Circuit Breakers vs. Fuses: Some Hard Fact |
||
| Although I don't want
to defend circuit breakers ("CBs") over fuses, per se, there are many broad
assertions on the subject (e.g., "Fuses are more precise, faster acting, and
more simple (sic) which increases overall reliability many times...").
These assertions are supported neither by empirical data nor testing,
so I decided to find out for myself. The problem with comparing CBs and glass filament fuses is that the original engineering specs used for fuse selection in our aircraft are apparently not available. If you try to order a fuse from Fletchair, for example, you'll be told to buy one from Radio Shack--which is exactly what most people do. So, comparing the Radio Shack fuse to a Klixon CB (for example) is impossible because one doesn't have any reference performance data for the fuse. (Besides the voltage/current rating, the only marking I could find on Radio Shack fuses is "Made in China.") The assertion that fuses are "more precise" means, I assume, that CBs have a wider activation range. It's worth observing that it's possible to select a CB whose activation ranges are identical to that of a fuse. (There are some exceptions, but they're micro-amp fuses, or other form factors not relevant to our planes.) Manufacturers' data suggest that fuses have a faster response time, but either device's response time is suitable for most aircraft applications. More on this later. Using a programmable, recording constant-current generator (borrowed for a long weekend from a corporate laboratory that shall remain nameless), a friend and I conducted some casual (but systematic) tests on high-quality Klixons circuit breakers vs off-the-shelf fuses made by Littlefuse, the world's number one maker of fuses (or so they claim). Our goal was to test a fuse widely available to aircraft owners against a Klixon CB with similar specifications. For fuse engineering data, we used the data from Littlefuse. We selected "Aftermarket Products Series," because that's what pilots would buy for replacements. The organization of data sheets at the web site was opaque, so we used 318 Series with form factor 3AG, the spec sheet for which is at http://www.littelfuse.com. For Klixon engineering data, we used http://www.ti.com/mc/docs/precprod/docs/acb.htm, which is Klixon's (i.e., Texas Instruments) engineer's selection page for thermal circuit breakers. There's a table on the left giving all the "series" of Klixon CBs. Clicking on any series provides all the performance data for that series, including a table giving its thermal/time/current performance. (An explanation of the terms used is at http://www.ti.com/mc/docs/precprod/docs/tech-notes.htm#cb.) We tested four values: 1, 5, 10, 15 amps. Tests were performed at an ambient temperature of approximately 65F. We tested how closely each device's activation point matched its rated value, and how fast it reacted to an application of current in excess of its rating. Our results show that CBs and glass fuses with similar specifications perform very similar in typical aircraft usage scenarios (see next paragraph). The only remarkable difference was the surprising phenomenon that I refer to as "thermal memory"; that is, when the devices were repeatedly cycled to 80% of their rated currents using a ON/OFF duty cycle of 20/10 minutes, the average activation current for the 1- and 5-amp fuses decreased PERMANENTLY while the equivalent CBs recovered to rated value. Practically, this means that in a stressful situation, the current rating of these fuses decrease slightly each time the fuse is subjected to a current near its rated value. Although we didn't perform enough testing cycles to determine whether this behavior eventually "bottoms out," we conjecture that this might account for the occasional unexplained blown fuse we sometimes find in our aircraft at startup. Due to the 0.100 second resolution in our testing equipment, we were unable accurately to compare the response times of the devices to a massive over-current condition such as would result from, say, a short circuit. That is, when we applied 10 times the rated current, both devices reacted within 0.100 seconds. This time suggests that either device would be adequate in applications where the goal is to protect circuit power wiring as opposed to protecting the circuit itself. MY CONCLUSIONS: Given a circuit where the goal is to protect the circuit wiring... Here are some of the reasons one might prefer circuit breakers over fuses:
Here are some of the reasons one might prefer fuses over circuit breakers:
For anyone who thinks I'm nuts for spending the time to do this, consider
I had both the fuses and the breakers on hand. Moreover, the test equipment
was miraculously programmable, enabling us to run unattended tests on four
devices simultaneously. Since my friend was expert at operating the test
equipment, our total time investment amounted to two hours--at most. |
||
|
Originally posted on the Grumman Gang February 27, 2003 |