• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Fuses do affect sound, the question is how much

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
AFAIK slo-blo fuses are used for mains AC,, but on amplifier's outputs I think I've seen slo-blo as well, even on solid-states (I've seen such fuses getting red/orange while peaks of power). For speakers there are only fast/quick blow fuses, for a better and faster protection.
 

Shadrach

Addicted to Fun and Learning
Joined
Feb 24, 2019
Messages
662
Likes
947
Sorry for the interruption. I was looking for a site called Audio Science Review and seem to have got directed here instead.
However, while I'm here and not wanting to miss the opportunity of making a few quid (well, you would have to be insane to pass up the opportunity of selling fuses at £175 each to this lot wouldn't you) I wondered if I could interest you scientific types in a few fuses?
I should be able to supply some data on changes in resistance for the various types and of course, given how it's all about the science here. I wouldn't feel right omitting such data. After all (and I am one of you guys, honest. I love your scientific approach to audio.) if it measures differently then it's got to sound different right (?).
I'll be out the back late next week with the fuses and data sheets.
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
[...]After all (and I am one of you guys, honest. I love your scientific approach to audio.) if it measures differently then it's got to sound different right (?).
I'll be out the back late next week with the fuses and data sheets.

If it's used in the power line or after the power supply it's almost impossible to change the sound, unless the voltage that drops on that fuse is at least 5% of PSU's output voltage (e.g.: 2-3 volts).

I'm looking further for your tests, I'm aware that few mV drop might affect how a speaker reacts, if fuse in the audio chain.
 

audio_tony

Addicted to Fun and Learning
Joined
Feb 24, 2019
Messages
567
Likes
690
Location
Leeds, UK
Expanding just a little on my last post on the topic, this is my take on (mains) fuses.

With a typical 100W per channel power amp, the mains power consumption will be around 350 to 400W (depending on efficiency), which means at *maximum* output, the amplifier will be drawing in the region of 1.5A from 240v mains.

Typically, a fuse will increase in resistance by about 3 or 4 times its cold value just before it blows.

So, assuming a 2A fuse for protection (more likely to be 3.15A in the above case) - the resistance of the fuse will still be low enough to have little to no effect on the current passing through it.

Additionally, there will be only be a minute voltage drop, right up to the point when the fuse blows.

Therefore, for the fuse to cause any appreciable voltage drop, the amplifier would probably have to be pushed right to the ragged edge of it's outright performance, by which time any unwanted effects caused by a fuse will be the least of your worries.

Ideally one would want to put a 'scope across the fuse, but without proper mains isolation it's not something I want to do myself, however to my mind this will give a good indication of what's happening across the fuse at high (peak) currents - but as I said above - driving an amplifier this hard must surely result in excessive THD from the amp at this point.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
There is a fundamental difference between fuses in a signal path and in the mains supply line.

inlet -> fuse -> transformer and transformer -> fuses -> rectifier both require Time-Lag or Slow Blow fuses.

Also there is a difference between :
smoothing caps -> fuses -> output devices -> speaker.
smoothing caps -> output devices -> fuses -> speaker.

I would argue it has no business being between the amp and speaker.
This should best be solved with output relays and protection circuitry (current limiters, DC protection).
However, many problems with power amps come from faulty output relays in my experience.
Fuses have a PTC characteristic and thus can add distortion when in the speaker path and driven quite loud.
Don't know how 'severe' it is though.
Are audiophile fuses somehow exempt from this effect ?
Do they really act as fuses ?

Fuses between power supplies and output devices are not really needed when the amp has decent current limiters.
When this is not present a fast fuse could potentially prevent output devices from being destroyed by shorts on the output.
This requires fast fuses.
To prevent fire I would argue fuses in the mains circuit (or secondary windings) should be enough protection.
 

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484
Slo-blo fuses are very uncommon in today's electronics. They would get a more use in tube products, things with motors and such.

That would certainly be my understanding. Fuses on the PS side would need to be able to handle in-rush current. This can be considerable for electric motors and, I suppose, tubes because of the current they need to become fully energized.
 

audio_tony

Addicted to Fun and Learning
Joined
Feb 24, 2019
Messages
567
Likes
690
Location
Leeds, UK
That would certainly be my understanding. Fuses on the PS side would need to be able to handle in-rush current. This can be considerable for electric motors and, I suppose, tubes because of the current they need to become fully energized.

I'm not sure tubes have a high inrush current? (but I don't know a great deal about them so could be wrong) - it's usually the smoothing caps that cause the high inrush current - although Class A might be a different story...

Toroidal transformers tend to have a higher inrush current than 'E core' transformers - this is usually made worse by the use of very high value smoothing caps, which is often why soft start circuits are required.

But slow blow fuses will often be found on the mains side in the above case, and the DC fuses are often fast blow when a soft start circuit is used.

If no soft start circuit is present, then typically the DC fuses will be slow blow types too.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
With tubes the start current depends on the type of heaters.
There are quick-start tubes (the ones that flash bright and then go dimmer) that draw a lot more current during start-up.
The normal heaters only draw a little more current during start-up.
The effect is nowhere near the current needed to charge the smoothing caps (SS rectifiers). When using tube rectifiers the smoothing caps are charged slowly.
Usually the start-up current of heaters is limited by the transformer.

Amps with huge smoothing caps and highpower transformers usually have a rush-in current limiter that by passes itself after a short while.

There is no need to worry about fuses except when directly in the signal path. And even then the effect is very small during 'normal' usage.
Audiophile grade fuses won't do much here either except ease your mind that those horrible sounding fuses have been replaced by goldplated 'audiophile grade' ones.
 
Last edited:

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484
I'm not sure tubes have a high inrush current? (but I don't know a great deal about them so could be wrong) - it's usually the smoothing caps that cause the high inrush current - although Class A might be a different story...

Toroidal transformers tend to have a higher inrush current than 'E core' transformers - this is usually made worse by the use of very high value smoothing caps, which is often why soft start circuits are required.

But slow blow fuses will often be found on the mains side in the above case, and the DC fuses are often fast blow when a soft start circuit is used.

If no soft start circuit is present, then typically the DC fuses will be slow blow types too.

The heater elements in tubes require a lot of energy to reach operating temperature. Some in rush protection will likely be used to prolong tube life, such as a negative temperature coefficient thermistor. But all this means is that incoming current will be used to heat the thermistor too.

I've heard that different transformer types have different in rush. I would also think that there is a size component to consider, as well as the need to fill caps (which you pointed out).
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
[...]Typically, a fuse will increase in resistance by about 3 or 4 times its cold value just before it blows.

I was reading on LitleFuses datasheet that it's 5-10% resistance increase. What datasheet has you seen, please?

[...]Ideally one would want to put a 'scope across the fuse, but without proper mains isolation it's not something I want to do myself[...]
I'm using a laptop on battery connected to a PicoScope, although not really recommended, so definitely differential probes are the best recommendation. Also, an isolation transformer should do as well if the voltage drop is over 1 volt; just need to multiply the output value with mains_input_voltage/output_voltage ratio.
 

audio_tony

Addicted to Fun and Learning
Joined
Feb 24, 2019
Messages
567
Likes
690
Location
Leeds, UK
I was reading on LitleFuses datasheet that it's 5-10% resistance increase. What datasheet has you seen, please?


I'm using a laptop on battery connected to a PicoScope, although not really recommended, so definitely differential probes are the best recommendation. Also, an isolation transformer should do as well if the voltage drop is over 1 volt; just need to multiply the output value with mains_input_voltage/output_voltage ratio.

I think I fudged my maths. :(

From Bob Cordell's book (also cited earlier in this thread on page 2 see below)

Bob Cordell said:
The cold resistance of a 2-A 3AG fuse was measured to be 78 mΩ, while its resistance when passing 2-A DC was 113 mΩ.

That's a 45% increase, or ~ 1.5x which is at 100% of rated fuse current if I read this correctly.*
It should also be noted this is with a "3AG" type fuse which is a 6.35mm x 32mm fuse (and hence 12mm longer than a 20mm fuse) so I suspect the numbers will be ever so slightly different for the shorter 20mm fuse.

Additionally, I found a datasheet (not the same one you refer to) and trying to calculate from that (again - maths fail) left me frustrated so I cited the paragraph from this post on page 2

*based on this - I agree with the figure of 3x to 4x it's cold value cited on page 2.

I hope this makes sense?
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
I believe that means a 0.07V drop between cold fuse vs. hot fuse (@2A), so not quite an impact; same applies for the added 0.035 Ohms as damping factor difference (if fuse is used in signal path, between output stage and speaker's drivers).

However, I was not expecting a 45% difference, but this is probably due to the fact that the fuse has a 2A rating and measurement was done at 2A, while manufacturers don't recommend loading a fuse over 75% of it's max. rating. For 2A measurements I strongly recommend a 3A fuse, where the diff. between cold and hot fuse will be lower.

Anyway, a fuse placed in audio signal path will definitely impact AP and scope measurements a little bit, but that doesn't means our ears will be able to notice any sound change. I also don't think any audio equipment manufacturer is building amplifiers with fuses in the signal path these days (at least not solid-state); most fuses are inside the power supplies and these will definitely not change the sound.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,273
Location
Gold Coast, Queensland, Australia
I believe that means a 0.07V drop between cold fuse vs. hot fuse (@2A), so not quite an impact; same applies for the added 0.035 Ohms as damping factor difference (if fuse is used in signal path, between output stage and speaker's drivers).

Just put the fuse in NFB loop with a high value resistor across it in case it blows- then all your non-linearities in your fuse are cancelled.
 

audio_tony

Addicted to Fun and Learning
Joined
Feb 24, 2019
Messages
567
Likes
690
Location
Leeds, UK
Just put the fuse in NFB loop with a high value resistor across it in case it blows- then all your non-linearities in your fuse are cancelled.

Rotel did this for years, not sure if they still do.

Curiously, some of their designs don't have a resistor across the fuse, so if the fuse blows, the NFB is open circuit!
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
Hafler also did that in the DH200 back in the 70s.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,273
Location
Gold Coast, Queensland, Australia
Hafler also did that in the DH200 back in the 70s.

Personally, I don't like fuses in the amplifier output at all, but if a proper relay protector (DC, over-current, temp etc) isn't in the budget, something is better than nothing. Might as well attempt to ameliorate its effects and keep it in the NFB loop.

HK had a thing for self re-settable thermal cutouts at one point too in the 70s and 80s. They unfortunately became a source of trouble as their internal contacts oxidized. NAD even used them in the original 3020 IIRC.

Here's the PM-640 with the cutout in the NFB loop (tone controls also in the loop- not shown to left).

1559689301926.png


NAD 3020:

1559689581989.png


In the loop.

1559689773086.png


@SIY So if the fuse is of a reasonable linearity, adequately high in value (to carry the maximum planned load) and resides inside the NFB loop of a fast enough amplifier, what would you expect to see in low level THD contribution? For example, if we took say the Benchmark and put a 7.5A(F) fuse in line and kept it in the feedback loop? Would be be looking at just increased THD at very low levels where the contact oxide (breakthrough) resistance comes into play from the fuse holders etc?

What about if we extend the feedback loop (a la Kenwood Sigma Drive circa 1980s) all the way to the loudspeaker terminals themselves and take the speaker cable out too? Have you ever experimented with feedback at the speaker terminals or negative impedance drive (lifting the 0v) and putting the driver in the loop? I have. It wasn't pretty.

Yamaha managed the so-called 'negative impedance drive' quite successfully with their AST (active servo technology) in the early 1990s. Combined with very basic crossovers and response matching cartridges (which plugged into the integrated amplifiers, customized for each speaker), they could achieve some phenomenal sounding combinations with very compact speakers. It didn't take off and we ended up selling the entire range very cheaply. People would lose the 'flat' cartridges and/or the ones that came with the speakers. They'd sell the speakers and forget to give the new buyer the cartridge which was still in the back of their amp etc. Great idea, but still-born.
 

Attachments

  • 1559689260564.png
    1559689260564.png
    87.1 KB · Views: 114
Last edited:
OP
tomelex

tomelex

Addicted to Fun and Learning
Forum Donor
Joined
Feb 29, 2016
Messages
990
Likes
572
Location
So called Midwest, USA
Hafler also did that in the DH200 back in the 70s.


yes, on the DH220 it was actually an RC circuit, 2K2 and .01 micro across the fuse/fuse holder although it was more to do with controlling HF oscillations form the output MOSFET sources than anything the fuse itself was up to.
 
Last edited:

egellings

Major Contributor
Joined
Feb 6, 2020
Messages
4,004
Likes
3,244
I had a power amp in which the speaker fuse was included in the feedback loop, and the amp had provisions for not losing it and ruining the speaker if the fuse blew. The feedback corrected for the fuse's misbehavior.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,834
Likes
16,496
Location
Monument, CO
I thought most amps these days were using something other than fuses for output protection? More circuits for SOA and such rather than a fuse. A fuse in the feedback gets compensated but that does not help the outputs; in fact, the fuse can introduce nonlinearity in the feedback path not in the main path. And if it blows you're left with the amp's open-loop transfer function. Hmmm... The times I have seen such a fuse, it was there so the output impedance remained low (did not want the fuse in the output path), and if the fuse blew there was some sort of snubber of clamp circuit to shut down the amplifier.

And yah, a Hafler without the little snubber circuit across the fuse, will oscillate wildly when the fuse blows. I used to have pictures... And there was one well-known big bad amp (ML, Krell, Threshold? can't remember but someone will) that would not only oscillate but exhibited the classic inversion at the output similar to overdriving early op-amps. You'd get this HF oscillation with LF "motorboating" polarity inversions at the same time. It tended to make speakers unhappy. At least my flame linear would just apply the power rail to the output when it died, destroying the woofer if it couldn't handle it until the power fuse blew, but the rest of the drivers usually had a series cap that blocked DC. When it smoked my Maggies (literally burning the foil off the diaphragm) it was history.
 
Top Bottom