We like to think that most common electronic components are essentially commodity items. We don´t buy premium wire or resistors. You just assume these electronic components are more or less the same from anywhere unless you need some very special characteristics. What about fuses? We would assume they are all essentially the same, but [Ham Radio A2Z] says he’s throwing away his generic fuses after he found they didn’t work as he would expect.
Of course, name-brand fuses are tested to very specific tests, and you get to see the plots of how the fuses are supposed to melt for Bussmann fuses. Then he takes out a generic assortment of fuses he bought at a hamfest. No Bussmann fuses in that batch!
Comparing the generic fuses with some from Bussmann and Littlefuse, they all work fine to carry current. That isn’t the problem. The problem is when you feed the fuses 20 A and expect them to clear. A 5 A generic fuse carried over 20 A for a very long time, and, as you might expect, it got very hot. We kept waiting for the fuse to blow, but after three minutes, he gave up.
For comparison, a 10 A Bussmann fuse in the same conditions blew almost immediately — about 350 milliseconds. None of the generic fuses blew, and, in fact, the fuse in the video had been subjected to 20 A of over-stress several times already. It seems like it is nearly impossible to blow them at that current level despite it being four times the marked current. Not much of a bargain.
As the video points out, fuses aren’t as much to protect your equipment as much as they are to prevent fires, so don’t forget to include them even on simple projects. Remember the TI 99/4A? The power supply for that vintage computer has an odd little box in the power cable very near the plug. Why? Because they forgot to put a fuse in until the UL reminded them.
Fake fuses are very much a thing…in a previous job I used to test products for electrical safety. A considerable number of cheaply imported electrical good came with fake fuses which would never have blown if required. Never buy cheap fuses. Always look for Bussman or similar.
That’s a great eye-opener, Tom! I’ve noticed the same thing, but haven’t dug into just how bad the no-name fuses really are. I’ve found this problem with all types of no-name fuses; not just the ATO ones. I had a multimeter literally melt down because its internal 5x20mm 2-amp fuse never blew!
5×20 fuses should not be in multimeters. Maximum rated voltage is 250v when most meters can measure at least 750v. Wrong lead placement and aptented measure of 400vac for example, will literally vaporise the fuse wire and coat the glass in a metal layer and keep conducting. Multimeter fuses are of bigger dimension, ultra fast and with the body filled with sand. Fuses are used to protect against meltdown of the cabling and they depend on the power supply impedance. A 5A smpsu will never blow. The power supply will limit the current.
I would say 5X20should not be the ONLY fuse. They can be the user replaceable, maybe, and often are, but there should be an HV rated fuse internally, as well. Funny, my good meters have that…
They absolutely should. It costs $12 to replace a “proper” 1000 Volt rated fuse each time a kid goes “Hey, let’s measure a battery on the 400 mA setting.”
+1
Another reason it’s just a hassle to mess with power stuff. I like to keep it simple and just run everything on USB. At worst, the fault current is whatever the USB supply can put out, and it keeps things modular.
You still occasionally need a fuse though.
If you get a roll of fusible link wire, you can destructively test a short length of it before putting the rest of the roll to use.
Geez. Now I think I know why the solenoid connector on my brother’s motorbike melted. That 30 amp fuse probably tried to pass 200 and just sizzled away for a while.
Fuses have never been ‘just fuses’. There’s fast blow, slow blow and even time delay fuses. So just dropping a fuse into a project is only doing half your homework. All fuses rely on heat to melt the fusible link inside and I’d say the author chose the wrong fuse for his project. He chose a standard or slow blow 20 Amp fuse. A fast blow 20 Amp fuse would have opened almost immediately after 20 Amps was crossed.
Breakers also have a time delay and a standard bimetal breaker can take even longer than a standard fuse to trip or open. This is because a bimetal breaker also requires heat to open. Magnetic breakers are fast but you never see magnetic brakers for low voltage operation.
The only advantage a breaker has over a fuse is a breaker can be reset. For speed, you need the properly rated fast blow fuse.
“It seems like it is nearly impossible to blow them at that current level despite it being four times the marked current. ” They were supposed to be 5 amp fuses, which should have blown at some point no matter what.
the slowest fuses should open at 4Xrating in about 2 to 5 seconds (depending on class). Typical autootive at the 5 to 30A rating a second or less. Yup. This is bad
I think the proper term for those is cheap Chinese s&$t. But don’t blame the Chinese, they are happy to manufacture whatever we are willing to pay for including that high quality cellphone you are using to read this post.
So it’s really us and our spending habits that are the problem. Us Americans can’t pass up a ‘good deal’ even if it’s suspect. Isn’t this why you tested these fuses? Because the price was too good to be true? If you’d found a deal on BUSS brand fuses then I suspect you would not have tested them unless you thought they were BUSS knock-offs.
Part of the problem with Chinese shit is Western companies not having “boots on the ground” in China to sift out shady subs and spot substitutions, unauthorized cost cutting, etc. And then when shit goes wrong, the suppliers know the Westerners won’t be able to navigate the landscape, spot the same people operating under new names, and bypass logjams in local government.
“Fuses have never been ‘just fuses’. There’s fast blow, slow blow and even time delay fuses.”
+1
For microelectronics, I do use “flink” types (German for quick).
For other circuits, like electric ones, normal types often do.
Additionally, I often add an 1N4007 diode (or a couple in parallel) into the DC power cord, to protect a device from reverse-polarity damage.
You have no idea how often something breaks due to wrong polarity. If it’s my own device, I do install a fuse and a protective diode directly inside the chassis.
1N4007 have quite high voltage loss. I would suggest, at least for higher currents, a P channel mosfet as reverse voltage protection. https://components101.com/articles/design-guide-pmos-mosfet-for-reverse-voltage-polarity-protection.
The better way is to use a unidirectional TVS diode (or a fat Z-diode) after the fuse, making the fuse trip on reversal and overvoltage. Losses are negligible under normal operating conditions.
Don’t buy fuses from China. Lesson learned.
The problem is that China companies can produce high quality, if being told. They have excellent engineers who studied in western vountries. The real problem is us, our European/American business people. They do out-source to China for maximum profits. The problem is extreme capitalism, greed, exploitation. And it backfires. Chinese companies can deliver for the minimum wages that are being paid. Let’s remember the old saying : “You get what you paid for”. Or, “quality has its price”.
It’s and uphill battle. They don’t even replace worn tooling unless told. You start at high quality, and then it goes down over time unless you’re literally there babysitting the process.
Hah, welcome to the world of over current protection.
When designing in a fuse you have to determine what you are trying to protect and trace the current limiting path all the way back to the power source. Note, I did not say overload, I said over current protection, they ARE NOT the same thing. Your design should tolerate an overload up to the point at which the over current device trips. Otherwise you get fire.
You use a UL listed or IEC listed fuse for primary over current protection (preventing short circuits from starting fires.) If you need supplementary protection (like most products that do not have AC powered outputs) you can use UL or IEC recognized fuses.
Each fuse has a datasheet for the family that shows you the AVERAGE trip time based on the current and duration of that current.
Fast blow fuses tend to not have a MINIMUM trip time specification above 400% of their rating, that is, the fuse may trip instantly (<1ms) and still technically be in spec.). That's bad for starting motors, so slow blow fuses were invented, where their MINIMUM trip time specification at a rated current is always specified. The trade off, because they are slower to react is that your design will need to handle the overload that occurs UNTIL the fuse blows. If your PCB traces or is too small of an "ampacity", they will overheat very quickly while waiting on the fuse to blow possibly leading to fire or shock hazard.
IEC and UL have different trip time curves for their respective requirements.
The fuse voltage ratings are specified by the UL / IEC standards, that's why you see 125VAC, 250VAC and sometimes both on the same fuse.
When measuring ANY circuits above single phase 240VAC you better make damn sure your meter is rated for the appropriate overvoltage category you risk an arc fault explosion hazard.
Here's a link to the categories:
https://www.digikey.com/en/blog/what-are-multimeter-cat-safety-ratings?utm_adgroup=General&utm_source=google&utm_medium=cpc&utm_campaign=Dynamic%20Search_EN_RLSA_Buyers&utm_term=&utm_content=General&gclid=CjwKCAjwqZSlBhBwEiwAfoZUIMAFrLPMUyNBUa2jllOlGcBi3yAicRnu5gpBV8H8nIySjhMNEKxuPBoCd6kQAvD_BwE
Now that you've entered the world of specifying overcurrent protection devices, you get to have fun dealing with inrush / starting currents of capacitors or motors weighed against short circuit protection for field wiring outputs.
Yep. Real fuses have a data sheet. It was a real eye opener. Always read the data sheet, folks.
And I do buy my fuses in a traditional 1960s era electronic store located in my local city. A Radio Shack like store, essentially. 🙂
Well, looks like the first real stress test for my DP5015 will be testing some of those fuses I already have.
As I bonus I may be able to learn if my cooling is enough to keep the stock fan from blowing (copper block between the board and chassis = on the other side of the actual mosfets with their stock cooler + fan).
Be sure to then put them on ebay as “pre-tested fuses”
Or pre-stressed fuses? A pre-stressed relay is a real thing…
I got it! Is a circuit breaker and fuse breaker on sale, too? 😂
Ok, ok. Jokes a side. It’s surely possible to use an ohm meter or frequency generator/oscilloscope to measure a fuse. It’s characteristics must be slightly different to a normal piece of wire.
Maybe, if it’s a glass fuse, it’s also possible to use a black light or UV light to examine the fuse wire more closely and distinguish it from an ordinary wire? 🤷♂️
You would need to know what metal was used for the fusible link and ampacity of the link (not the stamped on value but the actual current carrying capability of the fuse). The link for the Littelfuse guide is a very good reference to learn more.
“We don´t buy premium wire or resistors.”
Ah, grasshopper, you have much to learn.
https://www.digikey.com/en/products/detail/riedon/SM15-100KX/4832912
https://www.digikey.com/en/products/detail/remington-industries/64-38LITZ1/14310061
https://www.digikey.com/en/products/detail/alpha-wire/3111265-TN001/5291054
Sure, they are available. But you rarely spec such a thing unless you need something special. I’ve bought, for example, <1% resistors and non-inductive wirewound resistors. But I don't go overboard specing all my 10k pull up resistors.
Personally, I simply buy such good things depending on how much a project means to me. I also don’t use low-end parts if I do build something for a friend or a potential customer. Safety first. 🙂
Exactly. You spend the money where you need to, like on proper fuses. Or you can buy cheap stuff that breaks your product.
Cheap wire may break or have the insulation flake off. Cheap resistors might not be in spec, over a temperature range or at all. Cheap capacitors can ruin a power supply.
It’s no different for any component: if your product will do bad things if the parts don’t act like they should, then you need to make sure that those parts are well specified and properly sourced.
Some parts (like the ones I linked) are expensive, but appropriately so, and you need to pay that premium if you require their specs. Skimping can ruin you.
It’s actually Littelfuse… they weren’t able to trademark Littlefuse since it’s too generic sounding, so they swapped the e and the l.
Bonus reading: https://www.littelfuse.com/~/media/automotive/catalogs/littelfuse_fuseology.pdf
I’m curious about a plot of the voltage and temperature of good fuses vs bad fuses. It might be possible to design a non-destructive fuse tester that can see if a fuse has the right characteristics.
Check the Hackaday article from 20 April 2020…
https://hackaday.com/2020/04/20/test-unknown-fuses-without-destroying-them/
I have seen hobbyists building 3D printers buying thermal cut-outs for heated printer beds via ali express. There have no data sheets, no testing, and no certifications, but hey, you get 10 for a dollar instead of buying one from a reliable source with full specs and certification for $1. Talk about penny-wise and pound-foolish!
there is something wrong with your power supply, probably preventing from overload. Try to connect it to car battery true current-limiting devices or regulators. For now you just heat ip up until it pops with 12W while it can hold 20W
The most effective type of fuse is the old-fashioned wire fuse. Unfortunately, they don’t sell them any more. The fuse consisted of a small ceramic cartridge with a screw at each end. You purchased fuse wire by the roll. It looked almost identical to similar to solder that comes in rolls. The wire, made by companies like Buss, was rated by amperage: a roll of 10A wire, 20A, etc. To “mend a fuse” you unscrewed each end and removed the remains of the old wire, cut a new piece long enough to fit through both screw receptacles, inserted the new wire, tightened the screws, and re-inserted the cartridge into the fuse box.
Totally foolproof, as long as the proper size fuse wire was used. Like the old dangerous practice of placing a penny behind a plug fuse, people would replace the fuse wire with a piece of ordinary wire. Plus, the average homeowner probably doesn’t even have a screwdriver in the house nor knows how to use one. That may be the reason they were discontinued long ago.