Upgrading POWER Cord Theories and Discussion [Text View]

jneutron

05-13-08, 10:35 AM

What I would really like to get is a scientific explanation of how it would be possible for a power cord change to produce an audible difference.

A power cord completes the ground loop between two pieces of equipment. Any time there is a loop, there is the possibility that the loop can intercept some time rate of change magnetic field. Faraday's law of induction defines the amount of voltage that can be generated within the loop.

This is very often the culprit when hum is present. Single ended systems are always going to be susceptible to this phenomena.

If the power cord alters the ground loop resistance or alters the amount of coupling to extraneous fields, it can cause a difference. Hum is the most "popular" artifact to be created.

The idea of it is so wacky it fits right in with the worst of the high end audio tweaks.
If one considers only the path through the supply, then yes it is indeed wacky..PSRR is a well known parameter that the designers worry about.

How on earth do they explain it since they certainly don't chalk it up to bias?
Mostly by ridiculous and magical explanations that have nothing to do with the issue.
Oh, and by the way, power cords don't affect system sonics in a bias controlled listening test. That's a fact, not an opinion.

Actually, if you put a power cord in and get hum, that is factual and shows a difference.

Other than hum changes, I cannot ascribe an effect. Maybe yes, maybe no.. But not within my admittedly limited experience.

Cheers, John

scientest

05-13-08, 11:15 AM

A power cord completes the ground loop between two pieces of equipment. Any time there is a loop, there is the possibility that the loop can intercept some time rate of change magnetic field. Faraday's law of induction defines the amount of voltage that can be generated within the loop.

So how are two different power cords going to alter the characteristics of a ground loop? You either clip the ground and/or reverse the polarity or use an isolation transformer, all of which are physical changes having nothing to do with actual materials or construction of the power cord.

jneutron

05-13-08, 11:28 AM

So how are two different power cords going to alter the characteristics of a ground loop? You either clip the ground and/or reverse the polarity or use an isolation transformer, all of which are physical changes having nothing to do with actual materials or construction of the power cord.

1. Resistance. The total loop current will be a function of the loop voltage, the loop resistance, the total integrated flux, the loop inductance. If the new line cord has higher resistance, more of the IR drop will be across the line cord ground. (remember, aftermarket cords are not constrained to meet NEC guidelines for safety ground conductor ampacity or fault capacity.) If the new line cord has lower resistance, it forces more of the loop voltage to be present across either the other line cord, or the IC between the units.

2. Twist pitch. Zero pitch is the worst configuration for broadcasting a dipole magnetic field. A power amp generating high current haversines will couple more if it has no twist pitch. A twist pitch which is identical to a source cord twist pitch can couple more, the extent dependent on the relative pitch phase..remember, the ground conductor also twists helically in space.

3. Conductor spacing. Thicker insulation spaces the hot/neutral farther apart, the farther apart the higher the inductance. The definition of inductance is: the relationship between the current within the system (wire) and the amount of energy stored in the field generated by the wire (system).. E = 1/2 L I^2. The farther apart the wires, the more magnetic field it produces to couple to external systems. Teflon is touted at times, but teflon scares me because of creep, and because it is typically thinner than other insulations..but it can tend to provide lower inductance cords.

Cheers, John

scientest

05-13-08, 11:57 AM

1. Resistance. The total loop current will be a function of the loop voltage, the loop resistance, the total integrated flux, the loop inductance. If the new line cord has higher resistance, more of the IR drop will be across the line cord ground. (remember, aftermarket cords are not constrained to meet NEC guidelines for safety ground conductor ampacity or fault capacity.) If the new line cord has lower resistance, it forces more of the loop voltage to be present across either the other line cord, or the IC between the units.

Oh come on. You've got what 3, maybe 6 feet of power cord vs. 40, 100 or more feet of Romex in the wall. The only way the higher end cord could measure different to any significant degree was if it had a _higher_ resistance and this would have to be enough to essentially turn it into a heating element.

2. Twist pitch. Zero pitch is the worst configuration for broadcasting a dipole magnetic field. A power amp generating high current haversines will couple more if it has no twist pitch. A twist pitch which is identical to a source cord twist pitch can couple more, the extent dependent on the relative pitch phase..remember, the ground conductor also twists helically in space.

Again any contribution would be impossible to measure even if the the in wall cable contributions didn't totally swamp it.

3. Conductor spacing. Thicker insulation spaces the hot/neutral farther apart, the farther apart the higher the inductance. The definition of inductance is: the relationship between the current within the system (wire) and the amount of energy stored in the field generated by the wire (system).. E = 1/2 L I^2. The farther apart the wires, the more magnetic field it produces to couple to external systems.

Ditto, the above; the delta in the inductance between the two cables (the I in your equation above) will be swamped by the in wall power cables and insignificant in any case.

Seriously, if you actually understand this (and aren't just copying from somewhere) you've got to know that there's no way the delta of any of these contributions would be significant if it can even be measured?

J_Palmer_Cass

05-13-08, 12:08 PM

Hey, here is total proof that power supply cable mods work. If you have any doubts about the solid science behind this, post your response on the following thread.

http://www.agoraquest.com/viewtopic.php?topic=26561&forum=51&keyword2=electronics

Quoted from the link above:

"There was a new ultra tight integrity to the Harmonic Structure and the creation of lingering decay envelopes that transverse the entire soundstage instantaneously. The tonal character had become rich, lush and more full of instrumental bloom due to a substantial increase in dynamics provided by more instant access to the AC current now coming through the new Hospital Grade plugs."

"One of the most welcome improvements was that the Music was more mellifulous and natural sounding than I had ever heard it before."

"One thing I wanted to mention specifically about the effect that the new Hospital Grade AC Plug had on the Sony CX300 Megachanger and that was that it removed a layer of Digititus from the presentation which gave a new super smooth and relaxed feel to the Music that I have only heard on one other occasion and that was back in 2000 when I heard a Sota Turntable driving a pair of Cary SLAM 100 Tube Monoblocks. I never imagined that replacing a single molded plug on a stock captured cord on a piece of Sony equipment would have this much of a dramatic effect in my Audio system."

Bob Lee (QSC)

05-13-08, 12:12 PM

Hum problems, especially those that are exacerbated by a power cord, almost always indicate a pin 1 problem.

Bob Lee (QSC)

05-13-08, 12:14 PM

jneutron

05-13-08, 12:40 PM

Oh come on. You've got what 3, maybe 6 feet of power cord vs. 40, 100 or more feet of Romex in the wall. The only way the higher end cord could measure different to any significant degree was if it had a _higher_ resistance and this would have to be enough to essentially turn it into a heating element.

Why a heating element??

We are discussing the guage and resistance of the ground cable.. we are ignoring the ampacity of the hot and neutral.

Remember, it is not all the way back to the panelboard we are talking about. We are talking about the cords to the duplex outlet. It is there that the grounding scheme is attempting to force the system to refer all signals to.

Again any contribution would be impossible to measure even if the the in wall cable contributions didn't totally swamp it.

Contribution (re twist pitch) is quite easy to measure actually. I do not know from what experience you are speaking from. I do it for a living..(well, it's one aspect). Granted, most of my effort is currently geared towards 650 mcm and 250 mcm MV cables and switching supplies pushing 85 Khz, but maxwell cares not how big or small the wire...:p

Edit: almost forgot.:eek:.Cat5e upped it's bandwidth by going to different twist pitches for the four linepairs, this eliminated the pair to pair coupling that previously limited the cable's bandwidth..

Ditto, the above; the delta in the inductance between the two cables (the I in your equation above) will be swamped by the in wall power cables and insignificant in any case.

Edit: The inductance in that equation is the "L"...the "I" is the current..

Again, you need to sit down and re-think your argument. We are discussing a power delivery system which has attempted to force the equipment to consider the wall outlet ground as the signal ground. Please stop with the "miles and miles" argument which you are tending towards.

The inductance of a line cord is entirely dependent on it's geometry. Ignoring the internal 15 nH per foot inductance of a cylindrical wire for the moment, a typical 12 guage wire will have about 120 nH per foot, at almost zero spacing. Put it at 250 mils, that goes to roughly 180 nH per foot (this is a typical spacing for a UL #12 cordage..). If you wish, I can put the full terman equation up..

1.1 uH for a 6 foot cord gives what voltage drop at 60, 180, 300, 420, 720 hz, for whatever haversine is caused by the cap bank (remember, stiffness here alters the haversine components of the line draw..

Take that inductive reactance, calculate the possible reactive voltage, then HALVE that (the most a line cord can couple to the ground loop is half it's inductive reactance, so the max effective turns ratio is 2 to 1..

Take that voltage, and put it on the input of the poweramp. This is the worst possible case of coupling...an upper limit..

Seriously, if you actually understand this (and aren't just copying from somewhere) you've got to know that there's no way the delta of any of these contributions would be significant if it can even be measured?

All I can do is show you the door, you have to choose to step through it... (gotta love that line):p

We are talking about a mishmash of consumer equipment, most of it poorly designed with respect to ground loop isolation techniques. And equipment capable of 10 e2 to 10e3 watts output controlled by an input referred to some arbitrarily defined and controlled ground, by an input power of about 10e-4 watts. (ten Kohms/1 volt)

10e6 to 10e7 power gains..from chassis' which more often than not, looks like a bowl of spagetti?

If I copy from somewhere, anywhere, I footnote the source or provide the link. What I speak of comes from experience.

(sorry I didn't put units on the graph...X axis is inches, Y axis is nanohenries per foot. The graph stops at the wires touching..)

Cheers, John

ps.. A lot of this type of information is within IEEE STD 1050-2004, but they don't tend towards much of the equation stuff, just generic discussion of aggressor and victim circuits and generalities..

jneutron

05-13-08, 12:43 PM

Hum problems, especially those that are exacerbated by a power cord, almost always indicate a pin 1 problem.

Hi Bob.

Pin 1?? Which part of the RCA is that???;)

Seriously, I've limited my discussion to unbalanced consumer grade..

Cheers, John

Chu Gai

05-13-08, 12:58 PM

I conducted a seance with Maxwell earlier, John. He had the following comments for you. First, he assures you that he is quite stiff right now and it has nothing to do with haversines. Lastly, and most importantly, he wanted me to tell you this: "No matter how much you shake, wiggle, and dance...the last electron stays in the amps."

jneutron

05-13-08, 03:27 PM

Let's assign some numbers.

Line cord at .2 uH per foot. (includes the internal inductance).

6 feet, that is 1.2 uH total.

10 amperes at 60 hz...(sorry for the shabby notations..)

I = 10 sin(wt) = 10 sin(2 pi 60t)

dI/dt = 10 * 2 pi * 60 cos(2 pi 60t) = 1200 pi cos()

1200 pi = 3769.8

L di/dt = 3769.8 * 1.2 10e-6

= 4523 *10e-6 = 4.523 * 10e-3

4.5 millivolts is the inductive voltage of the full line cord..so, half that is 2.25 millivolts at 60 hz.

Haversines are odd harmonics, so let's assume 10 amps at 180 hz...(the actual haversine spectra depends on the source impedance even before the panelboard, the total inductances all the way up to the wall outlet, the line cord of course, the xfmr, caps, diodes, supply layout...)

At 180 hz, the reactive component is 3 times, or about 6.6 millivolts of ground bounce.

That, referred to a 1 volt signal is .6% of the input. And not spectrally related to the input signal, but instead, proportionally related to the delivered AC power to the amp.

It's magnetic coupling, so copper braid is useless.

Edit:notes:

1. I used #12 cordage, and assumed a reasonable spacing.
2. I assumed half the magnetic field was available to couple..a twisted cord will reduce that coupling...
3. I assumed there were NO supply diode switching transients. (not a good assumption, even for 2 uSec product).
4. NO accomodation whatsoever was made for any IEC style termination filters anywhere, the biggest concern would be the hardening of the 60 hz line pair at the expense of the ground integrity..the common mode choke would be independent of that.

Cheers, John

scientest

05-13-08, 03:34 PM

Why a heating element??

Because in order for the difference in resistance between two power cords to be significant you've got to have some large delta compared to the rest of the system. You're not going to have significantly different lower resistance, it's gotta be a higher resistance that make the difference significant.

We are discussing the guage and resistance of the ground cable.. we are ignoring the ampacity of the hot and neutral.

If you're ignoring the contributions of the hot and neutral then inductance etc. is irrelevant. Either you're looking at the system as a whole or you're not, you can't have it both ways. Even if you want to look at the resistance of the ground in isolation I can't see how two different power cords are going to have any large differences (unless, as I said, one has a high enough resistance to start getting dangerous)?

Remember, it is not all the way back to the panelboard we are talking about. We are talking about the cords to the duplex outlet. It is there that the grounding scheme is attempting to force the system to refer all signals to.

Uh, no. You can't ignore the rest of the system. The "grounding scheme" doesn't suddenly start or stop at the end of the power cord. For 120 VAC the ground at the outlet is not a reference point at all (unless it's missing....).

Contribution (re twist pitch) is quite easy to measure actually. I do not know from what experience you are speaking from. I do it for a living..(well, it's one aspect). Granted, most of my effort is currently geared towards 650 mcm and 250 mcm MV cables and switching supplies pushing 85 Khz, but maxwell cares not how big or small the wire...:p

Edit: almost forgot.:eek:.Cat5e upped it's bandwidth by going to different twist pitches for the four linepairs, this eliminated the pair to pair coupling that previously limited the cable's bandwidth..

Again, how's any of this going to be relevant to power cords? (I see we're now including the hot side once more, but no matter). In particular, how is the difference between two power cords going to be significant?

(As far as my experience goes: I've got a couple of years of EE for computer systems many, many years ago.)

Edit: The inductance in that equation is the "L"...the "I" is the current..

Whoops, I was in a hurry, but the question remains; how is there going to be a _significant_ difference in the inductance of two different power cords?

These things might make a difference to the propagation of a signal at several MHz but they simply don't have a _significant_ contribution to the delivery of 120 VAC at 60Hz.

Again, you need to sit down and re-think your argument. We are discussing a power delivery system which has attempted to force the equipment to consider the wall outlet ground as the signal ground.

You might be discussing that. If so, that's pretty much useless; a power cords isn't a "power delivery system" and a wall outlet isn't a ground.

Please stop with the "miles and miles" argument which you are tending towards.

Are you suggesting the rest of the system doesn't come into play? Do you _really_ think there is _any_ point in treating the power cord as complete system by itself?

The inductance of a line cord is entirely dependent on it's geometry. Ignoring the internal 15 nH per foot inductance of a cylindrical wire for the moment, a typical 12 guage wire will have about 120 nH per foot, at almost zero spacing. Put it at 250 mils, that goes to roughly 180 nH per foot (this is a typical spacing for a UL #12 cordage..). If you wish, I can put the full terman equation up..

1.1 uH for a 6 foot cord gives what voltage drop at 60, 180, 300, 420, 720 hz, for whatever haversine is caused by the cap bank (remember, stiffness here alters the haversine components of the line draw..

Take that inductive reactance, calculate the possible reactive voltage, then HALVE that (the most a line cord can couple to the ground loop is half it's inductive reactance, so the max effective turns ratio is 2 to 1..

Take that voltage, and put it on the input of the poweramp. This is the worst possible case of coupling...an upper limit..

So, what's the delta between the best possible case and the worst and how does it compare to the contribution of the rest of the power system? IOW, what is the max. voltage drop you can actually see over a 6' power cord and the minimum (for your ideal geometry) compared to the voltage drop over the rest of the system (let's say 100 ft of Romex to keep it simple)? Take your parallel 12 gauge example, the best case / worst case delta of maybe 700 nanohenries (for 6 ft or so) isn't going to translate into a real voltage drop compared to the rest of the system is it?

All I can do is show you the door, you have to choose to step through it... (gotta love that line):p

This is one door I've been thorough in the past; nothing on the other side, nothing at all.

We are talking about a mishmash of consumer equipment, most of it poorly designed with respect to ground loop isolation techniques. And equipment capable of 10 e2 to 10e3 watts output controlled by an input referred to some arbitrarily defined and controlled ground, by an input power of about 10e-4 watts. (ten Kohms/1 volt)

10e6 to 10e7 power gains..from chassis' which more often than not, looks like a bowl of spagetti?

Again, what does any of this have to do with the _difference_ between two power cords? The power cords isn't going to suddenly clean up the most significant portions of the mess. You're attacking something that contributes less than, what 0.01% of the problem?

Here's the relevant experiment; find some nice clean laboratory grade power supply and hook up 50 ft of Romex and plop a receptacle on the end of it with the other end hooked up to your reference power source. Now put a regular 6 ft power cable between your receptacle and some piece of equipment of your choosing. Now measure the signal delivered to your piece of equipment inside the chassis in comparison to the original reference for whatever parameters you think will matter. Swap out your cable and replace it with another "audiophile" cable and repeat the measurement. Do you _really_ think you you'll be able to measure _any_ difference between the two cables? Unless the "audiophile" cable is doing something to degrade the power supplied (and it very well may be), it just won't happen.

Bob Lee (QSC)

05-13-08, 06:04 PM

Hi Bob.

Pin 1?? Which part of the RCA is that???;)

Seriously, I've limited my discussion to unbalanced consumer grade..

Cheers, John

Devices with unbalanced interconnections also often suffer from the equivalent of the "pin 1 problem," the existence of which spurred the development of the AES48 recommended practice. (Of course, unbalanced connections also don't have the inherent noise immunity of a balanced connection.)

jneutron

05-13-08, 09:38 PM

Devices with unbalanced interconnections also often suffer from the equivalent of the "pin 1 problem," the existence of which spurred the development of the AES48 recommended practice. (Of course, unbalanced connections also don't have the inherent noise immunity of a balanced connection.)

It is unfortunate that consumer electronics does not follow the practices of professional equipment.

Cheers, John

jneutron

05-13-08, 09:44 PM

Here's the relevant experiment; find some nice clean laboratory grade power supply and hook up 50 ft of Romex and plop a receptacle on the end of it with the other end hooked up to your reference power source. Now put a regular 6 ft power cable between your receptacle and some piece of equipment of your choosing. Now measure the signal delivered to your piece of equipment inside the chassis in comparison to the original reference for whatever parameters you think will matter. Swap out your cable and replace it with another "audiophile" cable and repeat the measurement. Do you _really_ think you you'll be able to measure _any_ difference between the two cables? Unless the "audiophile" cable is doing something to degrade the power supplied (and it very well may be), it just won't happen.

As I suspected, you are not understanding the discussion. Your "experiment" shows that quite well.

PSRR is power supply rejection ratio. The ability of the active electronics to ignore so to speak, any variations in the supply voltage.

Line droop, or sag, is the lowering of the delivered voltage as a result of the current draw upon that line. The NEC recommends (but does not mandate) that wiring runs have a maximum drop of 5% (or 4, I can't recall at the moment) as a result of the branch circuit pulling the maximum load.

That is what you are speaking of.

But that is not what I am talking about.

Go back and re-read all my posts...you need to first understand what I am speaking of before you continue to fire shots in the wrong direction.

Tomorrow, if you still don't understand, I'll try to draw you some pictures to help you understand.

Cheers, John

scientest

05-13-08, 10:30 PM

As I suspected, you are not understanding the discussion. Your "experiment" shows that quite well.

Given, that as far as I can tell the discussion is about how an "audiophile" power cord is going to somehow change the quality of the power delivered to a piece of equipment can you explain what other experiment would be possible?

PSRR is power supply rejection ratio. The ability of the active electronics to ignore so to speak, any variations in the supply voltage.

Line droop, or sag, is the lowering of the delivered voltage as a result of the current draw upon that line. The NEC recommends (but does not mandate) that wiring runs have a maximum drop of 5% (or 4, I can't recall at the moment) as a result of the branch circuit pulling the maximum load.

That is what you are speaking of.

No. You're missing the point.

But that is not what I am talking about.

Go back and re-read all my posts...you need to first understand what I am speaking of before you continue to fire shots in the wrong direction.

Tomorrow, if you still don't understand, I'll try to draw you some pictures to help you understand.

I understand perfectly well what you're point is. I just don't think it's relevant. You've yet to explain how it is.

jneutron

05-14-08, 08:50 AM

Given, that as far as I can tell the discussion is about how an "audiophile" power cord is going to somehow change the quality of the power delivered to a piece of equipment can you explain what other experiment would be possible?

Quite incorrect. You really do need to go back and re-read the entire thread starting from the question that I responded to...I'll paste it here again....now read carefully.

What I would really like to get is a scientific explanation of how it would be possible for a power cord change to produce an audible difference.

Now, where in that question, is there any (and I repeat ANY) mention of the words "quality of the power delivered to a piece of equipment"????

As I said, go back and read the thread again. Read my posts CAREFULLY.

Get back to us when you understand what I am referring to. AND IT'S NOT POWER QUALITY...

I understand perfectly well what you're point is. I just don't think it's relevant. You've yet to explain how it is.

Given your words within your last two posts, you've made it perfectly clear what you think is being discussed, and what you think has nothing to do with the context of the discussion.

Sit back, compose yourself, re-read my posts, when you are ready, we can discuss.

Given your misunderstandings, you really should be asking a lot of questions.

Cheers, John

scientest

05-14-08, 09:40 AM

Quite incorrect. You really do need to go back and re-read the entire thread starting from the question that I responded to...I'll paste it here again....now read carefully.

No John, I really don't need to go back and read anything. You're trying to reframe the problem in terms of something you deal with every day and in the process of doing so you're keep missing the fact that the issues you raise are simply irrelevant for consumer audio equipment. There are three reasons for this:

1) Go look at the back of your equipment racks and count the number of power cords that are three conductor and are tied to a single socket. I have none, zero, zilch, nada, three conductor power cords in my main rack and one in my system as a whole on my HTPC. In the past it was more likely to find people with an amplifier with a three wire ground but that was it, nothing else had a three wire ground. For 99% of consumer audio equipment no one is pushing a star point back to the receptacle.

2) Even if someone does in fact have multiple three wire power cords going to a single outlet you have yet to show how the _difference_ between two power cords could make significant contribution to the overall signal quality unless one of the cords was defective in some way. You can't treat the power cord as an isolated component in such a case since the problem you are addressing arises from the fact that there is a second ground path running parallel to the power cord. Although I'm rusty on this point, as far as I remember isn't this second path essentially going to look very close to a short circuit relative to the power cord?

3) Go back and read the reviews of these "audiophile" power cords. Almost without exception the reviewers are using them with dedicated circuits! As I keep pointing out the contributions of rest of the system simply swamps the power cords at this point.

This is why I'm pushing back the simple experiment as the only thing that is worth discussing. It's not that your points don't apply in general, they just don't apply to the problem at hand.

jneutron

05-14-08, 10:27 AM

No John, I really don't need to go back and read anything. You're trying to reframe the problem in terms of something you deal with every day and in the process of doing so you're keep missing the fact that the issues you raise are simply irrelevant for consumer audio equipment.

Ground loop hum doesn't exist? After all, it is the first indication that a loop exists. It is a very real problem, with very real solutions. Ignoring it doesn't make it either irrelevant or non-existant.

There are three reasons for this:

1) Go look at the back of your equipment racks and count the number of power cords that are three conductor and are tied to a single socket. I have none, zero, zilch, nada, three conductor power cords in my main rack and one in my system as a whole on my HTPC. In the past it was more likely to find people with an amplifier with a three wire ground but that was it, nothing else had a three wire ground. For 99% of consumer audio equipment no one is pushing a star point back to the receptacle.

1. You need to review the entire concept of house wiring..
2. You need to review the entire concept of IEC filter theory.
3. You need to review the entire concept of agressor and victim loops and how they pertain to low signal level interference..

2) Even if someone does in fact have multiple three wire power cords going to a single outlet you have yet to show how the _difference_ between two power cords could make significant contribution to the overall signal quality unless one of the cords was defective in some way. You can't treat the power cord as an isolated component in such a case since the problem you are addressing arises from the fact that there is a second ground path running parallel to the power cord. Although I'm rusty on this point, as far as I remember isn't this second path essentially going to look very close to a short circuit relative to the power cord?

You are correct. You are indeed VERY rusty.

3) Go back and read the reviews of these "audiophile" power cords. Almost without exception the reviewers are using them with dedicated circuits! As I keep pointing out the contributions of rest of the system simply swamps the power cords at this point.

Again, you are very rusty. Who cares what the reviewers and vendors are saying.. Read my posts accurately please.

This is why I'm pushing back the simple experiment as the only thing that is worth discussing. It's not that your points don't apply in general, they just don't apply to the problem at hand.

Your experiment is useless for a discussion of signal ground integrity.. Line sag is trivially accounted for by the unit PSRR's, and regardless of what you are saying, has nothing to do with the discussion at hand..

Re-read the posts please..

Here...I'll attach a typical house wiring diagram from the pole to the duplex for you to see..I'll detail some aspects in a subsequent post.

Cheers, John

ps. Sorry, the font is rather small. And, I had ta use blue for neutral for obvious reasons..

jneutron

05-14-08, 10:39 AM

In my area, east coast usa, we use single bushing transformers. This means that the pole transformer primary gets most of it's neutral current from the common neutral that sits below the transformers on the pole. Some of it's current does however, go through the earth...every second or third pole is earthed. Out in California, I noticed they use two bushing transformers, so areas of the country can be different.

The house breaker panel internally busses both the neutral and the earth ground. All house outlets have neutral and ground tied to a common buss.

Any two outlets which are not part of the same circuit breaker, as I depicted, has a whopping loop that is formed by the two circuit grounds. If you plug two units one into each duplex outlet, then you have to worry about that loop picking up stray magnetic field. If the units have internally, any mechanism which ties internal chassis ground or signal ground to both hot and neutral (for example, line noise filtering components or transformer capacitance), you can form a complete loop for extraneous signals. Remember, this need not be a hard connect for signals to follow.

Note that even in one outlet string (multiple duplex's fed by one breaker), if you use a DVM to measure the voltage from one safety ground to another, you will typically find voltage. EVEN IF THERE IS NO CURRENT FLOWING IN THE GROUND.. This is due to the fact that your dvm leads complete a loop of conductor with the grounds, and that loop is picking up stray 60 hz field. In my house, I've measured anywhere from 400 mV of voltage ground to ground up to 1.5 volts.

Try it anywhere in your house... Even though you may get 200 mV to 1000 mV of voltage across 20, 30 feet of #14 or #12 wire, remember that there is NO CURRENT flowing within it. you are measuring a transformer's secondary voltage. Note that that transformer is actually an air core, one turn secondary, and the measured voltage is the result of whatever magnetic flux (rate of change of it actually) the loop is seeing as passing through. (faraday's law of induction).

Next, a drawing of one duplex with two units...

Cheers, John

jneutron

05-14-08, 10:48 AM

Here's the pic of two units being fed by one outlet.

The agressor loop in this case, is the amp power cord. It is radiating a dipole magnetic field around it as a consequence of the current draw of the amp. It is kinda pink in color...and because I depicted the poweramp ground as in the middle, the agressor and victim loops overlap. It is also obvious by inspection that the maximum flux of the agressor loop that can be trapped by the victim, is one half of the power cord's flux...this is because the ground of the power cord is (typically) along the midplane of the powercord between the hot and neutral, even if it is twisted.

Note that I stuck an IC between the amp and the source unit. Note that I show the IR drop as occurring entirely within the resistance of the IC shield.

Note that I show the ground currents simply as flowing through the unit chassis'.

Even with differential circuitry typical of professional balanced equipment, amplifier designers still have to contend with the possibility of chassis currents flowing. Some really well respected guy (I'm afraid I cannot recall his name) described a test somewhere to detect sensitivity of an amp to this chassis current. He used a 12 volt transformer, a diode, and a resistor, to force half sine current into the balanced connector ground through the amp chassis, to detect the amp's ability to reject the common mode CURRENT...not voltage...current.

A well thought out test...maybe Bob or speco will remember his name..

By inspection, there are several methods that can be used to mitigate the problem.

1. Make the amp and source impervious to the chassis currents. (balanced is a good start)
2. Kill the loop by feeding the source power from the amp chassis daisy chain style. (this is a technique I used very successfully for a PA application where I had the source 125 feet from the amp, and still ran unbalanced line levels to the amp twisted around the power cord feeding the source....unfortunately, it is limited to applications which are fed from one breaker, 20 amps..
3. Break the ground (NOT RECOMMENDED for safety reasons).
4. Twist the powercord to force the integrated trapped field to zero(limited to integer twists, unless you put the amp on it's side).
5. Reduce the powercord external magnetic field.
5a. Coaxial cord (watch out for ampacity ratings)
5b. multiple conductors for hot and neutral
5b1....remember, each conductor must have the ampacity to clear the breaker in the event of a fault. The use of individually insulated wires, like a dozen #22's, will have the ampacity but if one of the wires shorts to ground or neutral, the one wire may not be sufficient to clear the breaker before a fire.

Note: It must also be noted, that if two wire cordage is being used, both the agressor and the victim loops still exist physically....the only change would be the exact paths for error signals to take... After all, how many people out there have been bitten by ground loop hum even though they had nothing but two wire cordage???

ps...I specify ground loop hum, as I believe all humans have had that experience at one time or another..but in reality, we've all had to deal with dimmer noise, motor contactor noise, various transients, all getting into our audio equipment. This victim loop will indeed pick all that crap up and provide us very clear audible indications of the event. As is also clear by inspection, any noise caused by this loop, cannot be stopped by anything that is done to the power supply of the amplifier, as that is not the path that the error signals are taking to get into our music..

pps...scientest...you have mistaken 30 years of practical experience in the home audio and pa environment with 15 years of experience in a very hi-tech work environment. Luckily, what I do there is extremely valuable and entirely applicable to both consumer and professional audio technology. It's what I bring to the table, a good add on to the valuable experience of others who post here..

Cheers, John

scientest

05-14-08, 12:49 PM

Ground loop hum doesn't exist? After all, it is the first indication that a loop exists. It is a very real problem, with very real solutions. Ignoring it doesn't make it either irrelevant or non-existant.

Never said that. The _only_ point I'm trying to make is that swapping out one power cord for another doesn't contribute enough to the equation to make any meaningful difference to the problem.

1. You need to review the entire concept of house wiring..
2. You need to review the entire concept of IEC filter theory.
3. You need to review the entire concept of agressor and victim loops and how they pertain to low signal level interference..

Tell you what, first explain how the difference between two power cords can make a difference with respect to any of these? I've gone back and re-read your posts. At some points you want to look at only the power cords between the equipment and then in other places you want to look at the entire system all the way back to the power company. I'd certainly agree with you that if you really want to know what's going on then it's the latter case that matters. But I'd also argue that in this case the power cords to your equipment simply isn't relevant compared to the problems in the rest of the system.

You are correct. You are indeed VERY rusty.

So explain what I'm missing here? Are we looking at two pieces of equipment with common grounds or not?

Again, you are very rusty. Who cares what the reviewers and vendors are saying.. Read my posts accurately please.

The people who care about what the reviewers and vendors are saying are the people buying the equipment and reading theses forums. Its impossible to read your posts "accurately" simply because they don't include enough information for anyone not doing your work on a day to day basis to tell what factors you are or are not taking into account on any given post.

Your experiment is useless for a discussion of signal ground integrity..

Agreed. That's not the issue.

Line sag is trivially accounted for by the unit PSRR's, and regardless of what you are saying, has nothing to do with the discussion at hand..

It's got nothing to do with the issues you raise, but you've yet to explain how the issues you raise are relevant...

Here...I'll attach a typical house wiring diagram from the pole to the duplex for you to see..I'll detail some aspects in a subsequent post.

Don't bother on my account. This simply isn't relevant.

Bob Lee (QSC)

05-14-08, 01:20 PM

It is unfortunate that consumer electronics does not follow the practices of professional equipment.

It is indeed, but those deficiencies will not be adequately addressed by voodoo power cords, either.

J_Palmer_Cass

05-14-08, 01:21 PM

You should've warned us to wear boots before opening that. :rolleyes:

Boots will not help. You need full body protection. Good thing to know that power cables have to be "broken in"!

More of the same here:

http://www.agoraquest.com/viewtopic.php?topic=27042&forum=65&keyword2=electronics

Quote:

"Probably one of the most unexpected results of this round of AC Plug and power cord upgrades that I did was a phenominal increase in bass response. This is not flabby bloated bass either its very tight and articulate with great low frequency extension and at least twice as much volume as I was getting before."

and

"Its been 2 weeks now since I did the AC Plug and Power Cord upgrades and they are breaking in nicely and sounding even better all of the time."

and

"From the moment I first heard these new AC Power Delivery upgrades in place I noticed something that I had never heard before and that was the sound of constant real time break in.

After doing a considerable amount of research I learned that the Furutech FI-11G AC Plugs take about 150 hours to come into their own and 400 hours to fully break in and I have come to the conclusion that the FI-11G AC Plugs breaking in were responsable for what I was hearing.

I've broken in new signal wires and speaker wires before and I'm familiar with the gradual improvement in sound quality that occurs over the break in period, but these AC Plugs are very different in that I could actually hear the sonics improving continuously moment to moment.

Over the last 100 hours of run in the Sound Quality² of the Audio system has become increasingly richer and more Musical posessing even more lusciously mellifulous transparent liquidity and instumental bloom than I ever imagined an Audio system would be able to convey.

I was totally shocked at the transformation that these AC Power Delivery Mods brought because I've listened to 100's of top notch well wired and expertly set up Audio systems that have equipment that cost up to 40X what my equipment cost and they didn't sound anything even remotely close to the way this Audio system sounds now!"

Chu Gai

05-14-08, 01:38 PM

Must be something peculiar to Sony.

jneutron

05-14-08, 01:40 PM

Never said that. The _only_ point I'm trying to make is that swapping out a power cord for doesn't contribute enough to the equation to make any meaningful difference to the problem.

And you base that on what? Your opinion?

You are not a power engineer, you are not a low level instrumentation engineer, you are not an audio engineer, you are not a test engineer.

So what basis are you using? Books or internet forums?

Tell you what, first explain how the difference between two power cords can make a difference with respect to any of these?

I thought you said you re-read my posts... Try again..

I've gone back and re-read your posts. At some points you want to look at only the power cords between the equipment and then in other places you want to look at the entire system all the way back to the power company.

Strawman alert. Stick to the discussion please.. I detailed the outside power pole feed to the duplex outlet so that you could have a base upon which to learn. Clearly, you missed that mark..

Then, I detailed the agressor/victim coupling that I am speaking of.

I recommend you get a copy of IEEE 1050-2004, as well as a copy of the National Electric Code, and learn about what I am talking about before you continue to waffle about..I'll give ya page numbers if ya need em..

I'd certainly agree with you that if you really want to know what's going on then it's the latter case that matters. But I'd also argue that in this case the power cords to your equipment simply isn't relevant compared to the problems in the rest of the system.

Silly. The duplex forms the safety ground point. It's fruitless to argue the miles and miles from the powerplant schtick.

It's what happens at the local level that impacts the equipment. And like it or not, the safety ground is here to stay...giving us loop problems..

So explain what I'm missing here? Are we looking at two pieces of equipment with common grounds or not?

My goodness, I had thought that a drawing would have helped you to understand..

The people who care about what the reviewers and vendors are saying are the people buying the equipment and reading theses forums. Its impossible to read your posts "accurately" simply because tey don't include enough information for anyone not doing your work on a day to day basis to tell what factors you are or are not taking into account on any given post.

Which part of the drawing is confusing to you? Neutral, hot, duplex outlet, IC...did I make the font too small for you to read?

It's got nothing to do with the issues you raise, but you've yet to explain how the issues you raise are relevant...

Stating it is not relevant certainly doesn't make it so.

Relevancy has been clearly provided, you can stick your head in the sand if you wish...or you could ask questions, as you certainly need to.

It's not the people's fault that they listen to the vendor "white paper garbage" and marketing blurbs..it's the fault of people like you. Ones who provide blanket assertions such as "it can't possibly affect", or "it can't be audible", without any experience whatsoever in the discipline which lives and breathes this stuff on a daily basis..

At least in the pro audio market, real issues of grounding loops has been addressed...even if issues of image localization have not.

Sheesh, it's bad enough dealing with silly guru's and vendors, nevermind technical people speaking well outside their area of expertise or experience.

Cheers, John

jneutron

05-14-08, 01:47 PM

It is indeed, but those deficiencies will not be adequately addressed by voodoo power cords, either.

Nobody has claimed it could be.

What I posted was a scientifically accurate analysis and explanation of how a power cord can affect the quality of the sound we hear. That was the question posed, and the question answered.

As far as I can see, the aftermarket powercord market is a mismash of haphazard, random.... garbage..lots of excrement tossed onto a wall in the hopes that some sticks..

Fancy dielectrics, fancy litz (dangerous I might add), fancy silver, fancy shields. a real hodgepodge...

When all that is needed is attention to details. Cord inductance, cord resistance, cord capacitance, conductor geometry, external field coupling..and system interconnect geometry. Not snake oil. Tried and true e/m stuff, albeit culled from many sources like AES, IEEE, and some vendors like rane.

Cheers, John

J_Palmer_Cass

05-14-08, 01:48 PM

Must be something peculiar to Sony.

Sony supplies a warning not to use their equipment when you are under the influence!

jneutron

05-14-08, 02:50 PM

Sony supplies a warning not to use their equipment when you are under the influence!

Good. (or, is that bad??) I'll turn my stereo on after a martini.. (I assume you meant "legal" influence).;)

I never woulda thought that my stereo could get twice the volume by changing the plug...what was I thinking??? Twice the volume, now does that mean an order of magnitude more power??

If I take 5 cords, plug them in series, then power an Ipod, will I then be able to run the "absolutely last tour of Cher" in madison square without all that pro stuff? I mean, 5 orders of magnitude increase...that's nuttin ta sneeze about..imagine gettin rid of the 5 or 10 truckloads of sound equipment..what a deal...

Talk about wigglin the electrons outta the amp...:p

Cheers, John

scientest

05-14-08, 03:23 PM

And you base that on what? Your opinion?

A couple quick back of the envelope calculations, your contributions and common sense. Can you show otherwise, and if so why have you come this far without doing so? As I've asked previously; what's the max difference two different power cords are going to contribute in this situation (if neither of them is defective). Surely you've got the knowledge and tools at hand to answer this, yet after a dozen or so posts you haven't touched it. Why not?

useless rehash

My goodness, I had thought that a drawing would have helped you to understand..

Which part of the drawing is confusing to you? Neutral, hot, duplex outlet, IC...did I make the font too small for you to read?

Well, since that post came after the one I responded to I'd expect that you could figure out I wasn't addressing that diagram at all in my post. I think I've done a pretty good job of quoting only a single post at a time and trying not to jump around? I will however go back to the post you point at here, it seems the most relevant thing in this entire discussion so far.

It's not the people's fault that they listen to the vendor "white paper garbage" and marketing blurbs..it's the fault of people like you. Ones who provide blanket assertions such as "it can't possibly affect", or "it can't be audible", without any experience whatsoever in the discipline which lives and breathes this stuff on a daily basis..

John, if you truly want to make a difference in this discussion then the onus is on you is to answer the questions directly instead of ducking them.

Sheesh, it's bad enough dealing with silly guru's and vendors, nevermind technical people speaking well outside their area of expertise or experience.

I'm sure the fact that most of this don't live eat and breath this stuff on a daily basis is frustrating to you. However, I've got more than enough background to understand anything you've tried to explain so far. So far you've more than adequately explained that power cords could have some small contribution to ground loop issues. What you haven't done is show how the difference between two normally functioning power cords could change this situation in a significant enough fashion to actually make a difference to the final audible signal!

jneutron

05-14-08, 04:13 PM

A couple quick back of the envelope calculations, your contributions and common sense.

What back of the envelope calculations? You posted something?

As I recall, I was discussing safety ground resistance and you talked about heating caused by power draw.

Do us a favor. Read IEEE STD 1050-2004 (or a later version), get a feel for what is being discussed.

As I've asked previously; what's the max difference two different power cords are going to contribute in this situation (if neither of them is defective). Surely you've got the knowledge and tools at hand to answer this, yet after a dozen or so posts you haven't touched it. Why not?

Saying something incorrectly many times over doesn't make it so. Go back and review the analysis and calculations I provided.

1..Resistance

2. Inductance

3. Geometry.

As I said, it does not serve you well to stick your head in the sand and pretend it wasn't there..

Well, since that post came after the one I responded to I'd expect that you could figure out I wasn't addressing that diagram at all in my post.

Fair enough.

I will however go back to the post you point at here, it seems the most relevant thing in this entire discussion so far.

It has been the entire discussion all along. You just didn't understand, and continued with the line sag argument, which was not the discussion..

John, if you truly want to make a difference in this discussion then the onus is on you is to answer the questions directly instead of ducking them.

Again, stating something repeatedly that is incorrect still doesn't make it correct..no matter how many times you say it..

I'm sure the fact that most of this don't live eat and breath this stuff on a daily basis is frustrating to you.

Actually, not at all. Part of my responsibilities is to provide tours to groups ranging from physicists and scientists, to professors, all the way down to 10 and 12 year old kids. So your lack of understanding and diversionary tactics frustrating??? Nah. Seen it before.

However, I've got more than enough background to understand anything you've tried to explain so far.

You haven't demonstrated that at all. I talk of ground loop induction, you provide an "experiment" which measures line sag under load...and talk of "heating", even though the discussion centers around the safety ground. It's not a discussion of power line quality, it's signal ground integrity and paths which affect it.

Again, I recommend you read IEEE STD 1050-2004. It speaks of what I discuss..

So far you've more than adequately explained that power cords could have some small contribution to ground loop issues.

Finally, the light is beginning to dawn..sheesh, I thought you were gonna hide in the sand forever**.:p

What you haven't done is show how the difference between two normally functioning power cords could change this situation in a significant enough fashion to actually make a difference to the final audible signal!

1..resistance
2..inductance
3. Geometry

You already have the diagram in front of you ...figure it out..

Cord 1...Home depot powercord, p/n 64455. #12awg, 400 mil hot to neutral centers. Parallel non twisted cord. 6 feet. 225 nH per foot. Resistance...#12, why ask..

Cord 2. Nordost yada yada powercord (don't know the name, don't know the parametrics, so I'll creat a semi-likeness that IS NOT TO BE CONSIDERED A SAFE LINECORD!!!!) (emphasis to anybody who would try this unsafe cord)

Make the cord with 8 twisted pairs (or a braid) of inadequate guage wires, say #22. Each pair has 160 nH inductance per foot, the lumped cord...20 nH per foot.

Cord 1...6 times 225, or 1350 nH, 1.35 uH. Cord 2, 20 times 6 or .12 uH.

300 wpc, rail voltage 75volts, xfmr ratio 2 to 1.

Full power draw close to a kilowatt, 10 amps ave..15 peak, at PF=1.

Cap input filters never have PF=1, haversine currents will be anywhere from twice to almost 5 times peak, depending on supply stiffness.

20 amps at third harmonic, 180 hz...only twice what I provided earlier, eh?

Hey, isn't this derivation looking familiar????? Like, a coupla posts ago on this very thread???wow..

dI/dt = 20 * 2 pi * 180 cos(2 pi 180t) = 22618cos().

22,618 amperes per second..

Lets apply this to the linecords...

L di/dt at 1.35 uH

22,618 *1.35 *10e-6...30534 * 10e-6

30 millivolts total line cord reaction. 15 millivolts ground bounce.

Now...cord #2 is an order of magnitude lower, that's 1.5 millivolts ground bounce..

Hmmmm...

So argue that I've over-estimated the haversine first harmonic amplitude...halve it...7.5 millivolts...insignificant?

Try using a real diode, 2 uSec spec'd reverse recovery time. 1 uSec from full conduction (say, 5 amps) to off (the assumption is soft recovery, even though lots of 2 uSec product recovers in 100 to 500 nSec).

5 amps in 1 uSec, that's 5,000,000 amps per second. 2 to 1 xfmr, thats 2.5 meg amps/sec..core reaction plus inductive loop, say it drops to 100 Kamps/sec.

Now inductance of the line cord..1.3 uH will tend to dull this silly speed whereas a 120 nH cord will pass more through. But 100 kamps/sec and 1.3 uH, that's 130 millivolts cordage, 65 ground bounce.

And I dropped the di/dt of a real diode by a factor of 25 to be conservative, as it is unknown how much of the switching transient can get through the transformer....

Haversine buzz due to diodes? been there, done that.

As I said, the basic parameters of the cordage is all that is required.

Not magic.

And next time, stop ignoring that which is already in front of you**...

I did this calculation already. And provided it.

Get with the program**..

Cheers, John

ps..left the red verbage** in, as I certainly did say it..I apologized two posts afterwards...

scientest

05-14-08, 04:21 PM

Here's the pic of two units being fed by one outlet.

Ok, this is basically my case 2) from a couple of posts back. Not common in consumer grade audio equipment but given that the audience here on AVS may be using equipment like separate preamps and amps that in fact have three wire power connections it's seems reasonable to address it.

circuit description

Note that I stuck an IC between the amp and the source unit. Note that I show the IR drop as occurring entirely within the resistance of the IC shield.

Note that I show the ground currents simply as flowing through the unit chassis'.

Ok, it appears you're assuming the IC ground is tied back to some common chassis star type ground point and an unbalanced connection. Seems fair enough, but if you're running this kind of equipment I'd sort of expect you to also want to run balanced interconnects?

By inspection, there are several methods that can be used to mitigate the problem.

1. Make the amp and source impervious to the chassis currents. (balanced is a good start)
2. Kill the loop by feeding the source power from the amp chassis daisy chain style. (this is a technique I used very successfully for a PA application where I had the source 125 feet from the amp, and still ran unbalanced line levels to the amp twisted around the power cord feeding the source....unfortunately, it is limited to applications which are fed from one breaker, 20 amps..
3. Break the ground (NOT RECOMMENDED for safety reasons).
4. Twist the powercord to force the integrated trapped field to zero(limited to integer twists, unless you put the amp on it's side).
5. Reduce the powercord external magnetic field.
5a. Coaxial cord (watch out for ampacity ratings)
5b. multiple conductors for hot and neutral
5b1....remember, each conductor must have the ampacity to clear the breaker in the event of a fault. The use of individually insulated wires, like a dozen #22's, will have the ampacity but if one of the wires shorts to ground or neutral, the one wire may not be sufficient to clear the breaker before a fire.

Which _really_ makes me wonder why you'd use this configuration without balanced interconnects but ok, lets proceed...

So, given all this, the things a power cord manufacturer can actually address are 4. and 5. I've seen some claims for such things, so lets adress them seperately if we can:

Say you somehow actually achieved number 4 perfectly. Can you tell us what the real end result would be for some reasonable lengths of power cords and interconnects? Do power line frequency variations make it impossible to achieve 4. perfectly or are they too small to mater?

What do you think one coudl reasonably expect from such a design that addressed your point 5? Again, can you tell us if the end result would actually be significant?

Note: It must also be noted, that if two wire cordage is being used, both the agressor and the victim loops still exist physically....the only change would be the exact paths for error signals to take... After all, how many people out there have been bitten by ground loop hum even though they had nothing but two wire cordage???

In particular, when you introduce cable TV sources into the equation; you've now got 100's of feet of R59 or R6 with some minimal attempt at grounding at source connecting up at multiple points to your equipment often with leaky power injectors added to the mess. However, this time I really, really can't see how swapping the power cords would help anything? (Actually, I don't think I've ever seen anyone selling "audiophile" 2 wire power cords, but I haven't looked very hard...)

ps...I specify ground loop hum, as I believe all humans have had that experience at one time or another..but in reality, we've all had to deal with dimmer noise, motor contactor noise, various transients, all getting into our audio equipment. This victim loop will indeed pick all that crap up and provide us very clear audible indications of the event. As is also clear by inspection, any noise caused by this loop, cannot be stopped by anything that is done to the power supply of the amplifier, as that is not the path that the error signals are taking to get into our music..

Don't you mean with the exception of floating the ground? Otherwise, sure, but once again, I've got to ask; is swapping the power cord going to help with any of this noise? At this point you don't even necessarily know the frequencies of the noise...

pps...scientest...you have mistaken 30 years of practical experience in the home audio and pa environment with 15 years of experience in a very hi-tech work environment. Luckily, what I do there is extremely valuable and entirely applicable to both consumer and professional audio technology. It's what I bring to the table, a good add on to the valuable experience of others who post here..

I'm sure I've got enough knowledge to follow anything you have time to throw out here. However, it's confusing trying to follow all of this across multiple posts with minimal contextual clues and few common references (such as your last post with diagrams finally included) so bear with me if I seem to miss some of your points. Suggesting a re-reading or going to external sources really isn't useful. If you really want us to benefit from your knowledge then the onus really is on you to make sure it's conveyed clearly...

dae3dae3

05-14-08, 04:30 PM

All I can say on this subject is that I have run my system both with and without the power cord and I can definitively say that "with" is better. The difference is quite stunning actually.

jneutron

05-14-08, 04:41 PM

Arggh...missed this by just a minute or two..sorry bout that...

Ok, this is basically my case 2) from a couple of posts back. Not common in consumer grade audio equipment but given that the audience here on AVS may be using equipment like separate preamps and amps that in fact have three wire power connections it's seems reasonable to address it.

It applies also to two prong cordage.

Ok, it appears you're assuming the IC ground is tied back to some common chassis star type ground point and an unbalanced connection. Seems fair enough, but if you're running this kind of equipment I'd sort of expect you to also want to run balanced interconnects?

Much of the equipment used by consumers is not balanced nor gives the option. As I said, balanced is the preferred. But it must be noted, that I'm confident that most consumer equipment doesn't test CMRR with pin 1 current..so even consumer balanced equipment may be inadequate to the task..

Which _really_ makes me wonder why you'd use this configuration without balanced interconnects but ok, lets proceed...

As I said, much of the equipment doesn't give the option.

Say you somehow actually achieved number 4 perfectly. Can you tell us what the real end result would be for some reasonable lengths of power cords and interconnects? Do power line frequency variations make it impossible to achieve 4. perfectly or are they too small to mater?

WHAT??? Powerline frequency variation??

What in the world are you talking about?? At no point in the discussion or analysis does the powerline frequency variation play any role...Ahhh, I think I see what you mean...this is not a wavelength issue. Twisting only alters the amount the magnetic field varies along the cord length..

Quite honestly I would:
1. Eliminate all line cord magnetic field. Triaxial design, outer shield being safety ground.
2. Run the source power from the amp. Triaxial also.
3. Twist the IC's around the source power cord, reduce loop as much as possible.
4. If still required, modify the amp and source to zero the coupling between ground currents and the active circuitry..star grounds are inadequate for this.
5. All internal wiring (high current stuff) primary and secondary, as well as the high current rails, low impedance wire (4 to 8 ohm maximum). (this can easily be done)..

What do you think one could reasonably expect from such a design that addressed your point 5? Again, can you tell us if the end result would actually be significant?

Look at the numbers in my last post, the one that I (admittedly) ended with too much attitude...I apologize for that ending..:o.

In particular, when you introduce cable TV sources into the equation; you've now got 100's of feet of R59 or R6 with some minimal attempt at grounding at source connecting up at multiple points to your equipment often with leaky power injectors added to the mess. However, this time I really, really can't see how swapping the power cords would help anything? (Actually, I don't think I've ever seen anyone selling "audiophile" 2 wire power cords, but I haven't looked very hard...)

Glad you mentioned that. The first drawing is useful in that aspect. And no, nothing from the duplex to the equipment can stop cable hum. That has to be considered at the source.

That was why I detailed the neutral aspect at the pole. In this part of the woods, it makes 60 hz gradients in the earth, and that does tend to wipe cable to audio interaction.

Don't you mean with the exception of floating the ground? Otherwise, sure, but once again, I've got to ask; is swapping the power cord going to help with any of this noise? At this point you don't even necessarily know the frequencies of the noise...

Noise (lets call it error signal) caused by the amp is haversine with known spectra albeit unknown fallof vs frequency, as well as higher audio band stuff making it's way through the filters.

Floating ground may or may not solve the problem. Input filtering tends to shunt the noise, and that may just change the reference point for the noise from 0 volts to riding 60 volts..

I'm sure I've got enough knowledge to follow anything you have time to throw out here. However, it's confusing trying to follow all of this across multiple posts with minimal contextual clues and few common references (such as your last post with diagrams finally included) so bear with me if I seem to miss some of your points.

And I apologize for being impatient..

Suggesting a re-reading or going to external sources really isn't useful. If you really want us to benefit from your knowledge then the onus really is on you to make sure it's conveyed clearly...

Point taken.

Cheers, John

Randybes

05-14-08, 04:45 PM

All I can say on this subject is that I have run my system both with and without the power cord and I can definitively say that "with" is better. The difference is quite stunning actually.Actually, I was listening to some Tiny Tim, pulled the power cord out of the system and it sounded much better. I am sure Neutron has a good scientific explanation. I heard he was very tight with Tiny when he was alive. Could be just a rumor though. Tiptoe through the tulips!

jneutron

05-14-08, 05:04 PM

Actually, I was listening to some Tiny Tim, pulled the power cord out of the system and it sounded much better. I am sure Neutron has a good scientific explanation. I heard he was very tight with Tiny when he was alive. Could be just a rumor though. Tiptoe through the tulips!

YOU AGAIN!!:p

Cheers, John

Randybes

05-14-08, 05:11 PM

YOU AGAIN!!:p

Cheers, JohnJust trying to keep you on your toes:p

Ron Party

05-14-08, 06:08 PM

Actually, I was listening to some Tiny Tim, pulled the power cord out of the system and it sounded much better.
Precious.

Chu Gai

05-15-08, 12:58 AM

I'm sure TT sounds much better on a resolving system

J_Palmer_Cass

05-15-08, 09:27 AM

You guys need to burn in your cables.

http://www.hagtech.com/frybaby.html

Reviewer Excerpts

"I was in a shock, that was my first reaction. This is one of the best tweaks I've ever tried ... fuller tones; more refined treble; bigger/deeper stage; instruments with a bigger body; more air; better sustain of instruments ... only after two or three seconds, it's so obvious that bass got more breath and freshness; bigger, deeper, wider and richer soundstage; more breath and body in bass; richer in the middle; warmer vocal; better sustain; you can feel and touch the singer who is here in my room ... the difference wa SO obvious and audible." - Marko Pecotic, Audiopuls

"The sound that came out of the speakers was in all aspect - BETTER. The color of sound didn't change, but colors were richer and clearer. The background of soundstage was completely dark, so it was easier to see the musicians and instruments. The character of instruments has become more natural. FryBaby is a very recommended component! Each and every serious audiophile should own it." - Robert Kocijan, Audiopuls

Chu Gai

05-15-08, 09:53 AM

Proof that you can be mentally and/or physically ill and still get a job. Or run for public office.

krabapple

05-15-08, 11:41 AM

No, he only got 4 out of 10. :D

Perhaps he wasn't wearing his magnetic bracelets that day.

It's a testament to unreality of of 'high end audio' that that doofus finds gainful employment in its media outlets. There was something amusing yet inevitable about seeing his byline appear in 'Stereophile'.

krabapple

05-15-08, 11:50 AM

I understand perfectly well what you're point is. I just don't think it's relevant. You've yet to explain how it is.

Anyone else feeling a bit 'deja vu'?
:rolleyes:

krab

05-15-08, 11:58 AM

Anyone else feeling a bit 'deja vu'?
:rolleyes:

Yep, all over again.

scientest

05-15-08, 01:46 PM

It has been the entire discussion all along. You just didn't understand, and continued with the line sag argument, which was not the discussion..

John, I never made a line sag argument at all, you missed my point completely. My point was simply that you probably couldn't find two power cord with significantly different ground resistances unless the "audiophile" version was designed in such a way as to start to become dangerous. In making that point I did assume that any three wire conductor would have a ground resistance similar to that of the other conductors (ore more importantly vice-versa) which given the wacky geometries you're talking about may not be true, but other than that, the point was actually posed as a question and not a statement of fact.

Which leads us to:

Again, stating something repeatedly that is incorrect still doesn't make it correct..no matter how many times you say it..

Which applies equally to both halves of this conversation. If you're not sure of what point I'm trying to make or the question I'm asking you might want to stop and ask for clarification instead of making insults. We get that you know what you're doing, that's why I'm asking the questions!

lots of stuff, including some wacky passive / aggressive behavior: sheesh

Cord 2. Nordost yada yada powercord (don't know the name, don't know the parametrics, so I'll creat a semi-likeness that IS NOT TO BE CONSIDERED A SAFE LINECORD!!!!) (emphasis to anybody who would try this unsafe cord)

Make the cord with 8 twisted pairs (or a braid) of inadequate guage wires, say #22. Each pair has 160 nH inductance per foot, the lumped cord...20 nH per foot.

Ok, gotta ask how you calculated the inductance of the twisted pairs? I wouldn't expect you can simply use the formula for a parallel pair of round conductors? At a minimum wouldn't you have to apply some correction for length loss dues to the twisting which leads to a need for conductors that are overall longer in length? Second, do you have to start looking at the twisted version as starting to approach a coil and account for that somehow? Third, do you need to account for the fact that individual pairs are not somehow perfectly separated from each other? It seems that the conductors of each pair could be in as close proximity to members of the other pairs as they are to the other member of the single pair. Not sure it makes a difference...

stuff that follows from these assumptions

max190

05-15-08, 02:07 PM

This is the OP <baddgsx> after reading the last 2 pages...

http://i38.photobucket.com/albums/e117/smax190/68386240.jpg

Randybes

05-15-08, 03:27 PM

This is the OP <baddgsx> after reading the last 2 pages...

http://i38.photobucket.com/albums/e117/smax190/68386240.jpgThat is great-I love it ROFL.

jneutron

05-16-08, 09:38 AM

John, I never made a line sag argument at all, you missed my point completely. My point was simply that you probably couldn't find two power cord with significantly different ground resistances unless the "audiophile" version was designed in such a way as to start to become dangerous.

Your proposed experiment had no source equipment detailed. Without the formation of the victim loop in any clearly defined way, a "difference" can't be measured..as even if a difference is found, lack of control over the victim loop makes the test of no use.

As the drawing I posted later in the thread showed, it is not the resistance per se that causes the victim loop, but the fact that the loop is there.

The cord and IC resistances will impact the loop current that results, and the selection of the PC and IC can certainly impact the loop voltages and currents, both in intensity and distribution of the drops.

which given the wacky geometries you're talking about

Actually, it's not wacky at all. HOWEVER, it could be dangerous. As I pointed out on another thread as well as this one, the use of multiple insulated individual conductors to achieve the desired ampacity while optimizing some other parameter (this case was inductance), carries with it the scenario of one very small conducter shorting to another but yet not having the ampacity to clear the circuit breaker feeding it. If a shorted small guage conductor cannot clear the breaker magnetically, all that is left is the thermal response of the breaker, and guages smaller than #18 can have serious consequences in the event of a fault.

Which applies equally to both halves of this conversation. If you're not sure of what point I'm trying to make or the question I'm asking you might want to stop and ask for clarification instead of making insults.

umm, which part of my apology was unclear to you?? I am not one to simply erase verbage stated as if it wasn't stated at all, but highlited it instead with astericks..

Ok, gotta ask how you calculated the inductance of the twisted pairs? I wouldn't expect you can simply use the formula for a parallel pair of round conductors? At a minimum wouldn't you have to apply some correction for length loss dues to the twisting which leads to a need for conductors that are overall longer in length? Second, do you have to start looking at the twisted version as starting to approach a coil and account for that somehow? Third, do you need to account for the fact that individual pairs are not somehow perfectly separated from each other? It seems that the conductors of each pair could be in as close proximity to members of the other pairs as they are to the other member of the single pair. Not sure it makes a difference...

ok...one at a time..

1. Twisting a conductor does not alter it's inductance appreciably, for typical and reasonable twist pitches (2 to 5% longer wires will of course increase the inductance 2 to 5% per unit length, but that's small potatoes). Measurement of the inductance vs twist pitch for a specific length of wire produced zero change in total inductance, which certainly will translate into more inductance per unit length but the length change was quite small once the wires were allowed to relax to final form. Inductance wise, the biggest impact twisting will have is forcing the conductors to remain together, spacing increases the inductance. Using the terman equation results in the proper numbers.

2. No, for twist pitches that are reasonable, you don't worry about solenoidal inductance..Also, note that for extreme pitches, you have made two solenoids, and they are counter-wound. This negates the solenoidal fields. (like a bifilar resistor)

3. A very good question. If the twist pitches are exactly the same, then there can be proximity effects. If you space the pairs two or three wire diameters apart, the proximity effect plummets (the external field falls as 1/r^2). A braid, if done correctly (I think Kimber used braid, but do not know if it was done correctly) eliminates the proximity effect entirely. But a braid with insufficiently sized wires used as hot and neutral is a very dangerous construct. If teflon insulation is used as well (for whatever audiophile reason), this is even moreso a danger.

Cat5e got around this proximity effect by using different twist pitches for the four pairs. At 100 ohms per pair, if you took a cat5e cable and connected all four stripes together and all four solids together, you make a cable with 25 ohms characteristic impedance. (I think it was cat 4 that used an internal "x" of plastic to try to space them further apart...but I'm not sure about that)

What I hadn't pointed out with the multipair line cord is the fact that the lower inductance of the cord can easily cause higher current slew rates for the error signals such as diode transients, allowing more hash to get into the romex on the other side of the duplex. It did significantly reduce the external fields of the cord, but at what cost?

Systems which are large enough to require more than one breaker will have more of an issue with increased current noise being injected into the duplex.

So, while the cord could easily drop the agressor/victim coupling I've been discussing, it is only altering "this side of the wall". Random haphazard solutions such as a simple line cord, are exactly that...random and haphazard.

That is why I had a larger list of items I would address. The task is multi-pronged..

1. Eliminate the source of the local fields. (using (triaxial lines) and low Z internals.)
2. Eliminate the pickup of local and non-local fields. ( wrap IC's around source line cord (to make the victim loop as small as possible), daisy the source cord.)
3. Eliminate the electronics sensitivity to ground currents (modification of star ground techniques.)

ps.. quite honestly, I'm suprised nobody has hollered when they read "modification to star ground techniques". Quite honestly, that is the most significant aspect of all the posts I've made on this thread.

Cheers, John

jneutron

05-16-08, 09:39 AM

This is the OP <baddgsx> after reading the last 2 pages...

http://i38.photobucket.com/albums/e117/smax190/68386240.jpg

Hey, let him speak for himself!!! (when he wakes up...:p..)

Cheers, John

ps..this is an audio theory, setup and chat forum, no??

jneutron

05-16-08, 10:04 AM

One thing I forgot to mention.

You had questioned a discussion of 3 prong grounded cords as pretty much a waste of time because yours and everybody elses equipment uses two prongs only (embellishment on my part, but certainly a logical argument on your part).

This thread is entirely about aftermarket power cords. I believe most (if not all) of the aftermarket cords available use an IEC connector, as opposed to having the end user solder a new one in.

So it is my belief that for any discussion concerning aftermarket cords, it is inherently a discussion of equipment which has built in safety grounds..three prong, with all the problems caused by them.

Cheers, John

penngray

05-16-08, 10:42 AM

Thanks for the fun read guys....Im sure the OP left on page one when he posted that he saved "$500" on power cables.

I actually remember now that I had 2 years of EEE courses many, many years ago. Thanks for the walk through memory lane and to re-confirm I picked the right field to go into...IT consulting...ZERO EEE :D

penngray

05-16-08, 10:43 AM

You had questioned a discussion of 3 prong grounded cords as pretty much a waste of time because yours and everybody elses equipment uses two prongs only (embellishment on my part, but certainly a logical argument on your part).

Sorry but I think lots of people use Pro-amps, I know anyone that has built a sub system has pro-amps and they have three prong plugs.

I do have a hum in my system, unbalanced to balance, separate circuit issue, mismatched voltage issue.

jneutron

05-16-08, 10:44 AM

scientest

05-16-08, 10:53 AM

This thread is entirely about aftermarket power cords. I believe most (if not all) of the aftermarket cords available use an IEC connector, as opposed to having the end user solder a new one in.

I did a quick Google at one point in this discussion but didn't have time to point to the results at the time. The third most popular link for "audiophile power cords" yields up this gem: DIY Power Cord (http://www.venhaus1.com/diymains.html) which includes a shielded 2 conductor power cable. Any comments on these, um, "designs" ?

scientest

05-16-08, 10:58 AM

Sorry but I think lots of people use Pro-amps, I know anyone that has built a sub system has pro-amps and they have three prong plugs.

Yeah, but the main point under discussion WRT the 3 wire circuits only arises if you have a source connected to the amp that _also_ has a three wire wire power cord.

scientest

05-16-08, 11:02 AM

ps.. quite honestly, I'm suprised nobody has hollered when they read "modification to star ground techniques". Quite honestly, that is the most significant aspect of all the posts I've made on this thread.

Ok, I'll bite; what would you propose?

jneutron

05-16-08, 11:50 AM

I did a quick Google at one point in this discussion but didn't have time to point to the results at the time. The third most popular link for "audiophile power cords" yields up this gem: DIY Power Cord (http://www.venhaus1.com/diymains.html) which includes a shielded 2 conductor power cable. Any comments on these, um, "designs" ?

First..your "um, designs" comment is rather appropriate, I'd say.:p

1. Teflon's a no-no.
2. Teflon inside a hose? What's the NFPA's stance on AC inside unacceptable hose that isn't approved conduit?? Scary. The ampacity of a wire pair is heavily dependent on the wire being able to remain cool in it's surroundings. Uncontrolled tubes stop that. While possibly ok, I'd certainly not recommend it. Even if the teflon is rated for ampacity at 200C, what's the hose rated for?

3. Their two wire solution didn't show an IEC connector. Presumeably, the diy'er has to open the equipment and connect it in? Whatever happened to the concept of "no user serviceable parts inside...please refer servicing to qualified personnell??"

What prevents hot/neutral swapping?

Ok, I'll bite; what would you propose?

I was gonna say "my avatar" shows it all, but as I noticed, it's not on this forum that I have that avatar..:(

A star configuration is great when there are no magnetic fields to speak of, which is consistent with vacuum tubes and high impedance. At low impedances, there is much more current available, which causes more magnetic fields.

So star grounding is not that good at handling currents that have a high rate of change.. If a wire that has high current slew is connected to the star, the other connections of the star will see the magnetic field that is generated by that offending wire.

Two things.

1. A (edit: Hollow) cylinder of conductor has ZERO internal magnetic field as a result of the current in the cylinder. A braid has no internal field, nada..(that is part of my avatar..).

2. Take advantage of that fact by altering the input shielded wire configuration..

Here's a jpeg..

Notes:
1. This does not fix an IR drop of the input shield. If there's that much current, your flying a kite in a hurricane. Fix the hurricane by eliminating the external victim loop.
2. It can be modified to work with differential inputs...I leave that to anybody interested.
3. For stereo inputs, it gets a tad more interesting, but is certainly doable...again, whoever wants to try, go for it.
4. Note I didn't show the outputs. Don't connect the speaker neg to the chassis, but use low Z cable instead..either coaxial low guage, ribbon construct, whatever gets you low external inductance...twisted #12 would be next best..
5. Prevent any low impedance wire from generating a magnetic field... While the green reference ground is within a shield, it is an electrostatic shield only for the green and red wire, not a magnetic one. It can only prevent induced voltage on the red and green wires from currents within the shield, everything outside the braid can still be affected. The red and green wires will still be susceptable to stray magnetic fields caused external to their shield, especially ones from the supply rails, line cords, speaker wires, or chassis currents. But no matter what the current is that flows through the input shield braid, it will not impose a voltage across the green wire.

Cheers, John

ps..oops, I just saw that the jpg conversion left text out..The noise current is the purple wires, I show it going in the system at the rca (or pin 1) ground, and leaving via the power cord safety ground. Sorry bout that.