Fuses do affect sound, the question is how much

[3dbinCanada]

No, voltage drop that varies with signal causes distortion; it creates a nonlinear result. The modulation will be different at different power levels; the temperature is only constant for a given steady-state power (current) level so changing power (SPL) changes temperature and thus voltage drop.

I disagree. The voltage swings linearly with the input signal just like it does across a resistor. There is no modulation. Carbon filament resistors do generate thermal noise called Johnston (not sure the spelling of his name) noise but there is no modulation across the resistor, no phase shift, no crossover distortion.

 

[Speedskater]

Bob Cordell covered all of this way back in 2011, in the first edition of his book.

 

[DonH56]

I disagree. The voltage swings linearly with the input signal just like it does across a resistor. There is no modulation. Carbon filament resistors do generate thermal noise called Johnston (not sure the spelling of his name) noise but there is no modulation across the resistor, no phase shift, no crossover distortion.

Guess we'll just have to disagree. At low distortion levels real resistors are not perfectly linear either. Or at high frequencies. Etc. There is no crossover distortion but there is phase shift and a whole host of other things. None of which matter for audio except in pathological cases. And it is "Johnson", also referred to as Johnson-Nyquist or just thermal (4kT) noise; it happens whenever carriers flow, in all types of resistors, wires, whatever. But that is not what I was referring to.

Bob Cordell covered all of this way back in 2011, in the first edition of his book.

Do not doubt it one bit! Not sure where his book is at home, I have a copy someplace. But I first learned about it decades before when I was measuring power circuits. Fuses get very nonlinear before they actually blow.

 

[mansr]

I wasn't thinking of it tracking the instantaneous signal level. I agree it's a thermal issue with long time constant, sorry I did not make that clear. It does track the average power and that varies with listening level and so forth.

Right, so the effect, such as it is, will be more like dynamic range compression than harmonic distortion.

 

[3dbinCanada]

Guess we'll just have to disagree. At low distortion levels real resistors are not perfectly linear either. Or at high frequencies. Etc. There is no crossover distortion but there is phase shift and a whole host of other things. None of which matter for audio except in pathological cases. And it is "Johnson", also referred to as Johnson-Nyquist or just thermal (4kT) noise; it happens whenever carriers flow, in all types of resistors, wires, whatever. But that is not what I was referring to. .

Do you mean low signal levels rather than low distortion levels? To have a phase shift, one needs to introduce either a capacitance or inductance. Where is this reactive property coming from since a fuse is purely resistive? If your ignoring thernal noise, then what else is there?

I dont disagree with your assertions at the moment of fuse disintegration. I should have clarified my arguements that I'm assuming that operation is well within the fuse's current capabilities. Lots of weird shite happens at the extremes.

I also want to say thankyou for discussing this. I find this fascinating.

 

[DonH56]

Do you mean low signal levels rather than low distortion levels? To have a phase shift, one needs to introduce either a capacitance or inductance. Where is this reactive property coming from since a fuse is purely resistive? If your ignoring thernal noise, then what else is there?

I dont disagree with your assertions at the moment of fuse disintegration. I should have clarified my arguements that I'm assuming that operation is well within the fuse's current capabilities. Lots of weird shite happens at the extremes.

I also want to say thankyou for discussing this. I find this fascinating.

Look, I need to get back to work. Low distortion levels, and/or high frequencies. No passive component is really ideal, so a resistor actually has inductance and capacitance (among other things). And a capacitor has resistance and inductance, and so forth. Same applies to a fuse or piece of wire though for a fuse it does not matter unless it is operated in the signal path and near its limits. And if you start measuring extremely low levels of distortion tiny little parasitics cause problems. Capacitor values vary with the voltage across them, some types worse than others; resistors change with current, temperature, frequency, phase of the moon...

Those parasitic elements are critical in my world, which includes accounting for them in power supplies up to multi-GHz signals. For power supplies, the parasitic resistance in a capacitor limits its ability to deliver charge thus increases ripple and also leads to self-heating that shortens its life. The parasitic inductance in a capacitor causes the capacitor to quit acting like a capacitor, sometimes at fairly low frequencies (kHz for very large capacitors), and so we have to design for that. As you go into RF frequencies it just gets worse (look up things like skin effect, loss tangent and surface roughness, and so forth -- even the dielectric material starts being a problem), and as you reach for distortion levels say -80 to -100 dB and below tiny little things that don't matter become Big Deals.

Get Bob's book on amplifier design as @Speedskater suggested and I imagine he discusses fuses. The problem with fuses is that their resistance changes with signal; slowly, as @mansr said, but it does change. Where I have run into problems was power sweeps when you can readily measure the distortion a fuse adds; at lower current levels, the effects are there but much, much smaller (negligible). At high current (shortly before they blow) fuses act like very nonlinear resistors. I am no expert on fuses, but the time constant depends on the fuse, and I seem to recall milliseconds to tens of milliseconds or more. I don't have the measurement data any more, so treat this as heresy if you wish, but my audio-related example was my subwoofer design ages ago when I found at higher currents a 20 Hz signal or so (50 ms period) would modulate the fuse and increase the distortion I was measuring from the amplifier. I went from a large wire to my load to adding a fuse and noticed it. It was in the mud as far as audibility, but was interesting to me at the time.

Again, I am not claiming any of this is audible, or that it matters in a real-world audio situation, but it is real.

 

[scott wurcer]

And it is "Johnson", also referred to as Johnson-Nyquist or just thermal (4kT) noise; it happens whenever carriers flow, in all types of resistors, wires, whatever. But that is not what I was referring to.

Johnson noise exists with or without current flow, excess noise is a property of particular conducting materials at a given current density and exists over and above that. BTW Nyquist's 1929 derivation is a beautiful example of scientific reasoning.

 

[mansr]

I am no expert on fuses, but the time constant depends on the fuse, and I seem to recall milliseconds to tens of milliseconds or more. I don't have the measurement data any more, so treat this as heresy if you wish, but my audio-related example was my subwoofer design ages ago when I found at higher currents a 20 Hz signal or so (50 ms period) would modulate the fuse and increase the distortion I was measuring from the amplifier.

That sounds plausible. A fuse with a larger time constant, perhaps a slow-blow variant, might have produced less distortion. If I cared enough, I might rig up an experiment, but I don't.

 

[Cbdb2]

I disagree. The voltage swings linearly with the input signal just like it does across a resistor. There is no modulation. Carbon filament resistors do generate thermal noise called Johnston (not sure the spelling of his name) noise but there is no modulation across the resistor, no phase shift, no crossover distortion.

Resistors have a temp. coefficient that can cause distortion. Its small and many manufacturers spec it. You can buy low TC resistors. They change resistance with temp. so they can change resistance with signal and this can modulate the signal. The amount of distortion is very dependent on the circuit, amount of power, and the signal. Not possible to sim without heat transfer modeling. That said, its always very very small and inaudable. The difference with a fuse is there designed to melt so the temp goes up rapidly when close to there limit so the temp. coefficient also increases rapidly. So This will cause more distortion than a resistor, but still inaudable. As mentioned before its been measured at .0033% probably worst case, just before the fuse blows.

 

[DonH56]

Johnson noise exists with or without current flow, excess noise is a property of particular conducting materials at a given current density and exists over and above that. BTW Nyquist's 1929 derivation is a beautiful example of scientific reasoning.

Yah, why I said "carriers" not "current" flow... I have worked with some superconducting circuits and it gets all squirrely -- theoretically no noise when carriers stop moving ("vibrating"), and we were out to plot the curves, but measuring thermal noise in the single-digit Kelvin range is something I never hope to have to do again. In any event I was thinking of "carrier" (e.g. electron etc.) movement that happens even without applied voltage (thus no current flow).

Edit: "Flow" was a bad choice of word on my part and led to the confusion, sorry!

I saw the Nyquist derivation many years ago but honestly do not remember much of it. One of my professors had all of us read some of the fundamental papers that started it all; it was actually an interesting set of assignments.

At this point I have no desire to delve deeper into the weeds.

That sounds plausible. A fuse with a larger time constant, perhaps a slow-blow variant, might have produced less distortion. If I cared enough, I might rig up an experiment, but I don't.

Me neither. I could measure it, and it was a vexing if interesting exercise to find the source, but ultimately negligible as far as hearing it was concerned. Especially in a subwoofer! I remember people for a very brief period adding little heat sinks clipped onto their fuses to better dissipate the heat and supposedly reduce the distortion. One of those audiophile tweak things. I do not recall making any measurements showing that to be true -- all the action was inside the fuse, and they always melted in the middle since the end caps would keep the ends a little cooler than the middle. You could super-cool them with dry ice and raise their melting point a tiny bit.

 

[tomelex]

From previous posts, the temperature of a fuse remains fairly constant once warmed up. Voltage drop does not cause modulation and does not cause distortion. That is an incorrect assertion. I whole heartedly agree that putting a fuse in line with a speaker is poor design.

And yes, though, that is what this thread is about, and it was common back in the day, and speaker uses worked, lord knows I blew a few in my time. I hope you can agree that speaker fuses cause distortion after reviewing the posts in this thread. It is science and we as a team brought it out into the light. that's what I love about this site, facts and a place for technical folks to gather and also for those that want to learn to come to as well. Worth donating to IMO for all the quality information provided as opposed to touchy feely stroke your members sites ahahahahahah

 

[3dbinCanada]

Resistors have a temp. coefficient that can cause distortion. Its small and many manufacturers spec it. You can buy low TC resistors. They change resistance with temp. so they can change resistance with signal and this can modulate the signal. The amount of distortion is very dependent on the circuit, amount of power, and the signal. Not possible to sim without heat transfer modeling. That said, its always very very small and inaudable. The difference with a fuse is there designed to melt so the temp goes up rapidly when close to there limit so the temp. coefficient also increases rapidly. So This will cause more distortion than a resistor, but still inaudable. As mentioned before its been measured at .0033% probably worst case, just before the fuse blows.

As an EE, Im very well aware of everything you stated about resistors. I used resistors as an example to show that they cannot modulate a signal. I'm also saying that fuses dont modulate a signal.

Based on the definition of modulation....

Modulation is a process of mixing a signal with a sinusoid to produce a new signal. This new signal, conceivably, will have certain benefits over an un-modulated signal. Mixing of low frequency signal with high frequency carrier signal is called modulation.
{\displaystyle f(t)=A\sin(\omega t+\phi )}

we can see that this sinusoid has 3 parameters that can be altered, to affect the shape of the graph. The first term, A, is called the magnitude, or amplitude of the sinusoid. The next term, {\displaystyle \omega }

is known as the frequency, and the last term, {\displaystyle \phi }

is known as the phase angle. All 3 parameters can be altered to transmit data.
The sinusoidal signal that is used in the modulation is known as the carrier signal, or simply "the carrier". The signal that is used in modulating the carrier signal(or sinusoidal signal) is known as the "data signal" or the "message signal". It is important to notice that a simple sinusoidal carrier contains no information of its own.
In other words we can say that modulation is used because some data signals are not always suitable for direct transmission, but the modulated signal may be more suitable.

Types of Modulation[edit]
There are 3 basic types of modulation: Amplitude modulation, Frequency modulation, and Phase modulation.
amplitude modulationa type of modulation where the amplitude of the carrier signal is modulated (changed) in proportion to the message signal while the frequency and phase are kept constant.frequency modulationa type of modulation where the frequency of the carrier signal is modulated (changed) in proportion to the message signal while the amplitude and phase are kept constant.phase modulationa type of modulation where the phase of the carrier signal is varied accordance to the low frequency of the message signal is known as phase modulation.

My arguement is with the term modulation used in this discussion. Since fuses at audio frequencies do not behave reactively, they act more like a resistor. Operating well within the fuse's limits, any changes in the resistance is linear, not reactive. You get a minute resistive voltage divider going on. There is no mixing of signals together, hence no modulation.

Since this is an audio forum and not an RF forum, I'm not looking at transmission line theory. TL theory does NOT apply at audio frequencies as the parasitic inductance and reactance losses are neglible for the cable runs used. I'm not disputing that fuses add distortion. Anything that generates noise adds distortion. I'm saying that the term modulation is incorrectly used to describe noise generated within the fuse. One has to achieve absolute zero for thermal noise to stop. Everything else stops too at that temperature.

 

[scott wurcer]

As an EE, Im very well aware of everything you stated about resistors. I used resistors as an example to show that they cannot modulate a signal. I'm also saying that fuses dont modulate a signal.

This is not true, resistors have a temperature coefficient and are only linear at a constant temperature, it is trivial to demonstrate THD and IMD due to excessive self heating (which is BTW how a fuse works). The magnitude of the effects is another matter but their existence is irrefutable.

 

[3dbinCanada]

This is not true, resistors have a temperature coefficient and are only linear at a constant temperature, it is trivial to demonstrate THD and IMD due to excessive self heating (which is BTW how a fuse works). The magnitude of the effects is another matter but their existence is irrefutable.

Why does everyone keeping going to the extreme? Did I not say in an earlier post that things go to shite at extremes and all bets are off? Do you listen to your system at your volume set to 10 out of 10 with a heat gun pointed at your electronics? Within nominal operating conditions, Johnson noise is the distortion effect.

 

[scott wurcer]

Why does everyone keeping going to the extreme? Did I not say in an earlier post that things go to shite at extremes and all bets are off? Do you listen to your system at your volume set to 10 out of 10 with a heat gun pointed at your electronics? Within nominal operating conditions, Johnson noise is the distortion effect.

Science is about accuracy and attention to details that can ruin an experiment, i.e. the super-luminal neutrinos that ended up getting some scientists fired. There is nothing wrong with being a completest and determining that even though something exists in the context at hand it does not matter. Saying it does not exist only opens the door to problems. There have in fact been high power amplifiers designed with low power resistors in the feedback network where the distortion was measurable if not necessarily audible.

 

[3dbinCanada]

Science is about accuracy and attention to details that can ruin an experiment, i.e. the super-luminal neutrinos that ended up getting some scientists fired. There is nothing wrong with being a completest and determining that even though something exists in the context at hand it does not matter. Saying it does not exist only opens the door to problems. There have in fact been high power amplifiers designed with low power resistors in the feedback network where the distortion was measurable if not necessarily audible.

You have a talent of putting words in people's mouths. Where did I ever say it doesn't exist? I said the term modulation was incorrectly used to describe something to explain something that doesn't occur in nominal conditions, nominal meaning operating well within the current capability of the fuse. Thats different than stating that the condition doesnt exist. This is about audio in the home environment and not building HF amplifiers to be used in space conditions. If you pay attention to detail, then you wouldn't purchase a speaker with an inline fuse. Lets get real here.

 

[scott wurcer]

Where did I ever say it doesn't exist?

I used resistors as an example to show that they cannot modulate a signal.

You're angry over nothing, let it pass.

 

[DonH56]

Temperature changes in a resistor (or fuse, or wire...) change its value and that modulates the signal amplitude. Amplitude Modulation, AM. In your stereo or in a satellite. Nobody is applying transmission line theory to the fuse (we could, but why?)

In a speaker fuse within its normal operating range the impact is negligible.

 

[3dbinCanada]

Temperature changes in a resistor (or fuse, or wire...) change its value and that modulates the signal amplitude. Amplitude Modulation, AM. In your stereo or in a satellite. Nobody is applying transmission line theory to the fuse (we could, but why?)

In a speaker fuse within its normal operating range the impact is negligible.

The fuse is in series with the loudspeaker so the voltage change appears across the fuse linearly to the resistance change. Its a simple voltage divider. There is no carrier signal, just the original signal so there cant be a modulation. Modulation requires a carrier signal. There is none. The term modulation is being used incorrectly.

 

[3dbinCanada]

You're angry over nothing, let it pass.

I'm good as long as you don't mis quote me. If you are unclear of what I said, please dont hesitate to ask for clarification. I'd be more than happy too clarify.