Why are power supplies so complex?

That process would be open-loop. Your voltage at the output of the transformer would vary depending on the input source voltage (which should be able to tolerate some +/-10% to +/-15% variation, I suspect) and on loading. Rectifying and filtering would yield something. But what? And how would the voltage vary versus various time-varying load situations?

Closed-loop control helps to hold a precise voltage over a wide range of currents, provide protection against all-too-easily applied short circuits, guard against over-voltage to protect the attached circuit, deal with large capacitive or especially inductive loads like a motor, accept a relatively large range of line voltages, etc. It may also need to start up the entire system with all those protections without accidentally tripping them (for example, protecting against over-current/short circuit while also properly powering up your own internal system, when current loading is momentarily high as you provide the initial charging of capacitors.)

And, of course, there is _efficiency_. If you plan for the worst case (say, a -15% low input source voltage), then when you have it hooked up to a line with +10% high source your output would nominally be too high and you will need to burn off the excess, unless you apply a switching system.

It does depend on what you are powering, though. If you are powering a train set and have a human to adjust a knob to get the speed they want, it can be a very simple system and not much different from what you point out above.

But I'm a hobbyist and not a professional designer in electronics.

Jon

Hi, Don. Unregulated DC power supplies do exactly what you're saying, and they're exactly that simple (except for a fuse). Transformer, bridge rectifier, filter cap, voilla. This is similar to the large open frame unregulated DC power supplies, and also the small unregulated DC wall warts.

If you want a fairly small regulated power supply, you can do it without too much complexity. If you want a regulated 5VDC at less than an amp, take your 10VDC wall wart output, and just add an LM7805 3-pin IC in a TO-220 package with a good heat sink, and a small (10uF) cap at the output. Simple, straightforward, fairly large and heavy, and probably less than 50% efficient (meaning it dissipates more energy as heat than it delivers to the load).

It's when you need more power, and need to reduce the size and increase the efficiency of the regulated power supply, that things start getting complicated. It's a tradeoff.

Good luck Chris

All electronics is complex. That's why it's fun, but also why it's hard to learn and understand. I could ask you "why is a car engine so complicated?"

John

I believe the "simple" power supply you describe is what you would find in most equipment that is more than 20 years old, and it's still a viable design today. I think now-days it is called a linear power supply. The newer power supply design is a "switching" power supply.

Most components in a switching supply are much smaller than those in a linear supply, primarily because of the frequencies at which voltages are transformed in the two types of supply. In a linear supply the frequency is

50 or 60 Hz from the Mains. But in a switching supply the Mains voltage is rectified, filtered and then used to power an oscillator at a high-frequency (maybe 50 kHz) which drives the power transformer. The more sophisticated switching supplies use regulators that permit the supply to work over a wide range of input voltages, for example 90 to 220VAC.

Don

Regulate the voltage. Limit the current during overload. Correct the line current power factor.

In the '70s I was shooting with Ikagami HL79 video cameras. Originally the AC power supply used a transformer like you describe. The thing was a real bear since it was so heavy. Later in the model line they developed a switchmode power supply. We loved it since it weighed

1/10th as much as the original. A EE told me that a main reason that so many things are switchmode these days is the cost of making huge multitapped transformers. When I repaired these things replacing those transformers would easily equal the original cost of the equipment. Richard

Then I would explain part of the complexity. There is a lubrication system, and this needs a pump, and to be cooled. There is a cooling system, which needs to circulate and needs a pump and a whole lot more cooling. There is a timing system, which must be precise and many things exist to make this happen. And so forth.

If I knew the equivalent for a power supply, I'd post about that.

For a simple linear supply, there's

AC inlet, fuse, switch, EMI filtering

Transformer for isolation and voltage stepdown

Rectifier for ac-to-dc

Filter capacitors

Regulator, an IC (complex inside) or a discrete circuit with reference, error amp, pass element, current limiter

Output capacitor

Bleeder resistor

Maybe remote sense circuits

Maybe overvoltage crowbar

Heatsinks, maybe a fan

PC board, metal chassis, in/out terminals

Switchers are more complex.

John

There is a good comparison between engines and power supplies in that "environmental friendliness" drives some of the added complexity -- all the EGR components, catalytic converter stuff, etc. in an engine syste, all the power factor correction stuff in a power supply.