacitor to the Buck diagram and describe its function as a voltage smoother, or a bucket of water ;). \$\endgroup\$ – joribama Aug 15 '19 at 6:26
1
\$\begingroup\$

The basic-basics: input -> magic -> output, except the magic part doesn't violate the laws of thermodynamics, so power out = power in * efficiency. Meaning you can't get 10 watts out with one watt in, for example. It explains why a boost converter generates high voltage/low current on an output with low voltage/high current on the input, it explains why a buck converter does the opposite, and if you think about it hard it even explains why pretty much any switching supply with a regulated output has a negative input resistance.

After that: inductors are like tired children that don't like changing the speed they're moving (i.e. the current flowing through them), and they'll complain (with voltage spikes) if you try to change current rapidly. So most switchers work by getting current to build up in an inductor, then switching a transistor off suddenly which makes the inductor voltage change polarity and drive that current into something (originally a diode, but usually another transistor, these days) that puts it somewhere useful.

At the end, if you think you can do it without making people's heads explode, you can explain transformers and flyback converters and whatnot.

\$\endgroup\$
1
\$\begingroup\$

Short answer: if you can't explain inductors to your grandmother, then you cannot explain switching power supplies. Can you explain resistors? Capacitors? Yes? Well, first you have to "own inductors," and understand them as totally as you understand resistors, but without math and without lingo.

Most people understand resistors, but less about capacitors, but then have little understanding of inductors (understanding besides equations, or knowing how to use simulations. That's not 'understanding'.) To truly understand inductors, you must see them in your mind, see them working, as if all the invisible parts were visible.

First inductor lesson: a capacitor is not a capacitor if it's shorted out ...and an inductor is not an inductor if it's open-circuit. Just as capacitors are inherently open-circuit devices, inductors are inherently short-circuit devices. They are closed loops. If an inductor is not being shorted out, then it's not even an inductor at all.

If you visualize an inductor as a coil sitting on the workbench, but you don't every single time visualize the short-circuit between the terminals, then you don't understand inductors. (That's as bad as always viewing capacitors with a shorting-wire between their terminals. A shorted capacitor is not even a capacitor at all. A shorted-out capacitor is heresy, an abomination. Same as an open-circuit inductor.)

Inductors are flywheels. They are closed-loops, like a spinning wheel, or like a closed pipe-donut, full of water, with the water in rapid rotation. Inductors store charge-inertia, like a superconducting coil with no power supply. An ideal inductor is storing a loop of current: charges going in a circle, forever, with the inductor terminals shorted together. (Just as an ideal capacitor stores strained charges, the positive and negative held separated, forever.)

First SMPS lesson: what happens when you have a spinning bicycle wheel, and suddenly you jam a stick between the spokes? BOOM! Giant forces appear suddenly, the wheel halts, and the stick shatters. That's the SMPS "boost" principle: get your wheel going fast, then suddenly insert a stick (open the switch.) By suddenly blocking the flywheel, we can create a giant voltage. But it's a brief voltage. So, store it up, using a diode and a capacitor.

A battery, and inductor, and a switch, that's a high-voltage power supply. No need for step-up transformers.

\$\endgroup\$
1
\$\begingroup\$

You have already described linear power supplies transformers rectifiers and smoothing capacitors, then you are half way there.

A switch mode power rectifies the mains to give a high DC voltage. This voltage is then "switched" or converted back to AC at a high frequency. The high frequency AC is reduced in voltage using a transformer and this low voltage AC is rectified to produce low voltage DC.

The advantage of using the transformer at a high frequency is that transformers are much more efficient, smaller and lighter at high frequencies.

Much of the complication around SMPSs is about arranging feedback to the high voltage side to control the output voltage efficiently and in preventing the high voltage components from being destroyed. Operating these components in a high voltage high speed switching environment can be hazardous to their health.

Speaking of being hazardous to health your book should contain a safety warning about poking around in the high voltage side of SMPSs. They can contain large capacitors charged to high voltages, a hazard even when the mains is disconnected.

\$\endgroup\$
1
  • \$\begingroup\$ Good input. The book already contains a chapter dedicated to safety precautions, hazards with mains, capacitors, CRT, and the like. \$\endgroup\$ – bos Aug 15 '19 at 9:13
0
\$\begingroup\$

The short answer is, break it into smaller pieces, show how each one works, then tie them together as a system and show how they interact.

Have a look at this video as an example of a simple, but thorough walk-through of a switch mode power supply. https://www.youtube.com/watch?v=KIilQrgJ-gg

(Bonus: Commie test gear, and a cat.)

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.