EEVblog 1377 - The Amazing UNPREDICTABILITY of Fuses!
Вставка
- Опубліковано 6 вер 2024
- How long does it take for your 400mA multimeter fuse to blow at 600mA?
Grab a chair and watch!
The amazing unpredictability of fusing current ratings at low overloads.
Forum: www.eevblog.co...
Subscribe on Odysee: odysee.com/@ee...
EEVblog Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
EEVdiscover: / eevdiscover
Support the EEVblog through Patreon! / eevblog
AliExpress Affiliate: s.click.aliexpr...
Buy anything through that link and Dave gets a commission at no cost to you.
Donate With Bitcoin & Other Crypto Currencies!
www.eevblog.co...
T-Shirts: teespring.com/s...
Very useful to know! Thanks for doing the tests. I never considered thermal hazard from a near-limit fuse.
I think this is actually the limit for UK plugs- the outlet/wiring itself is rated for 20 amps, but the heat source of the fuse has to be limited to 1W so you get 13a, the fuse won't actually blow at all till 20a, but you're not allowed to draw more than 13a continuously, due to heat production in the fuse-13a is the point where it produces more than 1w of heat, and that's thus the limit.
The plugs often come with a sort of paper stencil over the pins to show the correct wiring, it is important to remove this before use, or the fuse might catch it on fire in the event of fault.
This is also the case for circuit breakers. A short while ago there was a video from AvE where he was confused with that. This topic is often overseen electrical-engineering-portal.com/lv-circuit-breakers-essence
I have burned my fingers many times changing just blown fuses for the house.
I don't think it is hot enough to light your own poo on fire though...
I've seen PCB mounted fuse holders melting. The fuse was 5x20mm 10A slow blow,, sand filled ceramic tube type, the fuse holder was rated for 10A as well, and the nominal current was 9.2A trough the fuses. A lot of them started to fail after ~6 months.
Dave: _Hi! Just a quick video…_
Me: _Looks at the duration…_
yep... just set playback speed to X2 ;D
guess im sleeping 34mins less this night :D
lmao I had same thought
I immediately paused and opened comments! Dave at his usual.
Quick for him maybe, not so much for us! Implying that is a bad thing though.
I always keep in mind Murphy's Law of Fuses: A $300 picture tube will protect a 10 cent fuse by blowing first.
That's very much the way Apple builds Macbooks. (Well, that and putting power lines on the connector pin next to data lines that go straight to the CPU with no ground pins in between, so if water gets in there it shorts power to the CPU instead of to ground. Burn the CPU to save the power circuit's fuse...)
Apple products cost a bit more where this happens 🤣. Ask Louis Rossman about that.
@@evensgrey , I remember a checker I had as a kid pushing a pen for the anti-missile system. I told him about a sheet I had which showed all the diodes on the relays were backwards. He told me, draw it as shown. If they catch it, we'll get to redraw the sheet. Holy shit.
WJAT HOW
Used to be common joke among DEC CEs; "transistor-protected fuse".
For electrical installations here in the UK, overcurrent is considered separately as overload and fault current. For a multimeter, the usual reason for the fuse blowing is when you forget you're on the current range and you decide to measure the voltage of your car battery. This is when the fuse's I2t rating is important as it must limit that burst of energy before it fries the A/D converter or whatever components are unfortunate enough to be in the way.
I'm glad you were patient enought to run these tests, they've been quite an eye-opener.
The fuses are there to protect the current measurement shunts, the over-range protection diodes around the uA/mA-range shunt resistors and the PCB traces. As long as those survive, there should be no way for sufficient voltage to make it out of the front-end and damage anything in the back-end.
I've ALWAYS decried fuses in favour of more sensitive and accurate electronics to protect equipment. Fuses are the path of last resort for the cheap minded. Cheap = prepare to lose your shirt because of bean counters.
Just TWO laws of any worth rule in any engineering field: Murphy's and Sod's. They are NOT the same, contrary to popular belief.
The former simply says: "If it CAN go wrong, it WILL"!
The latter says: "WHEN it goes wrong, it will happen at the MOST inconvenient time, cause the MOST damage, MOST expense to resolve and the MOST frustration"!
In over half a century of amateur and professional electronic practice I have never known BOTH to not be true together.
Simple example: "a 1 1/4" glass fuse will be fine!" Result? Fuse blows, the resulting plasma shatters the glass, explodes over the circuit board and sprays molten metal across the board some of which is not noticed during repair. After repair, such "hidden" metal vibrates loose and eventually causes even greater damage. This time damaging the circuit board itself to not just char but be blown through. DOH!
@@boblewis5558 Doesn't Murphy's law apply to MCB's with their moving parts waiting to corrode and seize after a period of inactivity? A glass fuse could sit happily outdoors in a cloud of salty acidic mist for ten years and still do the business when needed.
I share your confusion when I see newly-installed large expensive switch fuse boxes in commercial installations. They can cost more than £100 and be rated at 40A. No-one has been able to explain to me why a £3 MCB wouldn't have sufficed.
@@boblewis5558 , what's your favorite method of protecting circuits? have a favorite active over-current protection chip?
@@somedude-lc5dy nope. Take each case on its own merits and requirements but as an example:
I used to design linear DC supplies with foldback current limiting. Extremely large 70v to 80v electrolytics would frequently be used on the output.
Because of the foldback limiting there was absolutely zero requirement for any separate inrush (outrush?) limiting on the output. However the input side for high output amperage also required high value 60,000uF, high voltage 100v input capacitors that typically DID need inrush protection AND overvolt protection.
Now the classic solution utilised is to use an NTC on the input and a MOV across the input AFTER a suitable fuse, usually HRC type. That's the cheap way but consumes power and does not auto reset nor allow for simple manual reset on overload.
It typically means blown components blasting themselves over the board and sometimes all over the enclosure. Using a power resistor to limit inrush current is better but only during initial conditions, thereafter it's worse than an NTC. A less destructive protection solution is to use a suitable relay to short the power resistor after 30-50% charge on the caps is reached. The relay can be a pulsed toggle one taking zero power except whilst switching. Yes they exist, are only very slightly.more expensive than their more standard variants, and offer other advantages like enabling a thyristor crowbar input overvoltage and overcurrent trip that uses the relay (and or another) to isolate the input mains, the transformer output to the primary DC input supply and potentially to be either auto reset (more complex) or manually reset.
One such supply I designed could be shorted directly across the output terminals, arc welding a copper bar to them, with no adverse impact on the PSU - linear remember - whatsoever! After removing the short the output immediately returned to normal withr the supply able to remain in the shorted state indefinitely with no rise in temp, in fact it would cool down in that state.
Similarly shorting the DC input side was rendered safe via the input relay, as was any over voltage measured either at the mains input or the rectifier output. With some other minor tweaks even the transformer and bridge rectifier were protected and to a level of precision and speed completely impossible using fuses. Fuses are a cheap (all senses of the word) solution to a problem requiring a different level of thinking for quality of product and its electrical protection. MCBs are little better and certainly for their cost, inferior.
Not may channels on UA-cam where the audience agrees with: "that would be a fascinating video, tour of a fuse factory". :)
Kudos for sticking at this Dave! The SIBA was crazy! This video would have blown my mind back in the day when I thought a 300mA fuse would blow the instant the current reached 301mA. :D
You'd some sort of digital over current protection circuit for 1mA of precision.
@@simontay4851 and a fuse to protect it.
but doesnt that make sense?? isnt supposed to blow at 300!! what about the poor multimeter internals?!?!??
An older electronics engineer once told me more or less that lesson:"the only thing defined about a fuse is, that it should not blow at rated current"...
Another teacher told me, that it is quite futile to hope for a fuse to protect your semiconductors. It's only to prevent them from going up in flames. Because when your semiconductor has already burned out, the fuse is still in the process of getting it's temperature up... (compare a bond wire with a fuse wire, and you know one reason).
You know what they say "Anything's a fuse if you're brave enough".
And a fuse is also a shunt resistor as well! I have seen a simple circuit that compared voltage on a glass fuse with a reference in order to decide when to trip electronic protection. It appears that on low currents (like below 1/2 of the specified sustain current of a fuse) the resistances are stable enough for many applications and the fuse itself acts as an additional line of defence to protect the wiring in case the electronics does not trip.
@@Kirillissimus That's a pretty neat idea, actually!
And every circuit has a fuse, whether you put one in or not.
I always figured that fuses really are just to protect against shorts and near-shorts. Perhaps similar to a spark gap...only for when things are going really wrong. It seems that the trip tolerance is even worse than I expected!
I think that the main reason to use them is low price and simplicity. Anything that needs more sophisticated protection seems to get a breaker these days.
They are mostly there to prevent other wires from overheating and causing fires or sparks.
Right. Actually more likely to damage the meter itself with a fuse than without since as soon as it blows the sense amp sees the open circuit voltage across itself, which can normally never happen. You could protect it I guess with a high impedance voltage clamp at the cost of attenuating the input.
In power electronics, a blown fuse usually means an active device has shorted to protect the fuse.
That variance is insane. Much more than I expected.
Not enough smoke was produced during the making of this video.
Perhaps that's the real reason for the variation between them - differing amounts of magic smoke placed within the tube
That Siba was marked FF - this would be faster than normal fast blow (Frickin' Fast ?)
Only at >4In
F and T are German - Flink and Trage. M is Mitteltrage
While I have you both here. Do you know where I can get fluke parts in the UK. I need the input terminals for my fluke 89v, I can only find one selling 1 all in from the usa it's 60 squid fo a £15 part
@@UberAlphaSirus Contact Fluke?
@@perfumedmanatee6235 I would guess it means "Fast and Furious" for not having any arc suppression inside but who knows.
It’s one of those new Schrodinger’s quantum fuses; they’re both blown and not blown at the same time. It’s not until you measure it on the continuity/diode-test range on your meter that you know for sure. 😉
Except that when you measure it, there is a small probability that measurement alone will blow the fuse.
I think the SIBA datasheet might show maximums and minimums. Note the directions of the arrows. I think they are saying they guarantee that the time will be below those lines from 4-10 times rated current, and above the line from 1-4 times rated current.
And minimums might be valid for all sizes of fuses.
Wife: What are you doing ?
"Blowing fuses"
Very useful information! I've done some work with fuses in appliances, but I've never thought about this. Thank you!
One thing I learnt has also to do with "available current": If you design your fuse to protect your stuff, it's good to have a lot of current available. If you have a severe failure after the fuse, you're going to be at a very high current, which causes the fuse to trip really quickly (when you get up to that >3In).
Another interesting thing is that fuses sometimes will actually conduct a small amount of current after tripping. I've had loads of fun testing this where I used a fused voltage supply line both for power and for a signal. When I turned everything else off, the signal went high. I was very confused.
You should do a few more and plot a histogram like you did once when testing resistor value distributions in an old video. Because of the thermal physics invovled the time to blow will not be gaussian... Should be interesting to see!!
Yes, would be fascinating. Would blow (pun intended) a few hundred bucks in fuses though.
@@EEVblog that would be a cool video!
@@EEVblog Could you do it with a roll of fuse wire?
02:14 AM and watching this for insomnia. Now I'm ready for bed. Hope I don't have nightmares about waiting for fuses to blow.
I've seen many fuse holders on PV installations melted because of the heat that they dissipate. Thank you for this valuable information.
Open an online casino! EE nerds place their bets on when the fuse will blow. Huge money maker right there! 🤑🤓🤣
You could put a bunch of fuses in series and let people bet on which one blows!
@JM Coulon resistors in series, First is.
Big Clive did Resistor Roulette!
@@bigjd2k A second game! Progress on the casino. 🤣
excellent idea, i will be a habitual gambler
Fuses were one of the biggest eye-openers when my team started to develop intrinsically safe hardware for explosive atmospheres. The standard (IEC 60079-11) automatically assumes 1.7x the rated current can flow continuously. Then, if you want your fuse to be used as a current limiting device on which safety depends you need to multiply the I_rated * 1.7 figure by an additional factor of 1.5. So the current assumed to flow through the fuse continuously is I_rated * 1.7 * 1.5. For a 400 mA fuse this is 1.02 A. It can make it really hard to meet the power derating requirements of all the downstream components when 2.55 times the rated current can flow. No parts can be run at higher then 2/3 the rated power if safety depends on that component.
It's a bit extreme, but you are trying to limit ignition energy so your product does not become a detonator in an explosive atmosphere.
This video is a great illustration of why the standard does not trust fuses. Thanks for sharing.
WOW learnt a lot today! Thanks Dave. An interesting follow up video might be on how to design a circuit that will provide a way for a fuse to blow more reliably at around that 130% load where a high fuse temperature might melt your product, without the cost of the components/design being excessive.
Thanks for investing the time on this one Dave. Finally I have something to point clients to when trying to explain why designing fusing strategies is so hard!
I'm sure that the line of the SIBA fuse is showing minimum time for low currents and maximum time for higher currents. The triangles are inverted after all. The minimum also seems to be the same for low and high rated fuses which is why there is only one line.
Nice to see LIttlefuse provide good tech data. I recently designed a triac into a piece of hardware, and LIttlefuse's triac literature was very helpful!
I had never considered the voltage drop across a fuse. Now that I know that they can drop several volts I will have to reconsider my use of them everywhere. I cannot have a 3V drop on my 24V circuit. A very useful video.
Amazing. Couple of years ago I designed a custom UPS for video surveillance and network equipment. To keep it simple and safe I used regular fuses on each line with according nominals.
And it's only now that I realized that when I got some "mysterious" voltage drop due to those fuses. It actually worked just fine, but I could not explain a weird difference in voltage between UPS meter with the one that the router shows in its software.
"Trap for young players" as Dave says:)
Dave, I asked the fuse question in your forum at the 14.02.2021 (Frank987). This video is the best answer ever. Super big thank to you.
Another thing to consider with fuses apart from current/temp is fatigue, I have replaced many fuses that have blown purely due to fatigue. Which I believe is due to heating /cooling cycles.
Thumbs up - great subject! A lot of people think the fuse is used to protect the product;) Even a fast fuse is slow and sloppy. The fuse is used to take the product off line after it fails - not to protect it but to keep fires from starting after the product has failed.
Since you worked making test jigs, I would find it very interesting to see such a project, not so much to know how fuses are tested but the proccess involving the develop of a jig.
I would love to see the design process even if it is on paper and is not actually done.
(Sorry for my poor english).
this test seems like a great candidate for a little SCPI test automation
GPIB ;)
Fuses with straight elements are prone to thermal fatigue, blowing from repeated heating and cooling, even though there's no overload. Note how cars switched over to micro fuses with U-shaped elements. So 600mA repeatedly on off would be an interesting test.
Very surprising results. I'd love to see more science on this. It's yet another rabbit hole!
The voltage drop is so high that it could be used as a current sense resistor and trigger a SCR or a transistor to force it to blow.
It's like those machines that have a switch that activates a robotic arm to turn the switch off again
I know there are designs that use fuses as precision current shunts, as the resistance is very well controlled for such use.
Dave, this was great! So many people have such poor understanding of this topic!
Thank you for the informative video!
I had an Audio Amp that had High-Speed Relays on the output because fuses are slow.
Definitely try to do a fuse factory tour when the world mess ends. Would like to see how they create the wire element to exacting standards.
outstanding video! I always use quick fuses on my critical stuff ...and AFCI breakers in my home -(118 years old ) because wiring has been updated sporadically...its great you brought this to folks attention👍
Fascinating topic - thanks Dave!
I think the huge variability of the cut-out point clearly shows what you mentioned several times: fuses solely exist to protect circuits from catastrophic overloads, when the cut-out time goes below the second(s) range. For any sort of protection (or rather prevention) of low overloads, other components and circuits are required. For example 1-4 times overloads could be handled by a current limiting circuit after which the fuse takes over the protection functionality.
nothing makes me happy like a 36 minute video that starts out with the statement "just a quick video..."
Ya FLIR needs a batterizer me thinks
You should do fuse testing on other components, like resistors, capacitors, pcb traces, mosfets, microcontrolles, that kind of stuff. It would be great to see these components tested to see how much current they can actually carry before they "fuse".
I've played this game with 3 phase 480V AC at work with submersible water heaters on a parts wash line. The thing was cobbled together years ago by Jim Bob out in the Jack pines. The fuses are rated at the nominal current for the heaters. So every once and a while, it goes through periods where it eats fuses. The legs all check out balanced and in spec for voltage, resistance, and current. It's just every so often, one starts running hot, blows, and the rest may follow in a month or two. Replace them when needed, make sure the wires and holders aren't melted, and it is fine for another couple years.
Many, Many years ago I worked in a Littelfuse factory for a year.
I was responsible for the IBM system/36 and it's various software...but I walked through the factory every morning and asked questions. One machine was used to apply tiny blobs of solder at regular spacing along the fuse wire element so it does something clever with the metallurgy as it approaches blowing.
That's quite amusing, interesting and a more valuable in-depth info than expected! Nicely done!
With resettable PTC fuses it is the same issue but in addition to that temperature dependance is much worse. Something that somewhat reliably trips under around 1A at 23°C can require 1.5A at -40°C and start tripping at 0.7A at 60°C and in combination with even a simple LED they have a potential to cause output voltage oscillation when they try to trip making the LED blink with a temperature dependant period. Fuses can really be a major pain in the ass to properly put into a design sometimes.
This video remined be of this time I was ask to take a look at a broken car amp, and I was surprised to see that the inline fuse case had melted. I wish i had a thermal cam to show the guy at why it kept melting like this.
As for the amp, the guy ended up using over rated fused (despite telling him it was a terrible idea, but what do I know) and blew the amp and was surprised he ended up with a smoke machine in his car :)
Considering how fuses work, and how cheap they are, I'm impressed they follow any sort of curve at all. If you want something that has accuracy to it, like a protection relay, then you'll need a very different technology, which is much bigger, heavier, and more expensive. Normal fuses are designed to protect against high current shorts.
Nice video, I learned a lot of this stuff out of curiosity when I noticed a pretty big motor controller manufacturer was shipping units with fuses rated lower amps than the amp rating of the motor controller itself.
Can be slow blow types as well, so surges don't blow them.
Depending on the type of fuses, according to standards sometimes a fuse does only cover short circuit and not all the whole range. Therefore it is ok to only give the part of the curve that is relevant (and conforms to standards). In those cases a fuse only consists of a metal sheet- fuses with full-range protection are usually coated with some stuff like tin etc., that will melt with high currents, form an amalgamation with the underlying metal and then it will rip at that point.
I dealt with SIBA in the past, and bought fuses up to NH4a 1250A directly from their german main quarter, so i have some experience with them. SIBA is a highly respected company that also is quite knowledgeable in their trade- and basically is able to give very precise data of most of their fuses (and if you look closely, the curves from most Siemens fuses are verrry similar...)
We used them in a datacenter environment, so we really had to be able to trust them, and yes, we were quite happy with them- mostly NH3 protecting our UPS systems
Sometimes, when applicable standards themselves are a bit wobbly, it is sometimes better not to go into finer points that are not covered there.
I had some looong looks in the german VDE 0636, where common low-voltage fuses for general application are described, and in that area SIBA has quite outstanding documentation.
Also in those official standards there is a wide and big tolerance margin where a fuse is allowed to blow, some 45% above rated current is not uncommon to get it working immediately- and below 13% overcurent usually the fuse has to withhold at least an hour- between that it is left to the discretion of the manufacturer how to produce them, as this range is not standardized.
(At least not in german standards)
Ah the joy of testing! Tedious but you can get so much information. You learn something everyday.
I think that most of the viewers aren't even really certified/qualified to work on situations were the 1000/600V is needed or can be even an issue.
it would be cool to do a video on eFuses (active fet based fuses). I wonder if a hybrid approach would work well in practice. i.e active fet-based current cut-off and fuse for gross overloads.
Yes, a FET + real fuse would be best. FETs fail short.
@@simontay4851 they can fail short but generally the reaction time is microseconds and it’s well within the MOSFET’s SOA. The only reason to add the additional fuse would be if you needed some amount of fault tolerance but even then I’d recommend a different solution. Most battery packs have only eFuses (really discharge control). Fuses degrade especially if the load is drawing high peaks that aren’t necessarily shorts and over time it will oxidize and blow in a nominal setting. Even the NASA fuse de rating guide is complete BS (someone I know tracked down the original writer of the spec who claimed such) which is why we don’t design them into high reliability systems.
For some of us not directly involved in the arts, some of your most useful videos are the equivalent of watching paint dry. But good on you.
For larger fuses I have seen characteristics graphs in VA (volt amps) and time or rated for watts or others.
Very interesting to see the difference between even the same batch of fuses - I once had to spec resistors for an aircraft and in the end it was deemed not possible, due to all the vibrations etc. Even thermal fuses were out due to the pain of soldering them
We work with multi-kilowatt inverters with silicon carbide MOSFETs. A hall-effect current sensor with a 270 μΩ conductor makes a very effective DC fuse when the boost stage controller loses auxiliary power and decides to short the battery terminals. Obviously it's not encased in sand, and vaporized PCB traces will spew molten metal everywhere, but it does disconnect the battery.
hmm, I guess if you wanted to use a fuse instead of an active current monitor, but wanted to avoid the long blow scenario, you could put two fuses in parallel (each of half the rating you're specifying). the positive temperature coefficient should make sure they split the load evenly. this would ensure that the fastest of the two fuses controls because as soon as the first one goes, the current in the other would double and blow it.
When I started my electrical / radio apprenticeship in 1965 fuse ratings were chosen based on twice the maximum current that a device was likely to draw in normal operation. One would expect that a fast blow fuse would blow at or near it's rated current in several seconds. The element was made of copper with CSA (cross sectional area) chosen for it's melting and going open at the rated current. A slow blow fuse element was made of a lead compound which would blow after a little time at that current. The ceramic encased fuses usually meant that the fuse was a high current rupturing capacity fuse filled with fine sand as well as the fuse element which would fill the space left by the element when it blew.
It seems by your experiments current ratings of Chinese fuses mean bugger all. I have found the same thing with Chinese circuit breakers, they pop at much lower or higher currents than they have stamped on them. It seems as in a lot of electrical safety devices coming out of China, you just can't trust expensive equipment to them. Quality control just doesn't seem to exist, I try to use mains and DC C/B's made in countries where you can trust their manufacture has been quality assured. It will cost more but some equipment is worth it isn't it ?
Test circuit breakers yourself before installation, electrical safety means nothing in China or indeed most other Asian countries. I was serving with the RAAF in Malaysia, it was nothing to see Malaysian "electricians" blowing up side cutters by cutting through live cables with some current behind them. I started my trade with BHP in Newcastle , we apprentice electricians were shown safety slides every Friday morning. One story, a young apprentice I remember tried to open up a knife blade 415V 3 phase switch when the contactor (large relay) would not open due to being welded in the activated on position when it tried to interrupt a fault condition on a conveyor drive motor. Over 1000 amps were flowing, the knife blade switch blew up and bits of molten copper flying in all directions ended his life. The more sensible thing to do would have been to open the main OCB ( oil filled circuit breaker) for that sub station, quickly. Most of the mains rated circuit breakers and fuses sold in Australia are made in China, sad is it not ?
High quality Automotive fuses have less vertical curve with sub ohm resistance but its designed for higher currents you should take it in consideration
Very interesting video! Did not expect that they get that hot and still work
ElectroBoom's presentation on the subject was way more mind-blowing when his multi kW nichrome load burning in a bucket of water acted as a fuse before panel breaker at multiple times the nominal current of that panel.
Keep ya coffee warm! Good to know. Cheers for the perserverence.....
I will have it to chill out at nominal current .. good one :)
Really a Fascinating video. I never thought in doing such tests
I choose my own fuse wire size connecting each cell when building 18650 lithium packs. And yes you can only really protect against gross overload. In general my target is usually 4 amps but the lengths are different for each of the cells so they can blow anywhere from 2amps to 10 amps.
Point being there is a wide range for retail fuses. Even a wider range if you try to design in your own fuseable wire. It is fun testing them though and seeing the wire glow red hot but not break.
Siba is partial range fuse for protecting semiconductors (aR) and do not have overload characteristic, is only for short-circuit protection.
Think the standards writers are more worried about fire (in e.g. cable runs) than equipment damage when making specs for fuses.
Try using a frequency counter/timer on the time interval A to B function. Set it to common A and set your trigger levels so that it will start and stop when current is applied and when the fuse blows. This will also allow you to test the fuse's fast acting response time which will be in the mili-seconds.
Take a look at circuit breaker testers and protective relay testers. They can also be used for fuse testing. Companies like Megger/AVO/Multi-amp/Programma, Phenix Technologies and Doble/Manta/Vanguard are a few that sell this type of equipment.
Hi Dave, good job. I am interested in the way that these fuse can help us now. I think the most important thing is how to absorb the over flowed energy not just to shut the current. These ceremic fuse can absorb enough energy, melt itself but not exploded. And in some kind level, they performe as a ptc.
Hi, I have also faced with the same issue a times ago. I need to cut the power about 100mA and search for an appropriate fuse. I bougth a couple of different brand. However none of them was able to satisfy my requirements as in the video. Then, I designed a current limitter circuit with opamps and mosfets and solved the problem.
👍👍 for the landing and Dave's Perseverance!! 😉👌
Please do design the test rig as a video, it'd be great to see your design process for something like this!
Thanks Dave, really love these type of videos where you concentrate on a particular component type!
Even circuit breakers are like this Dave.
Yes problem is seen people running regularly pushing 20a through 16A mcb but after time the trip gets tired. So then keeps tripping at lower current.
What's the mechanism there? I though circuit breakers were based on magnetic fields, so no issues with heat transfer or similar?
@@max_kl depends on the breaker. Magnetic breakers are a subset. For instance take a look at klixon breakers which are thermal
@@max_kl MCBs have a thermal trip (using a bi-metallic strip) for overload, and a magnetic trip for fault currents (shot circuit, the circuit will have been designed to produce a high enough fault current to cause disconnection within in a given time period, 100s of amps). You can have a lot of thermal issues, but generally due to heating from other devices rather than heat sinking. The manufacturer's curves are are only valid at a given ambient temperature (e.g. 25 C), but sandwich it between two other MCBs operating close to their nominal rating for long periods of time and your ambient temperature goes up a whole lot!
Im kind of disappointed you killed the superfuse. That thing deserved better. It was a good fuse. lol
and..... at the going down of the Sun we will remember him.....
LEST WE FORGET
he was a good fuse
he was loved by everyone
Never did anything wrong UNTIL THE END LOL
he shall be remembered
Play at 1.5x and you get PROJECT FARM:which fuse is better?
The makita. Made in the usa. At 20 $ Blows at 10 seconds.
Please! Do a mini-series characterising those fuses. Blow-time vs current and temperature, perhaps burden voltage. Gaussian resistor redux is by far my favourite video, I do love those kinds.
I'm reminded of a "crowbar circuit" where active power measurement is used to short out the circuit and guarantee that the fuse will blow at a carefully measured overload condition.
Just looking at the datasheets of two common multimeter fuses reveals really different trip curves, there is Bussmann B-44/100 and Littlefuse FLU-44/100.
The former is rated at 440mA for 300 seconds, but the latter at 2A for 1000 seconds, that's quite something!
Dave shows the datasheets in "EEVblog #376 - Multimeter Fuse Diode Followup
" which also contains a primer of how the milliamp input protection works.
very interesting Dave, TNX 4 the video!
i replaced my blown fuses with no-blows lol
13:00 - One factor why this fuse lasts so long is because you use 4 clamps on it instead of 2. So way more thermal dissipation on booth ends there. So the whole fuse can heat up slower to a very "shocking" value though. Never thought something basic like this will surprise me.
it would be very great to run a comparison test within resetable fuses like PTC fuse and wire blow fuse
Very helpful video. I liked it
I remember doing a repair where I have put a low value test fuse in a product and forgot...Getting it back years later with the same underrated fuse still there and still intact and scratching my head. The very low current ones I remember have even wider tolerances...Then there are all the different types of time delay fuses..It gets a bit crazy. Above an amp things get a bit more predictable if I recall.
Are you using a constant current source? In most cases the fuse will be in a constant voltage circuit. The positive thermal coefficient would increase the fuse time.
I'm designing a personal project and I'm also testing out the various current rated cheap chinesium glass fuses. The product will deliberately blow these fuses and will need replacing. And I'll also have to build some test jig to test them. I want the glass ones so you can see when they are open. I also want to make the blowing be a bit "explosive" so I have to test out different voltages, currents and time. Sounds fun. I'm in blowing more fuses!
a test jig with a small micro to spit out fuse data would be gorgeous ❤️
In my experience you melt the meter before the fuse goes with constant current over time. That is why it would be smart to significantly underrate the fuse. Although it seems there is a very large variation in fuses. The spread at nominal current is very large.
Fascinating tests, the huge variation indeed. Something to really consider in the design.
Concerning the fuse marking on the Siba. AFAIK those are the DIN characteristics:
FF - Super Fast
F - Fast
M - Medium
T - Slow
TT - Super Slow
Those indicate time until trip at 10x rated current.
nice video Dave! Very interesting to see how much voltage drop you get at high temps! Thanks for the upload good sir! Cheers from Florida
30 minutes(+) vs 45 seconds. Hah! I certainly did not expect such a huge variance.
fuses are used because they are low cost. In electrical circuits like motors, fuses have a good value.
If you need a fuse to accurately perform at low over current situation you might need to design an electronic circuit measuring the current (or the heat) and trips a relay (also less problem with voltage drop). For the massive over current you use a regular fuse.
Siba fuses are typically a lower resistance than many other fuses. Very useful if the burden voltage is significant in your application
They are rated that way because there is a difference in sustained load and spike loads. They are built with in a tolerance range. If you want more exact fuses you can actually make them yourself.
Not exactly something I would call random.
I think these considerations are a good reason why an electronic fuse or current limit is a great supplement to using fuses for real protection. Protect the fuses between their sustaining current rating and the trip point.
This is also a good case for setting reasonable brown out detect/enable levels for switchers; if there is a lot of voltage droop on the input to a supply, there is probably something wrong enough to make it so the power supply shouldn't start up.
The melting energy is very important for fast current pulse. In that condition, the fuse does not have time to dissipate temperature through its pins. A2s figure is more linear and precise than clear curve. It is not an academic data, it is the main parameter to use to protect against current pulse and to coordinate a cascade set of fuses.