Capacitor cocktail - types, locations

[TUBEDUDE]

I taste electricity, but try to limit it to a helping of 9 volts at a sitting. Frequently my battery tasting reveals a somewhat weak sauce, but a potent alkaline is a salty jalapeno stew. Good thing we don't use C- batteries anymore. That would be a Ghost pepper!

[R.G.]

Here is what I know for a solid gold fact. I swapped out 6 caps in this amp of the same value, just swapping out Mallory 150s for the Dijons, and the difference was more than just noticeable. It was like a totally different amp. Better in all ways except the looser bass...

That is perfectly reasonable. Many people have reported similar things.

To me anyway. I get that that's a totally subjective and personal evaluation. I'm trying to achieve a tone that I like, but others will have a completely different and equally valid take on it.

Also completely reasonable; I have seen this kind of thing.

What I would like to understand why I heard a difference. Polypropylene vs Polyester? Construction? Quality of materials? Quality control?

And I'm completely with you here. People hear differences when they change caps. What's happening?

The hifi tweakos have sometimes been faced with similar issues. They replace - or even reverse the direction! - of a resistor, cap, or even a wire, and hear dramatic differences in sound. A great deal of ink has been spilled on both sides of this. The ideas from there are creeping over into the musical amplifier world.

There are some things we know for sure, as a result of repeatable testing:
- electrolytics are a mess; they show measurable distortion any time they have to support a significant signal voltage. This is fairly low distortion, but it's measurable. The jury is still out on whether other cap types show distortion in the sense of generating any harmonics or intermodulation products of a signal through them. Instruments don't detect this yet for film caps that I know of. Perhaps this is a "yet".
- There are measurable differences in Equivalent Series Resistance (ESR) and Equivalent Series Inductance (ESL) based on the conductive plates' materials and construction. This is a measurable fact, and the reason that metallized film, wound film, and stacked foil and film caps even exist as different things. They differ in the "side effects" of the internal resistances and inductances.
- There are measurable differences in the dielectric effects from different insulators. Polypropylene, polystyrene, and the rest of the poly-mumble-enes do have slightly different characteristics. Often this is in terms of DC soakage effects and the dielectric constant. They also vary in their response to moisture, pressure, history, recovery from over voltage, lots of things.
- There is a big, and repeatable effect of capacitor tolerance. The simple fact that any two 10% tolerance caps may be 20% different from one another is a big deal in terms of frequency response. It's possible for it to be even bigger if the circuit adds two previously filtered signals and there is some cancellation going on.
- There is the known, but shadowy mess of psychological effects. Humans are not instruments, and even the same human tends not to be repeatable over time. We have in-built (and probably in-bred) biases that are hard to even recognize, and even harder to defeat. That's the whole basis for double blind testing. The history of this is fascinating. I know that whenever I work on something I think it sounds better, just because I worked on it. I try to discount this and get third party opinions, because I have caught myself at it before. But it is quite difficult to get around. There are many other subtle, hard-to-even-recognize biases and byproducts of the human auditory system.

One thing is certain: humans do not hear or see accurately what comes in through the ears or eyes. We make it up depending on what else is going on in our heads, or what *ever* went on in there.

One thing the techies do is to try to make ever-more-accurate models of what the internal materials and construction details do to the circuit. The whole idea of ESR and ESL were invented to let designers get closer to what they saw in finished units while they were still doing a design. It's worth a lot of time (and hence, money) in industry to do that right. So they're very interested

Upon further reflection, it's worth a huge amount of money if someone can ever figure out what you're trying to do - find what kinds and types of caps sound better if you use this one here, and this one there, and over here it's better to use these other ones. An amp maker would leap all over that, like a chicken on a June bug as we say in the South.

[roberto]

Another point is how the same capacitor behaves when you have hundreds of DC volts on top of our beloved AC.

[Ken Moon]

Nobody really understands fully how the human ears and brain interact.

There is a whole field of study called psychoacoustic modeling which delves deeply into the subject, and is the foundation of the algorithms of audio encoders like LAME.

It is pretty astounding to me that a signal can be digitized, six out of every 7 bits deleted, and most people cannot tell the difference. An example is audio encoded by LAME with "studio"presets and a variable bit rate vs. the original WAV file.

If you are working in a studio that records at 24/96 or higher, or listen to high quality audio recorded and mastered in DVD-Audio format (24/96), most can hear some sort of difference from a 16/44 WAV file, but I have yet to be able to hear a difference in guitar amp tone caused by component type selection that has anywhere near the effect of guitar pickup, speaker choice, or a little fiddling with the tone stack.

I have tried. I built this amp many years ago in an attempt to use the "best" component types in all the right places. I built several amps with the exact same circuit but a wide variety of component types over the years, and I'd have to say that the easiest way to "get the sound you hear in your head" is to build a well-designed 1x12 cab (like the AX84 cab), then buy a dozen or more speakers, and blind A/B test them until you find one that makes you happy.

[img761]http://i514.photobucket.com/albums/t346 ... wiring.jpg[/img]

[Decko]

Good read here!

I've built the same amp many times. Each amp using the exact same components. Yet the amp tone was not 100% repeatable.

While the basic foundation for the tone was similar, the detailed nuances can be very different.

Today when I build an amp, I have accepted that each build will have its own character and that it will be up to me to experiment with various brands, types and values of caps and resistors, various types of wire, transformers, and speakers to create the tone that I am looking for.

Most of the time it will take much longer to tune the amp than it is to build the amp.

IMHO, amp building is art to the ear.

[R.G.]

That is an absolutely great way to think of it.

The number of variables is HUGE.

[lumox0013]

I have done more than a few amps each different but i find it to be alot of trial and error as well as personal taste. this youtube vid sheds alittle light on their aproach judging from the artists using their amps Id say it works. https://www.youtube.com/watch?v=x1xbvy0l1Pw cheers lumox

[WRC34]

I am a newbie to all of this so I may risk sounding like a turd here.

My humble opinion -

a) since we are dealing with instrument amplifiers it would seem that nothing comes close to substituting for an audio test or example, done in such a way that the only variable changing would be the area of focus (caps, resistors, speakers, pickups, etc). I am surprised by how little there is in the way of audio examples/experiments out there with this stuff...HOWEVER! what we are dealing with is much more elusive than that, and like the interactive tone/volume controls on a 5E3, when you change one thing everything else behaves differently as well. What a conundrum!

b) have any of you seen the film "What the Bleep...?" about quantum physics and how it relates to our daily lives? There are many fascinating experiments done throughout that film. One that stands out when thinking about capacitor types and conducting (as) scientific (as possible) tests is where they show that atoms, yes atoms, behave differently when observed than not. They 'propel' the atoms forcefully in a particular direction, through what looks like a poster board with slots in it, onto another board behind it which enables them to see the way the atoms 'land' after passing through the slots. They do this in a room with closed doors and no one/nothing present, then they set up a camera and have it rolling while they 'fire' the atoms, and the atoms land in a completely different pattern. That's particles behaving differently when observed by a camera, which unlike a human has no variance in it's results! If this is true, and it most certainly is, then you can imagine how difficult it would be to get an accurate result when testing different cap types in something as temperamental as a tube amp when using something as biased as the human ear to judge the results. PFFFTT!!

c) despite my believing b) to be true, I still think it's worthwhile to try different caps and blend them to taste. I have used a crazy combination of different types in the builds I have done with no concern or worry in regard to the junk box thing since I build point-to-point which looks unruly anyway. I build more like a crazed madman than a scientist, but still have been pleasantly surprised with the results of using different caps in different places. Almost to the point where I'd say that using all of one type of cap might produce less pleasing results than blending them. I say almost here because the best amp I've ever heard or played is a JTM45 clone with all Mullard mustard caps

d) years of buying used motorcycles from people in the local "Want Ad" has been beneficial in helping me to discern quality vs. crap when it comes to amp components. If it looks and feels like cheap crap, it probably is. And if it feels good, it is good. In this case the you can replace 'feels' with 'sounds'

[Structo]

One of the problems is the tolerance of components.

Pretty much impossible to build a true clone amp because component values will be different and drift with age.

You can get in the ball park where most Marshalls sound like Marshall and a Super Reverb will sound like a Super Reverb.

For test purposes you could set up a simple circuit to switch between two components to try and hear a difference.

I sort of agree about measurable differences should be observed but maybe we are looking at it the wrong thing when we test.

A change in a component can affect more than just it's new value but can affect the circuit around it.

[R.G.]

For test purposes you could set up a simple circuit to switch between two components to try and hear a difference.

We actually use this test quite often in my day job. We'll sub in different values or different parts with an "ABX" switch rig I set up for the purpose. It's quite useful. The office is full of guitar players and they laughed at me when I started this, but now they go and get it out to squint out fine differences. For critical stuff, we have a control cable that lets the guy who's controlling the switch be outside the room to eliminate them tipping off the listener to what is being switched. It's not a true double blind, but it's a lot faster, and does tend to eliminate the expectation effect and other biases.

I sort of agree about measurable differences should be observed but maybe we are looking at it the wrong thing when we test.

A change in a component can affect more than just it's new value but can affect the circuit around it.

This is correct. What can be measured should be, and that taken into account. If you try and **can't** measure it, you have to ask the resulting questions:

Are we measuring the wrong thing?
Do I understand what I'm seeing?
Is the effect too small to see?
Is it real or not?

All of these are possible explanations.

I read last night about how it was discovered that atoms are mostly empty space. Rutherford (IIRC) set up an experiment where he was firing helium atoms at a thin gold foil. Many of them sailed through unperturbed, and hit a scintillation screen. Some were absorbed. But *some* bounced back entirely. The only way that this could happen is if only a very few helium atoms were hitting something very, very small, and very, very dense - the atoms in the gold foil. In fact, the nucleus. People had always thought before that atoms were stacked together like a brick wall with maybe some holes to let a few get through. No one expected that they were almost entirely empty space, and their interactions in normal matter were all through the electric field of the electrons around the nucleus.

Rutherford spent a good long time with the four questions above, and guessed right about something no one had ever thought of before.

[Structo]

There are so many variables

Temperature
Humidity
Construction of component
Materials
Value of component (tolerance)
Voltage of component
Current
Test equipment
etc.


The hard part is duplicating a mixture of all these to recreate the same effect
or performance time after time.

Are amps that now use pcb and smd construction able to repeat performance easier due
to tighter tolerances?

[RWood]

[quote="R.G."]
the expectation effect and other biases.

The Science of Psychology likely has at least as much to do with this discussion as does the Science of Physics, as proven time and again by watching an episode of Brain Games.

[Gainzilla]

I'm not sure if it's the inherent pessimist in me, or my constant feeling of dissatisfaction, but I don't go into this with the expectation of sonic bliss. I need to hear it and feel it before I believe it. I don't expect that because it's a certain color or that it costs the most that it will make me happy. I do believe phenomenon such as bias exists, but I also don't see why it's such a stretch to believe we aren't all created equal.

The fact is, we are all snowflakes. We each have our own unique attributes, talents, and deficiencies. It's why Some people make us want to run our guitars through a wood chipper, while others struggle for mediocrity. Why some are such naturally gifted artists, but the rest of us have penmanship reminiscent of a ransom note. I believe some of us can hear better (or differently) than others.

I also believe, as mentioned in this thread, that it's possible we aren't measuring the thing that gives a cap it's mojo. i found it interesting when Doug Sewell singled out a couple treasured caps he thought sounded particularly better than most.

[Smitty]

http://www.sbelectronics.com/products/a ... ated-info/

[jam-mill]

I can't remember the source of this article, but I found it somewhere here on AG. Had to use the wayback machine to find the original. Steve Bench did measurements of the behavior of various capacitors: http://web.archive.org/web/201110150109 ... /caps.html

EDIT: Found a mirror: http://diyaudioprojects.com/mirror/memb ... /caps.html