You are exactly right about fuse tempco, and I should have mentioned that.
Cheers,
Bob
No, it is not, at least if you want to talk apples to apples. Fuses are characterized by the I2t product (at clearing and melting) which is proportional to the energy, not the power.
To talk about power in fuses would mean either considering the voltage drop or the resistance, both depending in a nonlinear fashion on the current.
For resistors, the maximum power dissipation is an average value, while the current surge capability is a completely different matter. An 1206 0.25W 2.2ohm thick film resistor can take say max. 0.3A on average and survive 1A peaks/pulses of say 10mS, while a carbon composition or wire wound resistor of the same value and power can take also 0.3A on average, but survive 5A peaks/pulses of the same width. The "I2t" for the latter resistor type is much higher, if you prefer. However, the resistor may remain linear even when overloaded, since the resistor thermal capacity is much higher than a thin wire in air (the fuse).
This is text book stuff. If one wants to use 1206 resistors as base stoppers "because it's not the end of the world if they blow", then so be it. Bad practices are allowed in DYI.
Last edited:
Codswallop.#
You mention I² twice.
That is effectively power dissipation, because whether it's a fuse or a resistor both have some resistance.
Adding a t just applies some time to the power dissipation overload limit.
Last edited:
Double codswallop. In fact triple, but I'll stop counting and defer to the readers to continue.
Simple single wire fuses melt through excess power dissipation. They get hot, they melt.
Slightly more elaborate types, those with the coiled wire that acts as a spring and that are soldered at one end still rely on heating to open the circuit, but they do so by heating the solder blob on the spring until it melts.
No power dissipation, no blow.
Slightly more elaborate types, those with the coiled wire that acts as a spring and that are soldered at one end still rely on heating to open the circuit, but they do so by heating the solder blob on the spring until it melts.
No power dissipation, no blow.
By the same logic, when the Vcemax is exceeded in a transistor and it blows, it is excess power that destroys the device. By the same logic, any electrical destructive process is due to excess power. Perhaps causally correct, but very limited in view and certainly not useful for understanding the underlying mechanisms.
Resistor blowing, secondary breakdown, etc... are local phenomena, treating them as global doesn't help. You can keep the average (global) power dissipation of a resistor within limits, but still blow it. You can keep the transistor power dissipation within limits, but still blow it. What is the average power dissipation allowed by a fuse? What? It's not in the data sheet?
I'm afraid I failed miserably in conveying this message.
What I am saying is that many devices (resistors, semiconductors) have the power limits defined by at least two parameters: one global (maximum power dissipation) and one local (surge, secondary breakdown). The local parameters depend essentially to the construction and materials (film resistor have much less surge capability compared to carbon volume resistors, secondary breakdown depends on the emitter geometry, providing an as much as possible uniform current distribution), and not by the ability to (globally) dissipate the heat/power. You can drive a dead cold transistor into secondary breakdown.
Not the case for fuses, it's all about "surge", or local phenomena. As I said, there is no "maximum power dissipation" parameter in fuses, and I can't imagine how one could be defined. Therefore, to tell that "fuses blow because of power dissipation" is essentially meaningless.
Fuses blow by localized temperature (typically in the middle, where there is the least heat removal effect from the end caps). They blow when that localized temperature reaches the melting point. That localized high temperature results from the dissipation of power, whether that power is a function of time or area or not. For this reason, it is not unreasonable to say that fuses blow because of power dissipation. It is just a very nonlinear process.
These semantical arguments get boring fast.
Cheers,
Bob
These semantical arguments get boring fast.
Cheers,
Bob
Here's a quick digest. Meantime, a lot of personal dislikes resurfaced.
Just thinking out loud here. Consider guelling test condition, but not a fault condition (like an output short circuit). I think a 1206 is rated for 300mW continuous dissipation. Consider a 20Hz square wave into a heavy load, where peak output current of each output transistor must go to 10A alternately for 50ms at a time. If beta is 50, base current is 200mA, which across a 5 ohm stopper (I prefer 0.22) is 1V and dissipation is 200mW on the 50ms peaks of the square wave. Average dissipation is 100mW. This may be OK on a dissipation basis, but if there is significant beta droop at 10A, things get ugly fast, since stopper dissipation will go up as the square of beta going down.
Just thinking out loud.
For an output short, mush will depend on the type of protection circuit, but the interval of the high current must be kept short by the protection circuit to be safe with the SOA of the output device.
Going to a larger base stopper would seem to be reasonable conservatism, given that with SMT space and cost are not a big issue.
Cheers,
Bob