"Sound" of Capacitors

[Sherman]

I built a stereo amp using Brian's boards and the basic kit using Panasonic capacitors. It sounds great and I've tried it on three different sets of speakers, two which I've built and one set of Advent/1. I've been able to compare it to a couple of mid-level solid state amps and find it definitely holds its own.

Now a friend of mine has built an amp basically following Brian's schematic but using perf-board and hardwiring the "traces" using 18AWG solid core wire. Since he used what he had in his parts bin the caps he used were cheap Xicons. They are 1000uF, 50V. He used Dale 1% resistors (about 3 times as large as the Phoenix resistors in Brian's kit).

My friend brought the amp over and we sat down and listened to it using the same sources and speakers that I tested mine with. I think his sounds just as good (though it doesn't look as nice and the boards are bigger ).

We tried his amp with his power supply and with mine. We also used his ps on my amp. The main difference we heard was when changing power supplies. His is bigger (200VA vs. ~100-120 for mine) and has higher voltage (+-31 VDC vs. +-27VDC for mine). His ps sounded a better on both amps with the Advents, the sound was closer with my higher efficiency home-built speakers.

So... I don't claim to have either a golden ear or high-end sources (a Pioneer turntable with ADC cartridge, an Aiwa DVD player for CDs, and a cheap DVD player for CDs but I thought my amp with the Panasonic caps should have sounded better. Then I read some things on Rod Elliot's site saying the "sound" of capacitors is grossly over-rated. Also the signal path, actually all the paths, in his amp are longer but that doesn't seem to hurt the sound. What do others think?

 

[Jay]

Off course Rod was right when he said that the “sound” of capacitors is grossly over-rated. In fact, it is not only relevant to that peticular saying. The reason is that too many people do not do the work, but reading and hearing And also caps do not sound, not even do amplifiers.

We don’t know what exactly happened when you said a sound is “better” than the other. Sometimes people have similar impression due to merely an increase in SPL.

Bigger transformer rating tends to sound better, when it is able to deliver current easily. For a good sound, you need transient or dynamics, so you often need an overated power supply. And so is voltage rating. And EI tends to sound “better” than toroid.

The next part in your chain is the rectifier. You need a fast diodes for good dynamics. Almost any rectifier constructed from axial diodes sound better than any bridge rectifier.

The capacitor, when it is only 1000uF, is unpredictable IMO. But I’m sure that your system won’t sound better if you replace the Panasonic with the Xicons.

Now, why not changing caps?

About the longer path, it was made from solid (copper) wire, wasn’t it?

The more complicated thing in audio is to know if
what sounds good is really good!

 

[Jocko Homo]

I'm not sure what to make of all that....

But........

If anyone thinks that any filter cap/transformer combination will "sound" just as good.....or as bad.....as any other, hasn't built enough amplifiers yet.


An amplifier is just a modulated power supply. And the power supply is ........?

Yeah, that's right.

Jocko

 

[carlosfm]

Look, Sherman:

With those sources (as you recognize yourself), you're not going to get reliable conclusions.
Where's the dynamics?
Where's the detail?

Ahhh... forget it.

Trash-in-trash-out, the source is the most important component of any system.
IMHO and whatever.

Note: I'm still waiting for the miracle (some people believe this! ) of a good amp and speakers making a bad source sound good.

 

[jan.didden]

Off course Rod was right when he said that the “sound” of capacitors is grossly over-rated. In fact, it is not only relevant to that peticular saying. The reason is that too many people do not do the work, but reading and hearing And also caps do not sound, not even do amplifiers.

We don’t know what exactly happened when you said a sound is “better” than the other. Sometimes people have similar impression due to merely an increase in SPL.

Bigger transformer rating tends to sound better, when it is able to deliver current easily. For a good sound, you need transient or dynamics, so you often need an overated power supply. And so is voltage rating. And EI tends to sound “better” than toroid.

The next part in your chain is the rectifier. You need a fast diodes for good dynamics. Almost any rectifier constructed from axial diodes sound better than any bridge rectifier.

The capacitor, when it is only 1000uF, is unpredictable IMO. But I’m sure that your system won’t sound better if you replace the Panasonic with the Xicons.

Now, why not changing caps?

About the longer path, it was made from solid (copper) wire, wasn’t it?

The more complicated thing in audio is to know if
what sounds good is really good!

You know what bothers me? In a power supply, the diodes conduct say 10% of the time (only when the sec voltage is above the cap voltage). That means that 90% of the time the transformer is effectively disconnected fom the amp.
How the hell does the transformer manage to sneak in in those 10% and make up for the sound when it wasn't there in the other 90%?

Jan Didden

 

[SY]

How the hell does the transformer manage to sneak in in those 10% and make up for the sound when it wasn't there in the other 90%?

Perhaps it's a matter of the degree of coupling between primary and secondary (passing noise)? John Curl was big on that, and it's certainly made me hesitate before using a toroid. Another possibility (and I haven't thought this one through, just shooting from the hip) is what happens near current peaks with the core; could that pass on garbage as it goes near (or past) saturation?

 

[jan.didden]

Perhaps it's a matter of the degree of coupling between primary and secondary (passing noise)? John Curl was big on that, and it's certainly made me hesitate before using a toroid. Another possibility (and I haven't thought this one through, just shooting from the hip) is what happens near current peaks with the core; could that pass on garbage as it goes near (or past) saturation?

SY,

There is indeed coupling from primary to secondary through a toroid. Toroids are wideband transformers. Am I to understand that this somehow radiates or conducts into the amp to change the sound? Why are we still using toriods then in the first place?

I don't really believe in radiation from a toroid; the tight coupling that promotes the widebanded-ness also prevents it from radiating outwards. Conduction would be a possibility, but that would mean that whatever enters has to take the diodes, reservoir caps, local decoupling etc.

As for saturation effects, yes, that would lower the transformer induction temporarily. What would that do? If anything, the coupling would become less, so junk from outside would be more blocked.

I don't say that different transformers etc cannot influence the sound, but the cases where I have heard it myself are in some tests I've done many years ago with deliberately undersized transformers. I though, hey, the average power of an amp is only 5 watt or so, so I can get away with a tiny transformer. It did change the sound in the sense that the amp sounded much less powerfull and more distorting. At the time I concluded that the severe supply modulation in the peaks caused it, but I didn't proof it to my own satisfaction.

I suspect that here also, as everywhere in audio, psychological effects play a large role if not a dominating one.

Jan Didden

 

[SY]

I suspect that here also, as everywhere in audio, psychological effects play a large role if not a dominating one.

Let me start with saying that I'm in 100% agreement with this, which should come as no surprise to you.

But claims about changing sound with transformers at least aren't beyond the realm of physics (though there's zero actual evidence that the claim is true). People use toroids because they're compact, low profile, and don't radiate much. But the tradeoff is that they'll transmit garbage from the primary just as efficiently as they'll transmit 60 Hz (ok, 50 Hz for you cycle-deficient Euro-types).

 

[jan.didden]

Let me start with saying that I'm in 100% agreement with this, which should come as no surprise to you.

But claims about changing sound with transformers at least aren't beyond the realm of physics (though there's zero actual evidence that the claim is true). People use toroids because they're compact, low profile, and don't radiate much. But the tradeoff is that they'll transmit garbage from the primary just as efficiently as they'll transmit 60 Hz (ok, 50 Hz for you cycle-deficient Euro-types).

Cycle-deficient?? Well, I'll be...! You got nothing to brag about, you 110V-potential-challenged lot! Jeez! 110V, in Holland we have bike dynamo's that do better than that.

Seriously, if it is transmitted junk, which only increases for bigger transformers, why is it that 99.99% of the posters report BETTER sound with bigger transformers?

Jan Didden

 

[SY]

which only increases for bigger transformers

Is that true? I thought that bigger transformers, in general, had looser coupling at high frequencies.

If the hypothesis that this is all psychological is correct, it's not unexpected that Bigger = Better.

 

[Jay]

I suspect that here also, as everywhere in audio, psychological effects play a large role if not a dominating one.

Yes, and psychologically, we tend to think that expensive parts sound better.

I have plenty of expensive transformers that are still not used. The build is beautifull. And I'm an artist, in that I think audio is also an art. I like beautifull things (I used to use acrylic to cover my amps). It is bothering me that I love the "sound" of an ugly EI transformer from an old UPS while the beautifull $100 is sitting there unused.

Oh, the "best" sounding toroid I own is a toroid that is custom manufactured for NAD (made in Canada). "HOLMGREN" TRANSFORMER is written on it. I don't know what the hell is that

 

[mAJORD]

very intersting topic,

something that interests me is the attittude that THD , signal-noise , and damping factor, among other measurments are considered useless by some audiophiles.....

if any of u are reading, perhaps u could state what other factors make an amp sound 'good' or 'bad' , and why they cant be measured..


something i thought myself was how music unlike a pure simple sinewave signal, consists of almost endles signals superimposed on eachother , and perhaps THD measurements couldnt take into acount how well a given amp could amplify such a complex signal without any impurities, especially when everything was sitting on top of a hefty low freqency bassline , but I am still learning so perhaps THJD measurments take this into account... or what im saying is irrelevent.. ? ? ?

 

[jan.didden]

What you say is highly relevant. Most designers and researchers I have read or met agree that THD is just a rough indication of amp linearity. Much effort has been spend throughout the last decades to find a test that correlates with listening preferences. Two-tone intermodulation, is one of them, but for instance the AP systems contain provisions for doing 10 or more tone IM tests. Then there other tests mixing sinus and square wave (low pass filtered) to see how linear the sine is reproduced in the presence of large lf levels. As far as I am aware, progress has been made, but there is not yet a surefire test suite that correlates with listening preferences (under controlled conditions of course).

Jan Didden

 

[Frank Berry]

"there is not yet a surefire test suite that correlates with listening preferences (under controlled conditions of course)."

Absolutely correct. Even after all amplifier parameters are tested, listening (and listener) preferences are added to the mix.

 

[mAJORD]

thanks , that clears a few things up!.

 

[CRFX]

I am interested to know what effect "better" caps have in the PS. When they are in the signel path "better" caps should have a noticible effect. But in a PS what differnence does it make? Do they make a smother dc voltage?

 

[Jay]

I am interested to know what effect "better" caps have in the PS. When they are in the signel path "better" caps should have a noticible effect. But in a PS what differnence does it make? Do they make a smother dc voltage?

Many said that I’m a very technical person. But in audio, I am a different person. I can read all the theories on how caps work on DC (which I think I have), but what does it give???

One thing I taught myself is that in audio every parts work together in a very unique way. Trade-offs are things you will always encounter, so you must know what to trade and when. You may need a PS cap that have the smoothing ability, or you may want a transformer to do that for you (don’t ask me how) and let the cap to do other things such as serving the output transistor with current in bass passage when the toroid is so poor on this respect. But if your amp is tube you may want to forget that.

In PS, too smooth a DC (or very noise free) usually means a bad transient. A “good” cap is a cap that can do many thing so you don’t have difficulty with the trade-off. With “better” cap in PS I want a silent system but doesn’t harm the sonic. I will make sure that the relatively cheap rectifier does not function as the bottle-neck for this. I will make sure that the trafo is not underrated also.

Many fake capacitors have damaging effect on sound, but I haven’t found that exotic capacitor is justifiable (in term of increased performance per dollar). I can use thick wire for supplying the board and use a small Black Gate on the leg of the output transistor.

Oh, yes, if we think an audio signal path as tunnel of water, you’re right that in the signal path “better” caps should have a noticable effect. But there’s no such tunnel. That’s just an analogy to make things understandable. But you can expand the analogy to uh… multi-dimensional space where uh… forget it

 

[sss]

something i thought myself was how music unlike a pure simple sinewave signal, consists of almost endles signals superimposed on eachother , and perhaps THD measurements couldnt take into acount how well a given amp could amplify such a complex signal without any impurities, especially when everything was sitting on top of a hefty low freqency bassline , but I am still learning so perhaps THJD measurments take this into account... or what im saying is irrelevent.. ? ? ?

THD stands for total harmonic distortion
the thing is that even the most complicated music signal is made of many simple pure sinewaves

but thats not all

lets say u got a power amp with a gain of 20 and u are measuring the thd at 1Khz freq and the thd at that freq is 1%
now u are doing the same test but under load conditions and the thd is still 1%

if the gain of that amp under load is only 10 the thd metter wount show nothing!! the thd remains 1% even if the signal is attenuated

hope that helps u to understand things better

 

[CRFX]

Oh, yes, if we think an audio signal path as tunnel of water, you’re right that in the signal path “better” caps should have a noticable effect. But there’s no such tunnel. That’s just an analogy to make things understandable. But you can expand the analogy to uh… multi-dimensional space where uh… forget it

::scratches head::
probably shouldn't have read that right after work.
I'll just stick with some decent panasonics for now.

 

[jan.didden]

[snip]In PS, too smooth a DC (or very noise free) usually means a bad transient. [snip]

That's completely new for me! Can you explain this? Do you mean a smooth supply without load or under load?

Jan Didden