Teardown of a PC Power Supply

But dielectric soakage runs in reverse, too. So even if you short out the capacitor once, it can recover some voltage after days to months left open again. (The energy comes from the molecules in the dielectric relaxing, very slowly.)

This isn't a problem if the designer has put in a bleeder resistor across the capacitor. But some designers are cheap, and don't want to pay the extra few cents to make their products non-lethal to technicians. Some are just stupid (yes, dumbfuck "senior engineer" coworker, I am thinking of you here, you colossal waste of oxygen).

What you want to do is short it out, with a screwdriver or otherwise, then keep it shorted with a gator cable or even resistor. Then you'll be safe even if you set the project aside for a week. Or three months.

Check for voltage - If not present / marginal, Then shorten, Else discharge through resistor and GoTo Check. A "capacitor discharge tool" automates this cycle.

it's also interesting to witness a bit glass thing acting like a capacitor

ps: for long I wanted to make a discharge rod but never finished it

Knowing that they could deliver a nasty shock, I discharged one once by shorting the terminals with a screw driver.

The sound of the spark, and the scorch mark left on the PCB and screw driver left me fearful of them.

Or was that caused by the knowledge that I had a friend who said he put one of those capacitors in a highlighter, thereby making a makeshift, single-use, "shelf-unstable" taser?

There are components in there that can exceed mains voltage when probed the wrong way.

When you start getting into the higher voltages, the body’s resistance is much lower. It only takes a few milliamps to stop the heart.

IIRC they were rated at 600v, and would leave black indents in the skin.

Kids are effing stupid…

You'll get a nice, painful reminder to respect electricity that'll last a few days, and not be in any actual danger.

(I also did the exact same thing as a teenager)

Although even putting this aside, capacitors don't seem to age well - anybody interested in 80s/90s electronics and gaming, will be aware that they should check boards before plugging them in, and if in doubt, but a 're-capping' kit

I don't know what -5 V was used for. Maybe some analog circuits somewhere, very early MOS logic also often had awkward supply voltages (charge-pumps made this obsolete too, just like high-voltage supplies for EEPROMs).

However usually there is. For example, the device might need +5V or +3.3V as well.

Well you can't supply +5V and PSU ground as separate leads, because almost always multiple ground connections are internally connected in a device.

Thus that would mean the +5V rail gets shorted to (PSU) ground by the device.

After quick searching is looks there are people reading this info with a MCU [1]. But I think it would be hard to find such PSU in ATX form factor.

[1] https://github.com/sxjack/dps750tb_psu_i2c

I don't really have enough data to outright blame the power supply. I don't think I connected all of the 12V lines to the motherboard (power supply didn't come with enough cables), and I can't be sure that the motherboard was designed properly for a mildly-overclocked 6950X (VRM banks, capacitors near the CPU, etc.)

It was all enough of a pain that I will never buy another Corsair power supply. My current best practices involve only buying Seasonic power supplies, and connecting every possible "optional" power connector to the motherboard. If the PSU doesn't come with enough cables to do that, buy them separately. My current Threadripper motherboard has a 6-pin and 2 4-pin ATX12V connectors, and a Molex connector. Many are marked "optional" in the motherboard's instruction manual, but I have them connected anyway. Probably overkill. Great stability. Without having designed the system or the test scenario, skipping optional things sounds like the sort of thing that's going to cost you a lot of time at some random point in the future. Best to avoid, even if you look like an idiot to the EE that designed the board.

The entry-level and lower-priced ones don't, AFAIK.

Here is my output from lm-sensors

    $ sensors
    corsairpsu-hid-3-1
    Adapter: HID adapter
    v_in:        230.00 V  
    v_out +12v:   12.00 V  
    v_out +5v:     5.00 V  
    v_out +3.3v:   3.00 V  
    psu fan:        0 RPM
    vrm temp:     +49.0°C  
    case temp:    +40.0°C  
    power total:  96.00 W  
    power +12v:   80.00 W  
    power +5v:    16.00 W  
    power +3.3v:   8.00 W  
    curr in:          N/A  
    curr +12v:     6.00 A  
    curr +5v:      3.00 A  
    curr +3.3v:    2.00 A

Even quality brands have crappy models.

They are not necessarily the cheapest ones, it is possible that the 400w is fine, the 550w is crap, and the 600w is fine again, even if they are in the same series.

They can also change the specs over time, without any indication on the packaging.

You have to look for reviews where they actually disassemble and check the insides of the psu to determine if it is a good buy.

There is a forum local to my language where they keep track of the small number of truly recommended models, there must be an English equivalent.

Though probably not due to current, likely someone took a torch to it to try to make it easier to remove, and then took a picture. Then again those connections are vastly undersized for the cables...

No heavy loads are connected to the 5 V / 3.3 V rails. Those are all supplied through the 12 V rail.

Obviously some components in a power supply still need to be through hole for mechanical reasons, but it isn't many.

[1] https://www.ti.com/lit/ds/symlink/uc2845.pdf Great generic current mode control circuit good for lots of things, not just supplies. Still in active use today. Datasheet is like a textbook. [2] https://en.wikipedia.org/wiki/Power_factor#cite_note-30

It's much more roundabout than that.

A transformer has a magnetizing inductance (you can measure this with an LCR meter by leaving all other windings open). This is generally a pretty big inductance (several henries for mains transformers). When you supply a waveform, that inductance will cause a small current to flow. That's the magnetizing current. The problem is saturation - if the voltage over time gets too large [=the magnetic field in the core], the transformer's core will saturate (that's also why transformers cannot handle DC; any DC offset will accumulate over time and cause saturation). This causes the magnetizing inductance to drop to approx. zero. When that happens, the magnetizing current becomes huge, basically a short circuit.

This means that you need to avoid saturation. Note that the strength of the field depends on the number of turns - each turn reduces the impressed field. So you want as many turns to avoid saturation as necessary, but as few turns as you can get away with (each turn adds resistance, and needs space inside the transformer).

And this is why mains transformers are so big. Because the frequency is low, voltagetime is large. So they need many, many turns (a few hundred to thousands) in order to prevent their steel cores from saturating. So the primary winding is a very long wire, which has a lot of resistance. So you need to increase the size of the core to fit a thicker wire, in order to avoid resistive losses. Now that bunch of wire doesn't fit any more through your transformer, so you need to use a bigger core, but wrapping N turns around that needs even more wire and so on. So you end up with a big, heavy transformer.

For a high-frequency switch mode power supply, the frequency is high, so voltagetime is orders of magnitude lower (e.g. 100 kHz vs. 50 Hz = 2000x lower), so they need far fewer turns. So their primary winding's wire is short, so it can be a) a little bit thinner and b) takes up much less space! So you get a small, light transformer.

But the core, it doesn't really care about power, like at all. It's all about not driving it into saturation, and having windings that can carry your current. Those two factors make the cores of mains transformers so big.

From https://www.electronics-tutorials.ws/power/switch-mode-power... "The Buck Switching Regulator" looks like an arrangement with a diode, inductor and capacitor is enough. Interesting.

A transformer converts electrical energy into a magnetic field and back into electrical energy. All the energy from a single 'sine wave' must be stored in the magnetic field inside the transformer briefly before coming out of the output. The transformer core is made of steel usually. For a given weight of steel, there is a maximum amount of magnetic energy it can store.

That means that at low frequencies like 50/60Hz, transformers end up very bulky.

The motivation would be to try to eliminate the full bridge rectifier and the high-voltage capacitors.

I'm not an EE, but I would guess the simple answer is it's because you can't pass a AC through a transistor, since a transistor not only switches but also acts like a diode.

I think you could work around that (build a circuit that switches AC with transistors), but there must be a reason why you wouldn't want to. Perhaps it's because it's simpler to just go ahead and rectify, or perhaps there's an advantage to having capacitors on the high-voltage side.

And I'm not up to date on modern electrical design, but a satellite-view on industry innovations & trends would be interesting.

https://www.google.com/search?q=lenovo+charger+gan

Electronic speed controllers for brushless motors are one area you can really see this too - they are amazingly compact, efficient, and high power now.

"Magnetic Amplifiers

Although called a magnetic amplifier; this application really uses an inductive element as a controlled switch. A mag amp is a coil of wire wound on a core with a relatively square B-H characteristic. This gives the coil too operating modes: when unsaturated, the core causes the coil to act as a high inductance capable of supporting a large voltage with little or no current flow. When the core saturates, the impedance of the coil drops to near zero, allowing current to flow with negligible voltage drop. Thus a mag amp comes the closest yet to a true "ideal switch" with significant benefits to switching regulators."