ISO Regen performance Improvement Cheap!

[mansr]

9 minutes ago, Abtr said:

Just for my understanding, doesn't the 'corkscrew rule' indicate that if two wires are twisted in opposite direction then current in the same direction will generate opposing magnetic fields in both twisted 'coils' that will (partially) cancel each other out?

That would be the case with two counter-wound coils. In a twisted pair, both wires turn in the same direction.

[mansr]

40 minutes ago, sandyk said:

That is NOT normal, which is why I asked the question !

Normally the sound will at least come from in line with, or a little in front of the speakers.

That depends almost entirely on the room and speaker placement.

[mansr]

2 minutes ago, Audiophile Neuroscience said:

Yeh I have recordings where sound circles around the room. A neat illusion done with tricky phase manipulations in the recording so I am told IIRC

Yeah, you can't do that with plain microphones.

[mansr]

8 minutes ago, Ralf11 said:

I think she's on MQA

"This is your brain on MQA."

[mansr]

56 minutes ago, marce said:

How... How can a DC cable have such an effect on the sound?

I think we both know the answer to that.

[mansr]

1 minute ago, marce said:

Pathetic... \Just resorting to name calling instead of taking part in an open debate...

Some people react badly when their faith is challenged.

[mansr]

20 minutes ago, Ralf11 said:

would not know what is magic and what is real - <insert Isaac Asimov quote here>

Make that Arthur C. Clarke.

[mansr]

1 minute ago, BigGuy said:

Can someone, perhaps John, please explain again the working theories behind JSSG and JSGT.

 

I have been reading quite a lot (dangerous in itself) about generation and shielding of EMI and RFI and it seems like JSSG might be acting like a quasi Faraday cage.  The difference I see is that a Faraday cage, IIUC, is always grounded.  Rather than a wire or another shield (with insulating layer) attached to both ends of the first layer of post added shielding, would it not be better to attach that wire to a ground?

A Faraday cage doesn't need to be grounded.

[mansr]

27 minutes ago, Speedskater said:

Well the Copper Clad Steel (CCS) wire is easy. It's a poor choice for anything much below 100MHz.

How do you figure? Insofar material matters at all, it is at high frequencies. Copper clad steel combines tensile strength with low resistance. What problem do you see there?

[mansr]

31 minutes ago, BigGuy said:

Please bear with me as I try to better understand the JSSG(360).

 

I DO understand that a Faraday cage is a 6-sided box or otherwise fully enclosed volume, e.g., cylinder with ends, that does not need to be grounded to be functional.

 

When I queried JohnS about the JSSG(360) being a quasi Faraday cage, he politely indicated yes but I surmise that would be true only if the shield extended a very long  (infinite?) distance from either end of the shield.

A Faraday cage can have holes as long as they are small compared to wavelengths one wants to keep out. Electromagnetic waves of 1 GHz have a wavelength of about 0.3 m. The holes at the ends of the shield tube are not a problem unless your cable is a foot in diameter.

 

31 minutes ago, BigGuy said:

I know JohnS also said that the insulated wire or second shield with insulator provide a path for electrons to move within the shield.

That part doesn't make any sense to me. Whatever an external field does in the lengthwise direction of the shield it will also do to a wire running alongside it. The wire may as well not be there.

 

31 minutes ago, BigGuy said:

I would propose that the JSSG360 does comprise a Faraday shield (thinking of it as thermos bottle of sorts) but that what is "protected" is the insulating material between them.

It does that too, not that it's good for anything.

[mansr]

1 minute ago, Ralf11 said:

the thermos bottle quote is a bit confused - not analogous to a Faraday cage really

Yes, the purpose of the shiny coating (more likely aluminium than silver) is to reduce the emissivity. Nothing to do with EM waves.

[mansr]

13 minutes ago, BigGuy said:

IF a thermos is a double cylinder of glass, etc., I  believe it is a good approximation of two layers of braid connected to each other separated by a insulating layer in the context of our discussions, IMHO.  A true thermos is two cylindrical layers of glass closed at the end with a evacuated layer in between.

There is a topological similarity, but the analogy ends there.

[mansr]

5 hours ago, Cornan said:

The caps is charged and supplies its own isolated power. If the cables are attached into the plugs together with the legs there will most likely be a combination of power from both the wire and cap, so not 100% isolated unless the caps provide all the power in my view.

Sorry, but I can't let that bit of misunderstanding go unaddressed. If the electrical connection is good, it makes no difference whatsoever where on the pins the wires are attached. Just imagine cutting the pins between the attachment point and the cap itself. If you did this, the pin ends would simply be extensions of the wires, and current would flow freely. Having the cap attached doesn't change this.

[mansr]

8 minutes ago, Cornan said:

Interesting! Let´s assume that there is a positive charge present in the negative wire and a negative charge present in the positive wire of the cable. If I squeeze the wires into the plug together with the Kemet legs inside a screw terminal DC plug. Surely some of the wires will be in direct contact with the DC plug giving away a bit of negative charge on the positive wire av vice versa. The Kemet will still be charged ofcourse but cannot actually provide 100% of the power and therefore not ensure that both polarities are clean(er) from opposite charges. Charging the Kemets further up the legs will however ensure that much more clean power is provided (even if there is still some leakage from the actual caps) in my point of view.

Unless you're dealing with extremely high frequencies, many GHz, the capacitor pins and wire ends, no matter how they are attached, can be regarded as a point connection with zero resistance. Moving the attachment a few millimetres simply makes no difference.

[mansr]

5 minutes ago, Cornan said:

Well it does make a subjective SQ difference according to my own ears on my first cable version.

I can't argue with that, but the explanation must be something else.

[mansr]

4 hours ago, marce said:

Yes its acting as a decoupling capacitor....

Obviously. The question is how bad the downstream regulators need to be before the addition of an external capacitor makes an audible difference. If a £0.15 capacitor provides such a whopping improvement, why didn't the manufacturer include it in the design?

[mansr]

3 hours ago, marce said:

I do keep saying that in reality the whole power delivery system should be considered, from main supply to device power pins, for any sort of circuit analogue or digital, with digital often being more problematic because of the high initial current draw, so the decoupling is as critical as the main supply, if not more so. Whereas as audio frequency analogue the main supply plays a big part, though local decoupling is still required. 

I've dealt with a few power supply issues in poorly designed devices. None would have been remedied by adding a smallish capacitor at the external power input, which is what people are doing here.

[mansr]

5 minutes ago, marce said:

Electrolytic/polymer for the larger reservoir caps, X7R ceramic for local decoupling caps next to a devices pins... For digital the smallest possible package for a given value, package size being more critical than absolute capacitance. Analog IC's etc again ceramic X7R next to the pins. Crystals/Oscillators very small COG next to pins, larger X7R next to it.

Add point of load regulators for the most demanding parts.

[mansr]

3 minutes ago, marce said:

Yep and pi filter isolation of the main distributed power before POL regulators help.

And when you've done all that, an additional 560 μF cap on the input won't make an iota of difference.

[mansr]

8 minutes ago, Cornan said:

If I get you right @marce you would consider Kemet´s polymer caps at one side and XR7 ceramic close to the powered device to be a viable option to try on DC cables if I do not want to rebuild my whole power delivery system? ?

The ceramic caps go as close to each IC as possible, preferably less than 1 mm.

[mansr]

2 hours ago, Cornan said:

Thanks! ? Since I will solder them onto the actual DC cable plug terminalI guess that´s close enough? Any specific value that I should choose for for example 12VDC/1,5A (Brooklyn DAC)?

The ceramic (X7R) decoupling caps need to be part of the product design and placed as close as possible to each IC. Adding one at the main power input isn't likely to accomplish anything. If it does, you should throw out the DAC and buy a properly constructed one.

[mansr]

1 hour ago, Cornan said:

So back to square one? Kemet’s cannot possibly improve because Brooklyn DAC is not properly constructed in your opinion? So LS-HPULN, BluWave USB to Spdif board & LPS-1.2 are not properly done either when the Kemet’s do imorove there as well? And the router? And Aqvox switch? And ISO Regen? And Aries Mini? They all improved with this tweak! 

You're the one saying adding an external capacitor improves the sound. Taking that at face value, the logical conclusion is that all those devices have inadequate internal power regulation.

[mansr]

3 minutes ago, Cornan said:

Logics does’nt always make sense!

Actually, it does. It's the definition.

[mansr]

1 hour ago, Abtr said:

According to @marce the quality of the main supply is as critical (for SQ) as is local decoupling and DC buffering close to each IC. The whole power delivery system has to be addressed. That would imply that if the main supply is inadequate then no amount of local DC buffering could remedy that. Now since polymer aluminum capacitors (such as Kemet A750) have very low ESR, low ESL and large capacitance, they have excellent response to transient changes in the load. So adding one directly at the audio device's DC input (close to the circuitry) makes sense and may well improve practically any (external) DC power supply and may therefore improve the *overall* performance of an audio device.. 

The further you get from the ICs, the less variable the current should be. In a typical design, there will be multiple regulators and a dozen or so capacitors between the 12V input and the critical consumers of power. The external supply needs to provide the required current without too much ripple. A marginal power supply can sometimes be made to work with additional capacitors. I guess the LPS-1.2 is marginal then.

[mansr]

1 minute ago, Cornan said:

I wonder how the Kemets would be like in a USB cable? A bit tricky to get that tested, but I will find a simple way sooner or later! ?

For USB powered devices? Don't do that. The spec only permits a maximum of 10 μF load capacitance. If you add more, it might trip the over-current protection when plugged in.