[blagie]

For the record: Some of these are a bit over-the-top, and Amazon is trash for anything safety-related, but for the record, virtually all fuses have a pretty significant margin before they blow. The risk here is very limited:

- A 2A fuse needs to conduct 2A, no matter ambient temperature, manufacturing variation, age, etc.

- 95+% of the time, what you're protecting from isn't a minor overload condition but a short.

This means a reliable fuse purchased from the best manufacturer in the world won't blow at 2A. Something like 3-4A is perfectly reasonable. 5x rated current is not okay, but 2x is well within norm.

Even at 5x, though, the risk here is not huge:

- You've caught shorts which is the most common failure mode.

- Everything carrying current typically has a ton of safety margin built in, so even 2x isn't really a big deal.

- There's a fairly narrow set of failure modes which (1) won't blow the fuse AND (2) will still be small enough to be less than 5x.

- Those failure modes still need to occur near something flammable.

The key reason why there is a risk at all is that power is a square law of current. If this was 5x heat / power, I wouldn't be worried /at all/. 5x current is 25x power and 25x the heat, which is more of a problem.

[rightbyte]

No.

https://www.nilight.com/products/nilight-272pcs-standard-min...

These are not slow burning fuses or whatever. In their product video 6A blows a 5A fuse in like subseconds. However 6/5 rated current should probably blow slower, like days, or not at all? The 2A fuse is probably just made faulty.

Edit: Of course, the Amazon seller could be a counterfeit of "Nilight", which might be a proper brand?

[blagie]

Proper brands are not "Nilight" but companies like Vishay, Littelfuse, Eaton, Bel, Bourns, and similar.

Those brands can be found on sites engineers shop on, like Digikey, Mouser, Newark, and similar (which also lets you price / order by the thousands, and feed into manufacturing channels, rather than by boxes of random parts). They will have datasheets, rather than Youtube videos. Better datasheets will often show:

* Typical and worst-case behavior

* I²t, which is used to compute how long it will conduct a given current

* Impact on ambient temperature on rated current

... and similar

Here's a few random fuse data sheets from Mouser:

https://www.mouser.com/datasheet/2/54/SF_1206HH_R-3304051.pd...

https://www.vishay.com/docs/28747/mfuserie.pdf

https://www.eaton.com/content/dam/eaton/products/electronic-...

Some of these are more complete, and some less. That's a fact of life. I can buy a fancy fuse which specs and tests everything for a bit more $$$ for a medical or military application, or one which doesn't for $ for e.g. a toy. If all I'm worried about is a short, that's good enough.

A Youtube video which shows a 5A fuse blowing at 6A means either:

1. The video was baked for Youtube.

2. The fuse will blow at 4A on a hot summer day, running in an even hotter enclosure, and you've bought a bad fuse.

My bet is on #1.

[quickthrowman]

In the line voltage (120v and up) world, fuses are made by Mersen, Littelfuse, and Bussman. Mersen owned Ferraz and Gould-Shawmut and sold fuses as Ferraz-Shawmut until a few years ago when they rebranded to Mersen. Bussman is owned by Eaton, and Littelfuse is an independent company.

Those are the only three manufacturers I ever see specified by electrical engineers in construction specs.

[lnsru]

In addition to other things I am certified electrician in Germany. Your claims are wildly oversimplified. Every fuse has a type, rating and trip current. I pick them by planned application, installation type and cable length. Copper cables cost money, the safety margin is THE fuse. Otherwise cable insulation gets too hot and wears too fast.

[blagie]

Is English your native language? You might be confusing fuses with breakers.

Fuses aren't typically used to protect copper runs; breakers are. Fuses are used inside of electronics. Breakers are rated as you describe, to protect copper runs.

For fuses, this is a data sheet of a random fuse (first one on Digikey search; I didn't pick anything special):

https://www.vishay.com/docs/28747/mfuserie.pdf

[lnsru]

The fuses are used in older installations. 25 A 500 V rated fuses are absolutely normal parts. They are usually slower than breakers. Breakers came later. The “normal electrician’s” fuses: https://ep-us.mersen.com/sites/mersen_us/files/2018-11/TM-10...

[blagie]

FYI: Page 6 is worth looking at.

* A 25A fuse will conduct 52A-110A for 5-10 seconds, which is about the 2x in my original post.

* For shorter time periods, this goes up a lot (150-260A for 100 milliseconds).

I'm getting down-voted, but no one is posting contradictory information. This discussion feels more like reddit than HN.

As a footnote, at least in the US, in the 21st century, I've never seen a house with fuses rather than breakers. As you point out, copper is expensive, and a breaker is $10-$15. It's not only more convenient, but saves more money elsewhere.

[rightbyte]

In many houses porcelain fuses are still in use instead of breakers.

[nucleardog]

And just to provide some perspective… typical household breakers work very similar to this.

If your 15A breaker tripped instantly when the current exceeded 15A, you’d be resetting a breaker every time you turned on a stand mixer or something.

Instead they have two mechanisms of action—thermal and electromagnetic.

The thermal trip time varies depending on how much you’ve exceeded the rated current, but for a typical household breaker could be a couple of minutes.

The electromagnetic trip is the “hard cutoff” where as soon as you exceed that limit the breaker trips near instantly. Depending on the type of breaker this is usually 5-10x its rated current. So your 15A breaker may allow you to briefly pull 75-150A through it.

So while these fuses are mislabeled and should not be sold, it’s not as much of a death trap as it might appear on first glance from what it sounds like.

[donmcronald]

> Everything carrying current typically has a ton of safety margin built in

The thing that worries me is what happens when everyone is cutting corners because they're relying on someone else building in safety margin. Where's the tipping point when everything becomes unsafe?

[blagie]

My suggestion would be to travel the world, and see how people live in places with different margins. Life is sort of okay. Look at Soviet-era Eastern Europe, India, the diverse approaches across Africa, China, the rest of Asia, etc.

People get by. There isn't a sudden collapse or implosion.

It's also easy-to-manage. If there are suddenly a lot of electrical fires, regulators will step in before it goes out-of-control.

For my personal tastes, the US is too liability and safety-conscious on a day-to-day level, and not nearly safety-conscious enough on a system level. Risks like fires, where I live, are small enough that I'm not worried about them.

Antibiotic-resistant super-diseases? AI apocalypse? Thermonuclear war? Cyber-Armageddon (where every network-connected device is maximally bricked in the span of 30 seconds)? Climate change? Some weird super-pollutant? Systemic economic collapse? Civil war? Genetically-engineered super-bug? Running out of water in Arizona (or your other local issue)? ...

We have major disasters typically around once a century in any given location (WWI/WWII/30 years war/Bubonic Plaque/hurricane/etc.). It's hard to predict which one will happen; some are very unlikely, and some are pretty likely. If something hasn't happened in 50+ years, we stop worrying about it, and we completely ignore future previously-impossible risks.

I'm much more worried by those sorts of things than by fire safety.

[FireBeyond]

> Everything carrying current typically has a ton of safety margin built in, so even 2x isn't really a big deal.

Unless of course you _also_ bought it from Amazon, in which case, who knows?

[postpawl]

I saw an example of this recently during the test of some large DC fuses for solar: https://youtu.be/6o9tbhgtws4?feature=shared

Some are slow burn fuses and don’t pop until you exceed the rated amperage for a long time and by a lot of amperage. Not really what I expected. Some of the fuses do work like you’d expect and pop when they exceed the amperage by a minor amount.

[blagie]

When I was a student, my lab professors had a saying: "Sometimes a fuse will blow to protect the circuit, and sometimes the circuit will blow to protect the fuse."

One problem is that fuses are thermal devices. They break due to overheating. They take a while to warm up and blow. If a fuse overheats at 2A at 100ms, it will get that same amount of energy at at 20A over 1ms, which is an eternity by the standards of transistors.

Most silicon devices can destroy themselves in milliseconds or microseconds.

[philsnow]

> 95+% of the time, what you're protecting from isn't a minor overload condition but a short.

But these fuses aren’t labeled “fuses that will blow when there’s a sorry but not when there’s an overload condition”, they’re labeled simply as “fuses”, which should blow over their rated limit, full stop.

It violates separation of concerns if I have to think not only about the fuse itself but also the rest of the system that it’s plugged into.

[blagie]

You're mixing disciplines. Separation of concerns is a computer science principle.

It very much doesn't happen much when one is designing analog electronics, which is where fuses come up. If I'm designing a piece of medical equipment, it cannot kill a patient with any single point of failure. That's a systemic design issue, and any device, including a fuse, has to be seen in the context of the system it's plugging into.

It'd be possible to separate concerns, but it would result in grossly inefficient, over-engineered systems.

You do get a bit more of that in electrical (rather than electronics) work, since you need to count on idiot contractors who will screw everything up.