The truth about caps

[Diablo]

will this sound better than brand "X" capacitors?

submitted for your entertainment

Aaudio Imports: Product

I love this quote from their "specifications":
Fuses always carry high electric current thereby causing metal fatigue. This would then adversely alter the conductivity behavior of the fuse element and hence the performance of the equipment.

Well, I'm a metallurgist, and I was never taught that electricity causes metal fatigue, and I never read about that possibility before. So, I guess we should also tear out the wiring inside our house after some period of time because the copper has been fatigued by the electrical current. I think I'll go purchase some stock in the copper mining industry. Maybe we should ask them why their gold plated fuses aren't fatigued by the electrical currents. This is fascinating science...

[guitician]

You have to ask yourself: What makes the fuse blow?

The thinness of the wire in the fuse! If your house was wired with 24 AWG, there would be some fatigue for sure. In the way of a fire....lol

[stokes]

$39 for a fuse!!!Its gotta sound better,no?What was it that guy said...theres one born every minute.Maybe for $39 it never blows,just repairs whatever would have caused those cheap fuses to blow.Never realized that slow blow was responsible for "best sound quality"

[Steve Conner]

I wonder if the audio industry will ever develop something akin to the mass spectrograph? Something that analyzes every possible thing occurring in a stream of waveforms. Like a Cat scanner for the instantaneous position of the electrons/air molecules as they change in every 1/3 octave. I guess hype is just a cheaper way to go...lol

Ha! This is in fact the manifesto of the Audio Objectivists. They say that if you have an oscilloscope and a spectrum analyzer, those analyze every possible thing that can occur in an audio circuit, exactly as you said. So as an objectivist myself I say we've already got this thing.

Cminor9: I don't buy the "Xicons= butt tone" theory. It'll be something else (that makes more than 3dB of a difference to something! ) If I were troubleshooting this, I'd hook a scope up to the amp and look for some kind of waveform that appeared at the same time as the "butt tone" was evident. From my experience of tweaking audio circuits, I have a fair idea of the relationship between waveforms seen on a scope, and sounds. In this way I'd eventually figure out what part of the circuit was faulty or in need of modifying.

Do you only get the unobtanium nut on 10-top PRSs? Mine doesn't have one

[defaced]

You have to ask yourself: What makes the fuse blow?

The thinness of the wire in the fuse! If your house was wired with 24 AWG, there would be some fatigue for sure. In the way of a fire....lol

That's not fatigue though. Fatigue is deterioration of metal's properties (usually mechanical strength and ductility) due to repeated long term application/release of stress. A fuse blowing is just increasing it's temperature (due to increased current) to the point where the material melts/explodes/whatever - it's basically uncontrolled welding. It's an instantaneous event, which is not fatigue.

Fuse fatigue, untill someone shows me micros of the material's structural change as a function of cyclic loading along with how that correlates to its electrical properties, is BS.

[rudutch]

I happy you all found it amusing, I was amazed at the marketing. I wonder how may people actually bought these and were then convinced they could 'hear the difference'?

[cminor9]

If I were troubleshooting this, I'd hook a scope up to the amp and look for some kind of waveform that appeared at the same time as the "butt tone" was evident. From my experience of tweaking audio circuits, I have a fair idea of the relationship between waveforms seen on a scope, and sounds. In this way I'd eventually figure out what part of the circuit was faulty or in need of modifying.

Hooking the amp up to a scope; now that would be ideal. But I don't have one. I don't have hundreds of dollars to invest in a new one. Maybe I could find an old one for cheap somewhere. Until then...

I want the xicons to be ok; $1 for a xicon cap vs $7 for a sprague vs $5 for an F&T. But there has to be some common variable between the two amps. But correlation does not equal causality, I know.

[J M Fahey]

who can measure 6800uf anyhow to know if it's off? I can't do it with my DMM.

You can, as anybody else.
Just drive your amp to clipping .
It's very easy to measure, hear and specially see, if you have a scope, the difference between, say, 6 Vpp ripple and 2 Vpp, which come from the difference between 2200uF and 6800uF caps.
We're talking a 3:1 difference in ripple/hum content mixed with your clipped musical signal, or 10dB, way over the usually accepted 3dB threshold of audibility.
Of course, I'm talking about capacitance differences here, the only significative parameter.
All others pale by comparison.
Modern components are usually *very* good and consistent, much more so than 20 to 40 years old components, with the exception, perhaps, of tubes.
The market is *very* competitive, we're talking billions of parts made, in a fully robotised way.
Just there's not space for a backyard, improvised factory, simply because its poor products will be more expensive than good generic ones.
Of course, fakes as those shown above, are a whole different game.
You don't even need to open the cans to spot such gross fakes as above:
1) they weigh less
2) they feel very different in your hand.
Those smaller caps floating , suspended by two puny legs inside an empty bigger can, vibrate physically at a couple Hertz frequency, will feel very funny when you move them, or when you put them on any hard surface.
The cheap fakers, might at least have used some plastic foam, hot glue, or even a cardboard ring to keep the whole mess steady.
Even more, they should have pulled the inner , self-incriminating plastic sleeve.
That alone, makes me think that this particular photo is a fake or a joke.
As a side comment, caps *have* been shrinking steadily for the last 40 years.
I still have, somewhere, a couple Campbell Soup can size electrolytics from the late 60's, Spragues or Mallorys, which are twice (or more) the diameter and twice or more the height of modern ones, we're talking an 8:1 size and weight difference.
Do they sound bluesier/warmer/better/britisher/soulful/transparent/heavenly/..../etc. than new, "inferior" ones? In fact, not, specially because they are probably half-dry by now.
Of course , I don't build "Audiophile grade" equipment, so maybe I'm not qualified to post these, admittedly, personal opinions.
Peace.

[R.G.]

I love this quote from their "specifications":
Fuses always carry high electric current thereby causing metal fatigue. This would then adversely alter the conductivity behavior of the fuse element and hence the performance of the equipment.

Well, I'm a metallurgist, and I was never taught that electricity causes metal fatigue, and I never read about that possibility before. So, I guess we should also tear out the wiring inside our house after some period of time because the copper has been fatigued by the electrical current. I think I'll go purchase some stock in the copper mining industry. Maybe we should ask them why their gold plated fuses aren't fatigued by the electrical currents. This is fascinating science...

Their explanation/advertisement is indeed pure male bovine excrement. However, fuses DO fatigue and DO fail because if it. It's just not the electricity (directly at least) that does it. As a metallurgist, you're familiar with the physical characteristics of low-melting-point metals, I'm sure. The thing about fuses is that they are designed to hold their specified current forever, but to open in about a minute, depending on the time rating of the fuse, on a 100% overload. As you could guess, this requires some fancy calculations of the resistivity of the metal, the temperature coefficient of the resistivity, and the melting point of the metal and how those affect the fuse element when it operates near the melting point of the metal.

But think about this: if you run a fuse up to 100% of its rating, it gets hotter, right? And all the fancy alloying in the world can't guarantee that a special melting point type alloy also has a zero thermal coefficient of expansion, right? For clear glass fuses, you can see the element expand into a curve instead of a straight line as it gets to its rating, and it will visibly wag when overloads push it over 100%, even if the overloads are so short they don't make it actually clear.

This mechanically flexes the ends of the fuse which are held fast to the metal and heat-sinking caps, which keep some of it cool and therefore not annealed by the heating. Any surface defect on the cooler ends can and does snap under fatigue cycling from near or just-over currents. And old fuses do die this way. Bending, spiraling, etc. the element helps and this is why many fuse element are formed this way.

But you're right - you can't HEAR the stressed metal in the fuses.

[Steve Conner]

You can measure it, though.

Douglas Self once published plots of THD generated by speaker fuses near the point of blowing. They produced a big rise in distortion at low frequencies. I think it was due to temperature coefficient of resistance, and the temperature varying wildly on the timescale of the signal.

I also remember once testing a brand new power amp at work, just out of the box. One of the channels sounded weak and crappy. I opened it up, and the speaker fuse for that channel was loose in its holder.

In either case, the solution is not to use speaker fuses. An electronic current limit is better anyway. The fuses should be in the DC rails as backup, to prevent a fire hazard if the output transistors do fail. And in the DC rails, you won't hear them because the power amp's negative feedback will get rid of any distortion they might introduce.

About the photo I posted: As far as I know, it's an example of some particularly nasty Chinese fakery. Or maybe it is a joke, you never can tell on the internet.

[R.G.]

I also remember once testing a brand new power amp at work, just out of the box. One of the channels sounded weak and crappy. I opened it up, and the speaker fuse for that channel was loose in its holder.

Yep. Loose contacts, especially on mechanically-shaken thing like a guitar amp will do you in.

In either case, the solution is not to use speaker fuses. An electronic current limit is better anyway. The fuses should be in the DC rails as backup, to prevent a fire hazard if the output transistors do fail. And in the DC rails, you won't hear them because the power amp's negative feedback will get rid of any distortion they might introduce.

Fuses in the DC rails are problematical. You have to ensure that both rails are interrupted within the time that would burn out a speaker if only one is blown. One obvious approach is to fire off a thyristor between the two rails when a defect is detected. This was once a popular protection scheme, but the brutal turn on of a thyristor creates a new set of problems, especially if the fuse either doesn't clear or arcs over.

My favorite scheme is to put high current arrays of MOSFETs in the + and - rails, clamping them saturated-on for normal operation and turning them off when the fault detector circuit goes off. I thought this was an obvious application, but I've never seen anyone else do it, even after I posted it on my web site. A first impression would be that it's no better than having the output transistors, which are after all in series. However, these devices are always saturated on, not active until tripped. So they stay (potentially, if you do it right) cooler and less stressed by normal activity. High current ratings are cheap in MOSFETs too, so one could over-design these MOSFET current switches to be massively better at carrying and interrupting the power rails than the output devices, which have other considerations to act on. If you use P-channels for the + rail and N-channels for the - rail, you can get by without high-side drivers, although LED/photovoltaic MOSFET drivers will let you do both sides with the same activate/trip circuit, no high side drivers.

Once you have rail-interruptors, you can gang up fault sensors and clean up some other issues. DC offset on the output? Fire the fault detector. Overheat on the output heat sink? Fire the fault detector. Overheat on the drivers? Same remedy. Want a remote enable? Hmmm...

I'm in the middle of doing the work on one of these for an amp I'm refitting. The nice thing if you use both N and P channels for switches is that you can bolt them directly to two strips of aluminum, which becomes a current buss bar for supplying the output stages or the entire amplifier. The inputs are on the sources, the outputs on the drains. The body diodes let reverse currents pass back into the filter caps. And the aluminum buss strips form all the heat sinking the MOSFETs need.

About the only drawback is that you have to use multiple parallel MOSFETs for the job, because getting really low channel resistance to keep any audio artifacts from being caused by the MOSFETs requires you to get the MOSFETs down to about 1/10 or less of the emitter resistors on the output stage. So you have to use several devices, more than you would think from casual examination of the power to be handled.

[Gtr_tech]

I got scammed on some caps a while back. I ordered 22u and 33u @450v. The 22s read ok on my cap checker, but the 33s read around 20. I got suspicious of the double rubber ends/double crimps like a non polar has, so I ripped one apart. Found this:
fake caps on Flickr - Photo Sharing!
DSC02035 on Flickr - Photo Sharing!
radial on Flickr - Photo Sharing!
DSC02036 on Flickr - Photo Sharing!

And one was even kinda smashed under the re-shrink....there's the original shrink under that. I suspect they were "recycled" pulls.

They were all 22u 400v parts.

Seems like you can't buy *any* electronic parts without getting fakes, unless you buy from a distributor you can trust. Fake STK chips....why would they even bother? Most of the larger/older ones aren't a big ticket item since they were used in 20-30 year old designs and not in high demand. As some of these parts get harder to find at the large distros, its gonna get real hairy for the tech who has to track down replacements.

[Chuck H]

Seems like you can't buy *any* electronic parts without getting fakes, unless you buy from a distributor you can trust.

I got my 30UF 500V caps from Mouser Electronics. Trusted and used by almost everyone U.S. on this forum. Yet I got a "small cap in a big can" thing. I'm more inclined to go with R.G.'s explaination of the whole thing. They're not "fake" caps, just smaller ones. The bigger can is what the MFG has been selling to their distributors so Sprague just made it work... Performance differance???

Chuck

[Merlinb]

That's not fatigue though. Fatigue is deterioration of metal's properties (usually mechanical strength and ductility) due to repeated long term application/release of stress. A fuse blowing is just increasing it's temperature (due to increased current) to the point where the material melts/explodes/whatever - it's basically uncontrolled welding. It's an instantaneous event, which is not fatigue.

Fuse fatigue, untill someone shows me micros of the material's structural change as a function of cyclic loading along with how that correlates to its electrical properties, is BS.

Fuse fatigue is certainly not BS- fuses break all the time, without blowing. In solid state amps the fuses pass enormous ripple current, which causes the wire to continually flex (from the force exerted by the magnetic field). Also, when you switch on, the inrush current will cause the fuse to visually jump. Over time this will lead to a fuse breaking of course, without blowing.
Even in a valve amp the inrush will eventually fatigue a fuse, although the ripple current is obviously lower.

While I agree that gold-plated fuses are unlikely to have any audibly effect on an amp, especially a valve amp, fuses certainly do have a measurable effect on amp performance, at least where powerful SS amps are concerned. See page 248 of Self's book:
Audio power amplifier design handbook - Google Books

Given the relevance to this thread, it is also worth reading the section about electrolytic LF distortion, which I believe manifests in guitar amps when small cathode bypass caps are used:
http://books.google.com/books?id=BRQ...age&q=&f=false

Page 44 also showed some differences between ordinary caps and an audiophile cap:
http://books.google.com/books?id=BRQ...age&q=&f=false

[Diablo]

I hadn't really thought about the service conditions of a fuse. I didn't realize they are operating normally so close to their melting point. I never watched them expand and contract in response to the current. Under these conditions, thermal fatigue would be the failure mechanism. Gold plating the ends surely won't make a difference to the thermal fatigue of the fuse filament wire.