rdf said:
That of course, only works if the difference within the cable is either R, L, or C.
It unfortunately, does not consider the possibility that the ground loop formed by the source, the amp, the line cords, and the IC, picks up via faraday's law of induction, amp line cord draw induced time varying magnetic fields. In which case, using a tape loop for in/out is not applicable for evaluating IC's, unless the final application is to connect the tapeloop out to the tapeloop in..
Division of ground loop current induced voltage drop will be dependent on the IC shield resistance as well as the line cord ground resistances.
The same applies to that silly abx speaker wire switching box. We're talking about 50 milliohms and below for R, 10 nH up to 2 uH for L, and capacitance from 100 pf give or take to a nF...and yet we are supposed to use a bowl of spaghetti switchbox?????
Bet the screws in it have permeability greater than 1..totally uncontrolled.
Cheers, John
jneutron said:
Exactly what I propose, save for the tape loop switching control circuit being replaced with one that generates random results and tabulates guesses. Regarding all the other potential effects of induction, RFI, etc., those would seem to me constants of all interconnect usage and valid elements for the test to capture. The intent is to keep peeling off variables until we hit the logical wall. The implementation of most ABX tests I've seen have been full of logical holes and unwarranted assumptions, though as always I appreciate counter examples.
rdf said:
No. As I read you, you are proposing using the tape loop as the switching mechanism. I point out that this configuration is not consistent with end use, having considerable differences. The tape loop has a single ground, the wire connecting the ground reference to just the same ground reference. Connecting two pieces of equipment is not the same, which inherently has large haversines included in the loop.
rdf said:
Yes, they are constants. But the IC's may not deal with them identically. Use of the tape loop does not include these entities. If no difference is detected using the tape loop method, all one has established is that using the IC in the tape loop produces no difference. Not whether using the IC in it's intended application makes a difference.
rdf said:
If you peel off the variables as such, all you establish is that the tape loop of the amp has no issue with the IC. Again, that is not the final target app.
Cheers, John
Yes and no. The tape loop's internal, unswitched ground reference is a very good objection I hadn't considered. Inconvenient to address, far from impossible to greatly reduce. The original suggestion was an integrated amp with an active buffered pre-amp section. If the pre and power amps were on individual PCB boards separated by passive wires and switching the front end could conceivably be powered by an independent supply and power cord. I recall the integrated version of the ASR Emitter, currently at the top of editor's choice lists, does something similar with a battery powered front end. Batteries are in fact another option for breaking internal grounds. The tape selector circuit can be adapted to switch ground as well as signal. Both these topologies exist in highly regarded audiophile equipment. Chassis handling would require some care but I could see butting two chassis side by side with a film or air layer to electrically isolate them. Other methods suggest themselves if this is objectionable. PCBs can be positioned to minimize the additional internal wiring required to connect the two devices.
I don't see how the above could be considered topologically so different from normal operation as obviate results. It is an electrically isolated pre driving an amp with an interconect switched in and out of circuit. Since the interconnect is outside the box and subject to all the environmental effects of normal use - including RFI, microphonics, bad vibes, etc. - I think breaking the power supplies and chassis cover most of those objections. (Not sure where haversines come into play.)
The game is to bring the test conditions within the realm of normal operating and build logical 'firewalls'. Someone could object for instance that the slightly longer internal tape loop wiring obscures the difference introduced by switching in the interconnect but to do so is to place its effect at a greater level than all interconnects. In other words it's the same as saying the effect of an extra inch or two of internal wire common in a very great many hi end products swamps that of all available high end interconnects of any reasonable length. This decimates the number of high end electronics with valid claim to exposing cable differences and places the impact of interconnects on the same level as a couple of inches of wire in a shileded box. I could live with that.
I don't see how the above could be considered topologically so different from normal operation as obviate results. It is an electrically isolated pre driving an amp with an interconect switched in and out of circuit. Since the interconnect is outside the box and subject to all the environmental effects of normal use - including RFI, microphonics, bad vibes, etc. - I think breaking the power supplies and chassis cover most of those objections. (Not sure where haversines come into play.)
The game is to bring the test conditions within the realm of normal operating and build logical 'firewalls'. Someone could object for instance that the slightly longer internal tape loop wiring obscures the difference introduced by switching in the interconnect but to do so is to place its effect at a greater level than all interconnects. In other words it's the same as saying the effect of an extra inch or two of internal wire common in a very great many hi end products swamps that of all available high end interconnects of any reasonable length. This decimates the number of high end electronics with valid claim to exposing cable differences and places the impact of interconnects on the same level as a couple of inches of wire in a shileded box. I could live with that.
rdf said:
Your last sentence says it all. I will explain.
Take an amp, a source, plug them both in. For 3 wire plugs, each carries a safety ground. For 2 wire, the neutrals carry back to the breaker panel, where they connect to the same terminal as the safety grounds. For 2 wire, there will also be some ground loop impedance.
Now, connect these two using an IC. You have created a physical loop out of the cords, the IC, and the two chassis. That loop will intercept some amount of time varying magnetic field, and as a result, it will create a voltage. Since the field is time varying, the voltage will track the rate of change of the field.
Now, where to get such a field? Romex is parallel wire construction, it creates a dipole field at 60 hz. Line cords are of various constructions, but very few go to the effort of eliminating external field, a braid over it does nothing at 60 hz.
60 hz...hum. How many people on this planet have experienced ground loop hum? Ground loop hum happens when the loop coupling is SO BAD that one cannot ignore it. How many people think that if they randomly play with their cords, or cut the ground pin of their cord, that the problem is over?? Pretty much everybody. Unfortunately, that is not the case...
Faraday's law of induction states that the voltage is proportional to the rate of change of the magnetic field. IOW, the higher the frequency, the more the voltage for the same field intensity.
Haversines are the high current pulses that are typical of rectifier into capacitor supplies for most amps available. The louder the music, the larger the haversines. Look at the frequency distribution of haversines..they are odd order harmonics, and can go very high in order. The 11th harmonic of a haversine will produce 11 times more signal on the IC input loop.
Haversines are a way of life. The worst part is they are present only when the music is loud..it is a confounding element that is not going to be reproduced via a tape loop. Your methodology is heavily slanted towards failure to find a difference.
To what ends must we go to, in order to prevent this skewing? Equipment placement without consideration of fields, movement of cords randomly. Variables to be constrained..
I haven't mentioned high slew rate output currents getting past the supply either.
To use the tape loop, you have to embed the IC itself within the two power cords of the equipment under test, into the romex inside the wall, through the chassis....
Isn't it easire to use the ic as intended??
rdf said:
I don't have any particular disagreement with this para..but I do have one comment in general...
Star grounds, a commonly used geometry, is a slipshod, horrible, ridiculous method to use for low impedance high gain circuitry. Anecdotal accounts of changing a wire or two or solder, whatever, within a low impedance amp chassis, is probably correct. But not because of metal, or alloy, or insulation, or solder, but because they didn't duplicate the wire draping within a high emf construct.
Cheers, John
Re: One thing for sure that makes a difference
Mike, how many kilometers of wire do you mean? 😉
Speaking of wires and bad amps. I've recently discovered bad-bad-bad distortions on high end of spector my Marshall MXL-770 mics show when loaded on some real cables and real professional console input, instead of being just connected to the measurement equipment through a short cable... 🙁
Now load some single ended vintage-tube amp on high caapacitance, and you will get the same rectification on highs... Somebody will really enjoy such coloration, but others would not...
Portlandmike said:
Mike, how many kilometers of wire do you mean? 😉
Speaking of wires and bad amps. I've recently discovered bad-bad-bad distortions on high end of spector my Marshall MXL-770 mics show when loaded on some real cables and real professional console input, instead of being just connected to the measurement equipment through a short cable... 🙁
Now load some single ended vintage-tube amp on high caapacitance, and you will get the same rectification on highs... Somebody will really enjoy such coloration, but others would not...
The full and proper understanding of the question -- begets the full and complete answer, as per usual.
I personally find that ABX testing is useless in the extreme and needs to be flushed as virulently as possible, in terms of its use in 'high end' audio. It has no place there, as the differences that audiophiles look for...are subtle, by their very nature.
Engineers who cannot see that, cannot devise a proper test. They are, in effect, unqualified for this task.
Engineers who CAN see that, generally -don't- embark on quests for ABX testing to prove/disprove their hypothesis or ideas. They understand the concepts & effects, and results well enought that don't need the idea or execution of an ABX test (in the classic 'ABX test' style).
The rest of this, is 70 pages of trying to gently show engineers with (potentially - 'potentially'..as this is a learning curve thing, not an absolute) tin ears and a lack of understanding of the depth, shape and character of the actual issues at hand... that ABX testing is crap and can't even prove correct the time of day.
Suffice it to say, that scientific modelling or testing, has it place and use, but not here. Not in the case and way that ABX is classically used. It is an invalid test in this case. Not applicable, as the subtulties are swamped by the testing proceedure. Dumbass and others seems to have covered this in some ways, via their knowledge of the actual issues at hand.
Do I think that the the entire thread is useless? YES, and NO. One, I know that ABX testing for high end audio is crap and won't usually chime in with my 'opinion' on such threads. Two, it does help educate the folks who have no idea of the essence of high end audio, and the subtulties involved-it does help them begin to conceive the issues.
Truth be known, in my opinion, all aspects of high end audio subtulties can and eventually WILL be fully quantified via science, sound engineering, good logic, deduction, and testing proceedures. Here we see a bit of that happeing, mostly on the individual level. (methinks, from my casual assesment of this thread)
But in totality of definition? But not yet. The subulties involved are in the realm of the highest levels of analysis and highest levels of sensitivity of measurment, and in some ways...in methods and ways we have yet to devise and implement.
A while back (about a year ago, maybe 4 months, I'm bad with 'time'), I proposed that I could give you a simple way to test all subtulties of cables and the like, and fully document and understand the behaviour of cables of 'audio' use, but in esence, I'd be giving away the testing proceedure for designing perfect cables, and I'm not willing to do that. (on pondering..the testing proceedure etc has use in multiple industries, and the test itself, as a concept, reveals far, far more..see the very first line in this post) Only a fool would give such away. It would also impact on existing companies and and the abilites of the owners to put food on the table for themselves and their employees. And I desire no such responsibility, for such an act of disclosure, on what is really such a minor aspect of human existence.
I personally find that ABX testing is useless in the extreme and needs to be flushed as virulently as possible, in terms of its use in 'high end' audio. It has no place there, as the differences that audiophiles look for...are subtle, by their very nature.
Engineers who cannot see that, cannot devise a proper test. They are, in effect, unqualified for this task.
Engineers who CAN see that, generally -don't- embark on quests for ABX testing to prove/disprove their hypothesis or ideas. They understand the concepts & effects, and results well enought that don't need the idea or execution of an ABX test (in the classic 'ABX test' style).
The rest of this, is 70 pages of trying to gently show engineers with (potentially - 'potentially'..as this is a learning curve thing, not an absolute) tin ears and a lack of understanding of the depth, shape and character of the actual issues at hand... that ABX testing is crap and can't even prove correct the time of day.
Suffice it to say, that scientific modelling or testing, has it place and use, but not here. Not in the case and way that ABX is classically used. It is an invalid test in this case. Not applicable, as the subtulties are swamped by the testing proceedure. Dumbass and others seems to have covered this in some ways, via their knowledge of the actual issues at hand.
Do I think that the the entire thread is useless? YES, and NO. One, I know that ABX testing for high end audio is crap and won't usually chime in with my 'opinion' on such threads. Two, it does help educate the folks who have no idea of the essence of high end audio, and the subtulties involved-it does help them begin to conceive the issues.
Truth be known, in my opinion, all aspects of high end audio subtulties can and eventually WILL be fully quantified via science, sound engineering, good logic, deduction, and testing proceedures. Here we see a bit of that happeing, mostly on the individual level. (methinks, from my casual assesment of this thread)
But in totality of definition? But not yet. The subulties involved are in the realm of the highest levels of analysis and highest levels of sensitivity of measurment, and in some ways...in methods and ways we have yet to devise and implement.
A while back (about a year ago, maybe 4 months, I'm bad with 'time'), I proposed that I could give you a simple way to test all subtulties of cables and the like, and fully document and understand the behaviour of cables of 'audio' use, but in esence, I'd be giving away the testing proceedure for designing perfect cables, and I'm not willing to do that. (on pondering..the testing proceedure etc has use in multiple industries, and the test itself, as a concept, reveals far, far more..see the very first line in this post) Only a fool would give such away. It would also impact on existing companies and and the abilites of the owners to put food on the table for themselves and their employees. And I desire no such responsibility, for such an act of disclosure, on what is really such a minor aspect of human existence.
To be clear, are you saying the adjacent yet electrically isolated coupling of two independent chassis will swamp the differences between cables? It's easy to address. Place the pre on top of the amp, a shielded tube electrically connected to the chassis of the pre extends through a hole in the chassis of the amp without making contact. This is much closer to normal system operation but does leave the objection the hole in the amp's case could admit RFI swamping the effect of the cable. Let's be cautious here though. If a tiny ring gap in an amp chassis obscures cable differences, more huge swaths of gear are removed from contention as high end equipment. Open RCA connectors, front panel displays, non-metallic enclosures, anything with these features must be removed. If placing a pre on top of a power amp does the same, yet more systems are removed. The circle closes ever tighter on valid reports of cable differences. If chassis placement swamps them, then they're exceedingly rare indeed.
No, not and retain a suitable null reference. That's the goal, to switch between a null reference or placebo and the DUT.
Why? That's not how we use interconnects between amp and pre and I think I've come close to replicating that topology with the additional feature a shielded path for a short, acceptable null reference.
BTW, I'm a ground plane guy. 😉
My brother was a prime in Nortel's audio department for many years before moving to optical. My recollection is engineers didn't design auditory test procedures or protocols in telephony, they implemented them. It's a huge difference. He's walked out in disgust on AES 'cable proofs' (for the record, he uses the regular RatShack stuff) for their lack of rigour and professionalism.
Among other assorted, boilerplate subjectivist nonsense:
When you make your first million on your perfect cables, come on back here and have pity on our poor, deluded souls.
Good one.KBK said:
When you make your first million on your perfect cables, come on back here and have pity on our poor, deluded souls.
rdf said:
For high gain low impedance, a common reference point is the way to go. The problem is how to get the ground reference wires there without causing loops w/r to magnetic fields. That can be daunting.
The rest of your post, you seem to be thinking in terms of RFI. I have not been talking about that at all. Yes, an issue of course, but I'm talking about stuff in the audible range, EMI below 20K.
Cheers, John
Dumbass said:
You seem to have taken it as an expression of ego. Point being, you looked at the thing through the window of your emotions, not what was written.
You missed the point - it was clearly expressed. 🙂 Thus endth my contributions to the thread. It only spirals downward from here on in. 🙂
Speaking of tests, once I tested fellow audiofiles.
There were tube single ended, tube push-pull, and transistor amplifiers on stage. I did a trick, sometimes switching on limiters, 2 or 3 diodes on input of a transistor amp (in parallel to the input, in opposite polarities).
When I collected papers, I discovered that some folks heard well differences, but most of them called "tube push-pull" a transistor amp with symmetrical diode limiter, and "tube single ended" was the same transistor amp with asymmetrical limiter (3 diodes). My conclusion was, not all audiofiles like low distortions, some of them prefer some type of them.
As the result, I put a big bold cross on tube output stages and concentrated on musical distortions of amps when overloaded.
There were tube single ended, tube push-pull, and transistor amplifiers on stage. I did a trick, sometimes switching on limiters, 2 or 3 diodes on input of a transistor amp (in parallel to the input, in opposite polarities).
When I collected papers, I discovered that some folks heard well differences, but most of them called "tube push-pull" a transistor amp with symmetrical diode limiter, and "tube single ended" was the same transistor amp with asymmetrical limiter (3 diodes). My conclusion was, not all audiofiles like low distortions, some of them prefer some type of them.
As the result, I put a big bold cross on tube output stages and concentrated on musical distortions of amps when overloaded.
I wonder why not just listen to a pair of speaker cables till u r sure of what you heard,switch the system off,change to another speaker cable,switch on the system and repeat the listening test.Choose a real system and go for it.This is how we hear the music,this is how a system works.Any test proposals that differ from this simple test to my opinion comes from the heads mostly of people knowing much about numbers and little if anything about music.Not to mention their refusal to spend more time on real comparative tests than calculations.Training your ears by listening to more and more components will eventually make you a better listener while making more and more calculations will eventually make you worse.And a better listener enjoys music.
Panicos K said:
What does this test? Auditory memory is very short, and very poor. You know which cable you are listening to, so that will have a large effect on what you hear.
In short this tests nothing, and can only generate a purely subjective result.
Currently we have nothing BUT subjective results.
very well put! All cables have different electrical properties so how can they all sound the same? Do all cd players or cartridges sound the same.........NO.
jneutron said:
I think we have a point to agree on here. You say not all are silly. I say I have no way to tell wich ones are silly, and which are not. Lacking evidence that there are non-silly claims, I make the determination that they are all silly.
How do you know that not all are silly?
audiophile36 said:
They may have different electrical properties, but that doesn't mean that they are relevant to the ability of humans to hear those differences.
Maybe all CD players and cartridges don't sound the same, but a great many do as well. There are people who claim to hear differences with that equipment too, but it doesn't separate whether the the audible differences are real, or imagined.
jneutron said:
assuming you can at least, hear something,
what else do you have in your stereo/sound system?
````````````````````````````````````````````````
somethin else-
http://www.youtube.com/watch?v=Xf-MF_SROOc&NR
macgyver10,an individual's system,taste,and understanding IS subjective.You mean you dont trust yourself to make an objective comment on how two things compare?It doesnt have to be your component against mine,it could be two unnown components between which you want to choose one.You cant be biased in such case.As audiophile says if it sounds good it is good,but surely one,you will find better to your ears,knowledge,and understanding of music.
- Status
- Not open for further replies.