[mbesto]

Curious - what actual real life issues do real world people encounter with dirty AC waves? Like I always hear the proverbial "this could cause harm to electronics" but are there real world tests of electronics failing? Does it fail over time or because of a one time instance? Same thing with under/over voltage.

[mikeyouse]

Modern furnaces are weirdly sensitive to ‘clean’ AC power. Mine won’t work on bad non-inverter backup generators and interestingly to me, doesn’t work on non-bonded (ground-neutral) power from an inverter generator. Had to chop the cord off a drill and build a bonding plug last winter when I finally figured out why it wouldn’t run.

https://rvelectricity.substack.com/p/diy-generator-bonding-p...

[ak217]

My furnace is protected by an old school fast-acting fuse. One day it blew and at first I thought it was an anachronism from the house's original wiring but then realized it's intentional - the standard breaker upstream of it is not fast enough. Not clear if it mainly protects the blower fan motor or the circuit board - I suspect it's the motor. At least one other fan motor in the house got fried previously.

I think the quality of your power is determined mainly by the size of the transformer serving your neighborhood as well as the presence of noisy heavy power equipment like AC with poor/no soft starters or big brush motors among the consumers. It's noticeably worse on our street compared to where we previously lived.

[cwillu]

Breakers are there to protect the in-wall wiring, not the connected equipment.

[alnwlsn]

I discovered the same thing a few months ago with my home's furnace (which was a bit of a shock, because one reason for picking natural gas is so you can still have heat when the power is out). It runs just fine on a pure sine wave inverter though.

[garaetjjte]

Could it be that it directly uses AC for flame rectification and gets confused with ugly waveform?

[quickthrowman]

I had a weird situation with a 480V three-phase 250kW generator, it passed an 80% load bank test (the load is resistive heat) with flying colors, almost rock solid 60 hz frequency with smooth operation.

This generator is meant to back up a 400A electrical service that runs a 480V 200 HP three-phase water pump motor, which draws ~225A and is controlled by a VFD. This pump motor and drive run just fine on utility power, but when utility power was cut to let the generator drive the pump, the generator output frequency started jumping between 57-63 hz and a manometer placed on the generator showed the gas pressure fluctuating wildly.

Now, a 480V 250kW generator puts out (250000/480/1.732)=300.712 amps, which should be more than enough to run the pump. I asked the generator manufacturer what size generator they would recommend for this 200 HP pump, and they said for a basic 6-pulse VFD, it would need a 350kW generator with a 600kW alternator, but the existing generator could be used if a harmonic filter was added to the variable frequency drive.

I don’t really understand exactly what the issue was, but something with impedance and harmonics that could be corrected with a harmonic filter, and once that harmonic filter was added, the generator was able to run the water pump just fine.

[dylan604]

If you get dips in voltage below the range that the PSU can handle, it will kill the PSU. If you get spikes higher than the range that the PSU can handle, it can kill not only the PSU but things attached to the PSU as well. Most people are familiar with spikes with things like surge protectors, but most are unaware of how damaging voltage dips can be as well.

[Dylan16807]

> If you get dips in voltage below the range that the PSU can handle, it will kill the PSU.

How does that happen? Let's say it's running and I drop the input to 80 volts. Why do I get any behavior that isn't "runs at reduced capacity" or "shuts off"? And what part of the circuit is failing?

Are we assuming a combination of missing current limiting and a heavy load? If I'm just watching youtube then it shouldn't overload any components even if it keeps running at a really low voltage.

[dylan604]

My experience is from older analog equipment. Specifically, I worked in a video/film post house that also had a VHS duplication department. Analog tape machines run at a certain rate. For those old enough, VHS had SP, LP, and SLP modes. Those were defined speeds that the tape was fed through the system. If the voltage dipped, the speeds would slow down. When played back at a constant rate, the signal could not be read as the signal wasn't recorded at that speed. So our meter would sound an alarm when the voltage dipped too low that would introduce that problem which would tell the operators to stop the recordings and rewind/restart. That's more of an annoyance than damaging, however, some of the records of a certain model would fail which was usually a power supply issue.

Same facility was fed 3-phase power, but due to some construction mishap nearby, one leg was cut. That lost us a lot of expensive power supplies that day for some of the more expensive equipment.

Those are examples, but not really an answer to the how question. I'm not an electrical engineer, so :shrug-emoji: I asked the engineers that question at the time, and they told me it was something along the lines of the equipment tried really hard trying to function instead of just shutting off. The dip was low, but not that low. Described like a ceased electrical motor where it keeps pulling more power where normally a breaker/fuse would trip, but something different. It wasn't a satisfactory answer then any more than it is now.

[Dylan16807]

Well, I can think of a lot more ways for losing a phase to break things than a voltage drop.

For a voltage drop, the main idea that comes to mind is something trying to keep up a constant wattage and drawing increasing current to make that happen. But you have to do quite a bad job to design that circuit and not have a current limit.

And a PC power supply is inherently flexible on the input voltage, so it would never have the problems you get with a fixed ratio transformer on that old equipment.

[dylan604]

Unfortunately, there are a lot more things in the world that need a power supply than a PC. Sorry if my use of PSU unintentionally narrowed the focus, just faster to type. The power supply for these high end video machines were not small, nor come with a cheap price tag. I would not have expected them to be made poorly as they are specialty units designed to run precision analog electronics. That would be comparable to expecting nylon seats in a Ferrari.

From all of that, I have learned the lessons of how dips can ruin electronic equipment (even if not the exact why back then), so for me and my house all electronics are hooked up to a UPS or power conditioner. Appliances are on their own though as that's the landlords problem! Multiple times a week, I get noticeable dips where the lights visibly dim and I can hear all of the UPS units kick in and back to mains a few seconds later.

[cwillu]

Low voltage requires higher current to maintain a given output power.

[Dylan16807]

Yeah but that's what the shutoffs are for.

[exmadscientist]

This really should not happen, at least for units that are qualified to the relevant IEC standard.

But, certainly, garbage devices are all over the place.

[toast0]

over voltage (beyond reasonable tolerances) has a tendency to let the smoke out of components directly.

under voltage can do lots of things. Browning out with partial functionality can cause lots of problems. Some devices will pull about the same watts regardless of input voltage, so lower voltage means more current, and significant under voltage may require much higher than rated current and can damage connectors, leading to thermal runaway (loosened connector has more resistance -> more current -> more heat -> connector loosens). Brown outs during control sequences can lead to controlled loads running for longer than intended and over current situations too.

[lazide]

‘It lets the smoke out’ is a classic, and happens periodically. Bad waveforms cause weird heating issues, (literal) audio noise, and sometimes sporadic stability issues with computers.

It typically shows up ‘randomly’ unless you know how to attribute it.

[bob1029]

Audio amplifiers can be strongly affected by noisy waveforms.

Class D amplifiers and other topologies that depend upon SMPS for power delivery are usually unaffected. Class A/B is where you will typically hear it.

[markus92]

Capacitive touch screens when plugged in with cheap chargers to crappy waveforms, behave weird.

[krunck]

I've had servers that would not power on with non sinusoidal power.