-
There's always polyfuses and the like. (if you thought normal fuses behave unpredictably...) -
Of course every component comes with tolerances and some parameters are not specified.
Siba 7017240 does not list any standards this product applies to. Since the curve has not been provided for currents under 4In (the dashed curve 1In to 4In is not part of safety but nuiscance tripping) this product is not designed to be used/relied as a fuse for currents under 4In and it should not be expected to conduct above dashed line. Same applies to voltages exceeding 1kV - it was not designed to operate there. So relying on a parameter of a fuse that lies beyond its design characteristics (a.k.a. abusing a component) is a poor engineering practice.
As for putting a fuse into an application where at advertised parameters it has 50% chance for nuiscance tripping.. It is ok if the advert/user manual states "400mA for 1h, 600mA for 1 second with 1minute cooling, log derating" but it is not ok if someone buys 600mA meter to find out it is useless for 600mA continuous measurements.
Both: overrating a component and operating inside of the tolerance of components, is a problem of the manufacturer, not the customer. -
For the sake of science, I blew a few 30mm fuses from AliExpress (they were sent me by mistake and of no use for me anyway).
1. I put 3A into 2A fuse and closely monitored drop voltage. It was increasing from 610mV to (shortly) 620mV but then dropped to 605mV (or may be power supply warmed up). After three minutes I stopped. I conclude that 150% of rated current it's not enough to reliably trip the fuses I have. So, I increased current to 4A (for 2A fuses), and here is what I have:
7s (probably was weakened by previous tests)
14s
37s
12s
14s
All were tripping fast after achieving ~1.3V drop.
2. Next I took 5A fuses, and run them at 10A. Here are the results:
19s
15s
110s (and was very hot, glass cracked where the wire was touching it)
11s
... And one fuse refused to die even after four minutes.
All 5A fuses except the last one were quite hot. I tried to measure the temperature of the last one, but flir was showing ">260C", even after removing power. The wire inside was visibly glowing. The wire was touching the glass. It seems to me that glass tube works as a support for the wire, that's why it didn't want to die. The glass cracked at the point of contact. I put fuse on the other side, and it finally died after 32s. It got even more shatters, I'm too lazy to take pictures of, the attached picture is from another fuse, but you get the point
.
3. Now 1A fuses at 3A: all blew in less than 1s.
Conclusion: the more current goes through, the faster it blows. And also more predictably.
-
For the sake of science, I blew a few 30mm fuses from AliExpress (they were sent me by mistake and of no use for me anyway).
1. I put 3A into 2A fuse and closely monitored drop voltage. It was increasing from 610mV to (shortly) 620mV but then dropped to 605mV (or may be power supply warmed up). After three minutes I stopped. I conclude that 150% of rated current it's not enough to reliably trip the fuses I have. So, I increased current to 4A (for 2A fuses), and here is what I have:
7s (probably was weakened by previous tests)
14s
37s
12s
14s
All were tripping fast after achieving ~1.3V drop.
2. Next I took 5A fuses, and run them at 10A. Here are the results:
19s
15s
110s (and was very hot, glass cracked where the wire was touching it)
11s
... And one fuse refused to die even after four minutes.
All 5A fuses except the last one were quite hot. I tried to measure the temperature of the last one, but flir was showing ">260C", even after removing power. The wire inside was visibly glowing. The wire was touching the glass. It seems to me that glass tube works as a support for the wire, that's why it didn't want to die. The glass cracked at the point of contact. I put fuse on the other side, and it finally died after 32s. It got even more shatters, I'm too lazy to take pictures of, the attached picture is from another fuse, but you get the point.
3. Now 1A fuses at 3A: all blew in less than 1s.
Conclusion: the more current goes through, the faster it blows. And also more predictably.
Which is why a crowbar seems attractive: it can make the fuse blow faster, but it can never make it blow slower!
-
Chinese fuses have a very bad reputation for being fake. I've seen videos online where they required ridiculously large currents to make them blow. That's why I tested the allegedly SIBA branded fuses that I bought from a Chinese seller. While cruising ebay for fuses, some of them were obviously fake. Either the printing was smeared or printed at an angle, or you could see a copper wire sticking out from the metal cap on the end. The ones I finally bought looked and tested fine.
If you're working on a personal project you can cut a few corners and test fuses like I did. If you're working on a commercial product, the liability issues of using questionable fuses is just too much of a risk. I think Harbor Freight had to recall a kit of car fuses because they were fake and required massive overloads before they'd blow. -
Chinese fuses have a very bad reputation for being fake.
Ok, interesting contribution but I belive this topic is not about fake fuses but about alleged unpredictability of safety-critical component. Both SIBA and DF are reputable manufacturers and "unpredictability" in title is misleading, bullshit and clickbait.
There is no unpredictability in these products, these work exatly as specified in the standards/regulations/datasheets. If there was an error in Littelfuse/ESKA/DF datasheet or fuses were counterfreits or did not meet the parameters - this has not been shown in EEVblog1377.
-
Chinese fuses have a very bad reputation for being fake.
Ok, interesting contribution but I belive this topic is not about fake fuses but about alleged unpredictability of safety-critical component. Both SIBA and DF are reputable manufacturers and "unpredictability" in title is misleading, bullshit and clickbait.
There is no unpredictability in these products, these work exatly as specified in the standards/regulations/datasheets. If there was an error in Littelfuse/ESKA/DF datasheet or fuses were counterfreits or did not meet the parameters - this has not been shown in EEVblog1377.
When I watched the video, I was left with the impression that even the best fuses have an extremely wide tolerance for what is considered "in spec" - in other words, it is normal that individual examples can vary a lot from their nominal specs? It seems fair use of English to call them "unpredictable".
A "fake fuse" would be one that perhaps also varies widely, but not in accordance with the specs? -
Since the topic is 'fuses' I think it's appropriate to highlight the issue of fake fuses so that they won't confuse the discussion. It's easy to create a fuse that looks like SIBA or Littelfuse. Combine that with the wide tolerance of genuine fuses and things get even more confusing. Anytime you're questioning a fuse's performance, you have to consider the possibility that the fuse is a fake.
-
Chinese fuses have a very bad reputation for being fake.
Ok, interesting contribution but I belive this topic is not about fake fuses but about alleged unpredictability of safety-critical component. Both SIBA and DF are reputable manufacturers and "unpredictability" in title is misleading, bullshit and clickbait.
There is no unpredictability in these products, these work exatly as specified in the standards/regulations/datasheets. If there was an error in Littelfuse/ESKA/DF datasheet or fuses were counterfreits or did not meet the parameters - this has not been shown in EEVblog1377.
Yes, this is normal- especially in overload conditions.
I spent some time years ago whilst getting deeper into short-circuit calculations (400V AC, threephase systems) and read the german standard (VDE 0636) quite intensely.
Normal fuses for line protection allow for about up to 45% of their rated current as an overload that from the standard can be tolerated up to 1 hour...
As some metals have to melt and weaken up the copper alloy to make them finally beginning to blow, it is understandable that a fuse that covers the whole spectrum of low overload up to short circuit/crowbar event has some tolerances, which is also reflected in the official electric standards and testing procedures.
In a short circuit condition, the energy in question is so much higher than it is when it is barely above normal operating temp, so the material will literally evaporate at rated breaking points inside the fuse.
When I watched the video, I was left with the impression that even the best fuses have an extremely wide tolerance for what is considered "in spec" - in other words, it is normal that individual examples can vary a lot from their nominal specs? It seems fair use of English to call them "unpredictable".
A "fake fuse" would be one that perhaps also varies widely, but not in accordance with the specs? -
This video was about mA fuses. I was wondering if larger value (80A) fuses used as MPD's for electrical circuits in residential house mains are as unpredictable as these smaller fuses? Are residential circuit breakers more predictable at tripping than fuses?
-
The residential circuit breakers are also not very accurate, at least not the more common ones. There are different characteristics and some are relatively good. Especiall some older type are rather poor and for this reasone phased out / to be replaced.
For a more accurate limit there are motor protectors. These can be quite accurate and some have an adjustable trip level. However they are latively slow and often also large. As they are usually made for AC motors the DC breaking capacity is poor. -
There is nothing wrong with fuses, small or big, mA or kA. All of them are fully predictable in whole design range, as specified in datasheets/standards, if not abused.
Dave took a fuse designed to deal with 4In to 10kA and abused it by heating it to insane temperatures at 2In. Typically such fuses/fuseholders just crack and crumble from overheating, well these just popped gracefuly. If you take a "partial range breaking capacity" fuse (these start from "a" as per IEC60269), you can destroy it exactly same way, by overheating. Mind this is stupid and most likely dangerous, these are fuses to deal with shorts, not overloads, and require overload protection. In residential environment much more often you can find "full range breaking capacity" fuses, (these start from "g" as per IEC60269) and can work in circuits from 0 up to 10kA,..,100kA of prospective short circuit current, whatever design parameter is specified. One cannot overheat them when operated/installed as designed, the list of possible abuses is shorter but still quite long. There are various fuse standards around the world, "residential fuse" is quite a wide term, I am not sure about Australia.
Residential fuses and fuses in general are much more reliable than circuit breakers. I have never heard a fuse to fail. These do trip but always as designed, this is not a failure. As for tolerances - that is another subject and a matter of comparing datasheets. Not to be confused with discussion about a tolerance for abuse - this is yet another, but not very interesting subject. Especially in the context of handheld multimeter protection and by ab-using reputable brands fuse manufacturers as reference of such. IMHO.
If you want to take a look at circuit breakers "amazing unpredictability" - I can recommend bigclivedotcom YT channel. And no, these do not always fail open.
As for abusing circuit breakers - take a look at mikeselectricstuff near-death experience where he was "extreme overload"-ing RCBO.
Abusing fuses - EEVblog 1377 of course, but seriously, this is the job for Photonicinduction! How about 1kV across low rupture capacity 250V CAT V multimeter fuse, wouldn't that be entertaining?
OTOH, I'd love to see an abuse of a proper 10kA fuse in action, @mikeselectricstuff, do you still own the Destruct-o-tron? -
This video was about mA fuses. I was wondering if larger value (80A) fuses used as MPD's for electrical circuits in residential house mains are as unpredictable as these smaller fuses? Are residential circuit breakers more predictable at tripping than fuses?
Reliable==maintaining precision according to the defined characteristics curve?
Thing is that the fuses used for circuit protection follow some very well defined standards. These define some values (and points that are tested in context of the certification of a fuse (German VDE 0636) and the allowable tolerance.
Depending on the company, some companies like SIBA guarantee some tighter manufacturing tolerances on their fuses than applicable standards might allow.
I worked at my last job with SIBA NH and Neozed fuses in an industrial environment and was quite happy about them.
(Mostly used for (household) circuit protection are class gG/gL fuses)
Circuit breakers are a bit different. When we talk about the standard RCCB, you have two triggers implemented: a bimetal one for long lasting overload, and an electromagnetic one for short circuit currents. Those have also standardized characteristics, and the tolerances especially with slight overload is quite high.
Of course the standards also account for manufacturing tolerances.
In my experience, the big issue is with only slightly higher currents than the rated current- there the mechanical elements have lots of tolerance compared to electronically controlled switches like motor protection switches etc.
In case of extreme high currents I have seen some fast stuff- we did the experiment some years ago: In a new room in one of our datacenters, we wanted to test the freshly installed UPS systems. So we initiated a crowbar event (triggered via relay) on some circuit that led to the (future) server racks and measured the reaction via a ION 7600 (or 7650, don't remember).
RCCB was an Eaton PXL C16, and with about 800 A measured short circuit current the breaker completed the whole cutoff in about 3 Milliseconds.
-
There is nothing wrong with fuses, small or big, mA or kA. All of them are fully predictable in whole design range, as specified in datasheets/standards, if not abused.
Dave took a fuse designed to deal with 4In to 10kA and abused it by heating it to insane temperatures at 2In.
What's your point? It's not like he took an IC with a Absolute Max Vcc of 6.5V and fed 12V in to it. These are fuses used and found in multimeters. These fuses are not in place to protect against a short-circuit fault inside the multimeter causing a huge current to be drawn from some power supply. These are to protect the multimeter from being damaged by the user attempting to measure a current beyond the limits of the meter. In such a situation, 1In, 2In, 4In, 10In are all broadly equally probable situations multimeter fuses will find themselves in.
It doesn't matter if it's your opinion that feeding 2In or 1.1In to a fuse is "abuse". Such a label is of no interest when these are the sorts of things multimeter users will be subjecting their fuses to. As such, it's fascinating to see why a 600mA fuse is used for a 1A jack, and it's fascinating to see just how much the fuses vary within the broad bounds/guarantees set by their respective datasheets.
Having enjoyed the video, I find your rant to be completely divorced from the reality of what a multimeter is subjected to in real life. If such conditions are "abusive" by your definition, then multimeter fuses get "abused" in real life, and it's interesting to see what the consequences are. Again, what's your point? -
I use fuses to protect my devices from fault currents. My biggest concern is a fault condition when the current is enough to destroy the device and set it on fire, but not enough to trip the fuse. That's why I try to use tight current rating on fuses. But, apparently, even this doesn't guarantee anything. Besides that, a hot fuse melting the plastic around is a source of danger by itself.
I guess a smoke detector is a good addendum to my lab. -
I have a smoke detector in my lab area. So far so good... it has never triggered, and doesn't seem to mind soldering! -
I have an off grid setup with a 60 Amp charge controller. My solar panels occasionally get to near that 60 amps. I chose a fuse a bit above this 60 amps just in case something shorted somewhere and I chose an 80 amp fuse. The fuses I first had were the automotive plastic fuses with two bolt on tabs. I have decided these are dodgy as at around 50 amps the 80 amp fuse plastic body started to melt, smoke and go black and deform. I now have some far more expensive ceramic bodied fuses with a little glass window to see if the fuseable section has blown.