-1
\$\begingroup\$

I have a slow-blow 3A, 1.2A2s (I2t), 32V SMT fuse [1206 package] in my design. The power is supplied by the 12V battery in a car. Questions:

  1. How can I verify that the fuse would protect my circuit - i.e., verify that the fuse will blow in case current x time exceeds the I2t limit of the fuse?
  2. In case the fuse I selected does not work well or has a spec mismatch with my design and I need to eliminate it, can I simply short the two pads of the fuse using solder?

Thanks.

\$\endgroup\$
  • \$\begingroup\$ You'd better make sure the interrupting capacity (probably something like 35 or 50A) is adequate for the maximum fault current. Sure you can short it (maybe with a bit of bare wire as well as solder), but you won't have a fuse if you needed one in the first place. That might be a safety issue.. \$\endgroup\$ – Spehro Pefhany Apr 29 '15 at 2:39
  • \$\begingroup\$ Thanks @SpehroPefhany - i am also wondering how I would prove to myself that the fuse would blow in case of a bad situation - for example, a current surge that exceeded the spec of the fuse I am using. How do I generate a controlled current surge to test the fuse? This is obviously a one-off test - on one board, just to prove that the fuse works. \$\endgroup\$ – NK2020 Apr 29 '15 at 3:22
  • \$\begingroup\$ What exactly is the fuse intended to protect? Wiring? PCB? Semiconductors (probably not)? \$\endgroup\$ – Spehro Pefhany Apr 29 '15 at 4:33
  • \$\begingroup\$ @SpehroPefhany: my main concern is wiring - and the risk of fire since I would be installing this in a car. \$\endgroup\$ – NK2020 Apr 29 '15 at 6:19
1
\$\begingroup\$

Fuses have datasheets, which have "maximum interrupting current" specs, as well as graphs of "how soon will it blow given X% over-current".

You need to consider what is downstream from the fuse, what the likely failure-mode of those components is, & thence how much fault current could flow. It could be as simple as a tantalum cap in the PSU that'll go nicely short-circuit, or it could be as fuzzy as something your circuit is powering off-board that, "for reasons", is drawing more than the expected current, particularly relevant if you're powering multiple such peripherals if there's potential for multiple simultaneous failures.

You don't really need to "prove" that the fuse will blow simply by applying some controlled resistance that will result in, say, 150% x maximum normal current, and confirm that the fuse does actually work as advertised - that's not much value. You need to foresee what likely fault currents could be (by considering the characteristic fault modes of those components), and size the fuse accordingly.

You might also need to consider inrush current (e.g. charging caps at the moment of switch-on) to select the right style of fuse, & even what influence inrush current might have on the fuse's amp rating - you don't want a fuse so close to the normal operating current that switch-on inrush current is going to blow it.

If your fault currents are "high" (because some component has a propensity to go very low R when it fails), then the maximum interrupting capability of the fuse becomes relevant, especially for DC voltages >30V. 40+V is the ideal arc-welding voltage, and an arc can be sustained through the fuse even after it's blown. HRC, 'High Rupture Current' fuses are common here. But you're at 12V, so this is not so much an issue.

Yes, if you must, short out the fuse. But then the lawyers come for you...

\$\endgroup\$
  • \$\begingroup\$ Inrush current has been limited using a resistor - there might still be current surges due to voltage surges on the 12V battery, but thats why the slow-blow fuse with a certain I2t rating. So, I have actually spec'd all that out - however, since this is the first time I am using a fuse, I want to see it blow in a controlled situation, to be confident that it would protect my wiring (my main concern is risk of fire). The question is: what would be a good way of generating such a controlled situation where the fuse blows. \$\endgroup\$ – NK2020 Apr 29 '15 at 6:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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