7
\$\begingroup\$

If I would like to use a piece of copper wire as a 150A fuse, what size of wire should I use?

It doesn't have to be 150A very accurately, the parameters I need the wire to meet are:

  • it needs to conduct continuously 70A 10V-30V DC without getting red or overheating
  • it needs to conduct 120A sometimes for 15 seconds without melting
  • it needs to surely be melted before the current reaches 200A

Easily available wires over here are: 0.5 mm^2, 0.75 mm^2, 1.0 mm^2, 1.5 mm^2, 2.5 mm^2, 4 mm^2, 6 mm^2, 10 mm^2. Will any of these do the job? I can combine a few if the required value is not in the above series.


Edit: now I read more into the specifications of the device, and it is in fact rated to be fine with 600A for 5 seconds. Besides that I will be only ever using it continuously with 50A. Is that gap (50A - 600A) large enough to make a fuse out of copper?

This table of AWG wire sizes suggests that a 2.5mm^2 copper wire will melt with about 750A in 1s, and with a little under 200A in 10s, so it looks like it will melt in time for the device not to be damaged. The device itself is wired with 10mm^2 insulated wire, and I mostly worry about that wire not to be damaged.

Now I only need to find out to what temperature will a short piece of uninsulated 2.5mm^2 copper wire heat-up with 50A continuous current flow, and will it not melt the thing it is secured in at that temperature.

By "continuous" use I mean max 1 hour at a time, with full attendance of me, so it will not run unattended like this.


More info: The machine came originally with two 50A fuses wired in parallel. In fact I want to use a wire, because the fuses blow so often it starts to get expensive, on average I need 1 fuse per 1 hour of charging, so the fuses cost more then the electricity to power this. I don't know why they blow, because I have an ammeter wired in series with the fuse, and I never seen anything over 60A on the ammeter! The fuses don't blow randomly, I just can see the fuse slowly get red, it stays red for some time, and at one point it just melts. I've been watching this process and I didn't see over 60A while the fuse was melting, I was watching the ammeter all the time in the slow process of the fuse being melted. So if I need 1 fuse per hour, so be it, but I need some cheaper option then 50 cents a fuse if I am to use them at that rate.

\$\endgroup\$
  • 5
    \$\begingroup\$ @miernik, the reason we purchase fuses is so that a fault event is quickly interrupted, you need to make sure copper will flash off fast enough to protect your load/source. \$\endgroup\$ – Kortuk Mar 29 '11 at 18:33
  • 5
    \$\begingroup\$ @miernik, why not a fuse? As Kortuk states, that's the safety function they perform. Wire...dodgy at best and it will not breakdown fast or reliably enough IMO. \$\endgroup\$ – kenny Mar 29 '11 at 18:39
  • 8
    \$\begingroup\$ If I was on a dessert, and it was so big that there wasn't a place I could get a fuse nearby, I would start eating it, and not worry about electronics for a while! \$\endgroup\$ – Kevin Vermeer Mar 29 '11 at 21:38
  • 3
    \$\begingroup\$ I think you should look at why the fuses are blowing since they are supposed to be OK. Are you sure that your ammeter is reading correctly (e.g. AC vs. DC, high-frequency variation, ...)? Are both fuseholders actually in the circuit (does only one of the fuses ever blow)? \$\endgroup\$ – Kevin Reid Mar 30 '11 at 23:16
  • 5
    \$\begingroup\$ Don't use copper. It's thermal conductivity is too high, and as a result, everything near the pseudo-fuse will get hot when you approach the fusing voltage. Something like steel or another metal with poor thermal conductivity will work much better. \$\endgroup\$ – Connor Wolf Mar 31 '12 at 0:30
14
\$\begingroup\$

Finding a fuse with those characteristics is going to be difficult. I'd use a current shunt with a suitable MCU and circuit breaker and monitor the current taken by the device, shutting off the current if it rises above 120A for 15 seconds.

Two fuses in parallel can cause problems unless they are well-matched and the holders are properly designed. Any difference between them can cause one fuse to take more current than the other one, and fail.

\$\endgroup\$
8
\$\begingroup\$

It sounds to me like like you're about to violate some NEC/NFPA* fire and safety regulations. Replacing a fuse like this with bare copper wire is ILLEGAL period. Machine downtime is always preferable to property damage, injury or loss of life.

EDIT * Assuming you're in the USA.

\$\endgroup\$
  • 5
    \$\begingroup\$ Well, but a fuse is just a piece of wire. OK, its probably something else then copper, but still its just a wire, just of the right size to melt at the right current. Fuses are not God-given and do not fall from the the sky as miraculously self-created objects from a secret source - no, they are made by humans, out of some wire. So I am asking "how to make a 150A fuse". \$\endgroup\$ – miernik Mar 29 '11 at 20:42
  • 4
    \$\begingroup\$ @miernik, how to make a 150A fuse and how to make one from copper are different, you are asking something more specialized. @chris is worried you are going to violate laws or safety. Copper is not ideal as it will not enter thermal runaway as well as devices design for this. Manufacturing fuses is still quite a bit of work. \$\endgroup\$ – Kortuk Mar 29 '11 at 22:19
  • 8
    \$\begingroup\$ make sure you put a glass or ceramic envelope around your copper wire. Copper melts at 1085 °C and I have scars to prove it. There will almost certainly be a fire if anything flammable is near your melted wire, or underneath where the molten copper splatters. \$\endgroup\$ – markrages Mar 30 '11 at 5:13
  • \$\begingroup\$ @miernik I guess I'm really just concerned about your application. If this if for home, hobby, prototyping some fuse design then cool. However your comment "I've got. Assume you are on a dessert, far from shops with fuses, you have only wires, and not running the device until you get a proper fuse is not an option. What size wire you would use?" \$\endgroup\$ – Prof. Meow Meow Mar 30 '11 at 12:28
  • 7
    \$\begingroup\$ The key difference between wire and fuses is that fuses are designed to fail above a specified current, whereas wire is designed to work up to a specified current. These are not the same thing. A copper wire will likely work well past its "rated" current, with a large tolerance in current levels at which it fails. The fail-open current for a fuse, while not perfect, is much better controlled, and as others have pointed out, will not create as much of a fire hazard due to the lower melting point of the non-copper (aluminum?) material. \$\endgroup\$ – Mike DeSimone Dec 11 '13 at 0:21
4
\$\begingroup\$

Forget trying to design your own fuse, leave that to experts whose livelihood depends on safe and accurate designs. Call a fuse manufacturer's technical department and discuss the parameters that you provided on your load (600A 5 secs maximum, 50A continuous).
Fuses protect the load AND the cable feeding the load. Replace your parallel fuse holders with one fuse that is designed for purpose.

\$\endgroup\$
4
\$\begingroup\$

Fuses don't work well in parallel, especially if they are installed in mechanical holders and not soldered. Holder terminals inevitably oxidize, which increases the resistance and affects current distribution.

Now, if there's a single fuse, oxidation doesn't become an issue. Increased resistance results in voltage drop in oxide layer which gets destroyed, restoring the contact.

That doesn't happen if you have a second fuse in parallel: that fuse will shunt the high resistance spot and let the oxide on the other fuse grow. Eventually, the oxide layer will completely isolate one of the fuses, forcing all the current through the other fuse and destroying it.

So, get a single 100A fuse. If you can't, try soldering two 50A fuses to the terminals. Also make sure that each fuse has a significant (and roughly equal) length of wire before the joint: wire resistance will help to equalize the current in both fuses:

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • \$\begingroup\$ It should be noted that R1 and R2 should be rated for 25 W, of course with some derating. \$\endgroup\$ – Uwe Sep 2 at 14:35
2
\$\begingroup\$

Sounds like you're planning ahead for a tragedy. Kinda like, "If a robber comes through the door should I shoot him, or stab him?" Forget that and just lock the door! :)

What you should do with all your planning is simply buy extra fuses and keep them handy. If on the odd chance that you run out of fuses, then kludge something up however the situation calls for it (keeping in mind the safety and legal aspects of it). But at that point, you're not worrying about what amperage the copper wire is going to melt.

\$\endgroup\$
  • 6
    \$\begingroup\$ No, strike that, don't kludge anything up! This is 150A we're talking about here. \$\endgroup\$ – Kevin Vermeer Mar 29 '11 at 21:40
  • 3
    \$\begingroup\$ I once used a bunch of "wire fuses" in series with a "normal fuse" when experimenting with a motor-control project. I had a bunch of turns of wire wrapped around one post, which then went to a second post which had a few wires were wrapped around. Each time the thing blew, I could simply unwrap some more turns from the first post and connect to the second. The wire let through enough current for operation, but consistently blew before the normal fuse. Not something I'd do in an unattended device, but the wire was probably cheaper than replacing the fuse a few dozen times would have been. \$\endgroup\$ – supercat Apr 9 '13 at 17:31
2
\$\begingroup\$

Fusible links in automotive use are normally made with a short length of wire (25-35 mm) that is two wire gauge sizes smaller than the wire supplying the load. For example, if the wire is 12 AWG, use 14 AWG wire for the fusible link. (Convert American Wire Gauge, AWG, to whatever system of measurement you are using in your part of the world.) This will prevent a fire that destroys the vehicle, but the wiring in the system may still overheat and be permanently damaged.

If super-fast response is not required, consider replacing the fuses with an appropriately rated thermal circuit breaker. A thermal-magnetic circuit breaker, such as those used in house wiring, can be used for faster response to overloads. You may find that replacing the fuses two or three times costs more than the circuit breaker and its enclosure.

\$\endgroup\$

protected by Kortuk Dec 11 '13 at 3:27

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.