[Folsom]

Ok guys... Here's a link that starts to talk about it the physical effects, then wonders off into inductors.

http://www.allaboutcircuits.com/vol_1/chpt_15/1.html

Inertia isn't something you want to build with capacitors. Doing that creates electrical resonances that are strengthened. Here's a pic of a prototype I'm playing with. It's designed to reduce resonations instead of encourage them.

[G Georgopoulos]

the capacitor field is electrostatic produced by voltage rather than current(electromagnetic),the field is usually confined between the plates of the capacitor and there is little or not emissions...

[randytsuch]

So when I saw the thread title, I thought you were referring to how to connect caps electrically.

But then the link talks about the e and m fields around conducting elements.

And when I look closely at your pic, adjacent caps are at 180 degrees to each other.

So are you saying that you can hear, or measure a difference if the caps are mounted 180 degrees, versus mounting them all in the same direction?

And I am not trying to dispute anything, I'm just trying to understand the point. 

And if I'm getting it, it's a interesting point that I never considered before.  For transformers, I always try to mount them at 90 degree angles so the magnetic fields won't couple, but I didn't think about e fields and other types of devices.

EDIT:
I read the crown thread after I typed this, gave me a little more background into this thread.

Randy

[Folsom]

This is pretty OCD.

It's subtle. However my experience is reduction in resonations, even small ones, is pretty serious on quality of sound. Think about it, everyone is putting wood pucks under their XLS amps...

There's limits however. I could use a different combination of caps for less resonation, but low resonation caps don't necessarily sound better, in fact their often cheap crap.

[lowtech]

Wow.   

[Folsom]

Wow.   

Sometimes the subtle things keep you listening.

[*Scotty*]

I do not think I will fry that fish at this time, there are bigger fish still in the pond. Checking Nichicon Solid Polymer Electrolytic capacitor series through hole.. There is no continuity with either the positive or negative lead and case and there should be no external electrostatic field to cancel out.
Scotty

[Folsom]

Sure, I'm speaking primarily of magnetic. The cases on caps aren't thick enough for some LF though.

I wish you could hear the amp ATM.

[BobM]

A small barely discernable improvement here ... a small barely discernable improvement there ... if you get enough of these and add them all up you get an improvement that is noticeable and worth pursuing.

Plus, many other engineers will say things like "It's a basic circuit that I've seen in many products. So how did he get it to sound so much better?"

[Davey]

This is "The Lab" circle after all.  The onus on Salis should probably be to demonstrate differences objectively.  A simple baseline experiment using another board with capacitors orientated the other way should suffice.  A measurable capacitance parameter should change, yes?

Noting "barely discernible improvements" is simply subjective evaluation and doesn't prove anything one way or the other.

Cheers,

Dave.

[randytsuch]

This is "The Lab" circle after all.  The onus on Salis should probably be to demonstrate differences objectively.  A simple baseline experiment using another board with capacitors orientated the other way should suffice.  A measurable capacitance parameter should change, yes?

Noting "barely discernible improvements" is simply subjective evaluation and doesn't prove anything one way or the other.

Cheers,

Dave.

In a perfect world yes.  But I'm not convinced every audible change is measurable, or at least measurable by "common" test equipment.  In this case I would consider a decent oscope as that test equipment, but most diy people don't even have one of those.  A DMM wouldn't cut it.

But I would be interested in a little more description of the testing performed.  Such as what was powered with the circuit (I'm guessing a class D amp), total capacitance used, and subjective findings of "normal" versus 180 degree orientation.

Randy

Randy

[*Scotty*]

I am not sure if this is supposed to help reduce ringing in the power supply or not. If you were looking for ringing you would need a digital storage scope to catch the event. Ringing in the supply can be made much worse if measures are not taken to minimize the effects of inductance in the power supply board traces or wiring. It can also be a problem in analogue amps. A snubber circuit consisting of a small cap in series with a sub 1 ohm resistor across the caps will work wonders.
Looking for a change in capacitance value due to capacitor orientation on the board or board layout is probably a futile act. The total capacitance value is not the variable in play.
Scotty

[Davey]

Fair points.  Although his assertion comes from a somewhat different premise.

Regardless, some lab testing showing orientation effects utilizing whatever test equipment necessary would be enlightening.  Otherwise this is all just speculation and not pertinent to this Circle.
That's not to say there isn't anything to the speculation or that the effects can (or can not) be proved with bench testing.  But, IMHO, speculation should be left at the door when objective testing.

Dave.

[Folsom]

Measuring may be difficult, hearing isn't always, but prevention is simple. The basics in a PSU are just orientation such as I've been doing, and no extreme differences in sizing of capacitors.

The transformer is likely to have a resonance. The cheap way is to dampen on the secondaries, as you say Scotty. However the best way is on the primaries but there's no "standard" for that, its variable greatly based on what parts you have.

Sometimes it's just easier to compare some stuff by ear, or just do what's easy and won't result in a negative benefit. The more capacitors, the more the potential for inertia. For example it's a bigger deal for 22x capacitors in one of my conditioners.

Dave, I'd love to give you measurements but I don't have a DDO. But if what you say is true and that everything is objective and only of value if we can measure it, well shit we can all throw away our capacitors that cost more than $.02 and save thousands of dollars.

[Davey]

But if what you say is true and that everything is objective and only of value if we can measure it, well shit we can all throw away our capacitors that cost more than $.02 and save thousands of dollars.

Did I say that?  Boy, I hate those kind of straw man replies.  They're insulting.
Is it possible for posters to NOT drift off on tangents with their replies? 

I'm simply asking if you can demonstrate  this capacitor orientation effect with some lab/objective testing.  How difficult is that concept to grasp? 

Dave.

[*Scotty*]

The transformer can certainly ring but the power supply caps will also ring if the instantaneous currents are high coupled with significant inductance in the wiring to the output devices, this where a snubber right on the caps themselves comes into play. This is what I did on my DIY Tripath Eval board based amp a few years ago to good effect. I used 4 pole Jensen caps in that supply and the addition of the snubbers on each rail immediately adjacent last capacitor was quite audible. I only had two 10,000mfd caps per rail to deal with but I did have 6inches of power wiring to the point on the circuit board where the B+ and B- connections were made and nothing I could do about it.
Scotty

[shadowlight]

@Salis,

I am coming in blind (no electronic background either) here but looking at the pictures you are trying to alternate the orientation of the caps.  Looking at the second row it seems that the dark side of the caps are facing each other and shouldn't they be setup the same way where light  faces dark?

[*Scotty*]

This may be less a matter of how the caps are oriented and more about how the supply traces are laid out on the circuit board. If they are laid out to increase the capacitance the traces might have this would tend to mitigate the inductance the traces also have. When you put a capacitor bank together with the idea of using a bunch of smaller capacitors because they are faster than one big one, you immediately run into the law of diminishing returns due to the rapidly increasing inductance of the board traces needed to connect all of those capacitors together. That inductance acts to increase the high frequency impedance of the entire capacitor bank. Sometimes it will sound better with half as many capacitors in the bank as you would like to use. Local storage/decoupling capacitance right at the point use next to the output devices is also very helpful.
 This is one reason I went with Jensen 4 pole caps, very low HF impedance in one big package to minimize the power supply rail inductance problems.
Scotty

[Folsom]

They're on copper pours, power planes, however you like to refer. The inductance is very low. (It may get split, will do check soon on this, if each VCC pin is a channel)

Dave, let me borrow your DDO, I'll have a heyday of testing (my engineer would probably develop some new concept). I'm not trying to drift the conversation, I'm just saying you can't measure everything all the time and/or will the measurements tell you good/bad for sound in all circumstances. This thread is an offshootish from Crown XLS amps that can't tell you how they sound by measurement because they have lower distortion but vastly more detail than many people have heard. The relationship of the measurement doesn't coincide with anything useful to you; and this isn't an abnormal thing. I consider getting really nit-picky over numbers to be as off track as some other things, when there isn't anything to really gain. We're better off discussion the physics of what happens, and weighing experiences of the differences with some things (as brutally egotistical as that road can become, anyway).

Jensen 4 poles are great, but not as good as these guys this close to the chip. In a PSU they'd be ideal, if you can afford them. But the copper pours will give you pretty decent inductance performance at a price that may be more appropriate depending on what you're doing.

Shadow, they're correct. The switching of row patterns is on purpose.

[*Scotty*]

Its always a question of how close you can get the capacitance to the point of use and how much storage you need for adequate ripple suppression under transient loading. The circuit layout would have to be purposely setup to use the Jensens effectively due to their form factor.
So far the solid organic polymer caps are a very good cap for high speed circuits but I have noticed that unlike conventional electrolytics they have no tolerance for over voltage conditions, for that matter neither do the Jensens.
Scotty