Ah, but what is the copper jacket thickness and conductivity?
Is the copper polished or is it solder coated?
And what about the transition from the lead to the endcap...
What's the endcap material? Is it kovar so it matches the BeO TCE? Or is it Molybdenum or tungsten?
Was the endcap plated?? A little nickel flash perhaps followed by copper?
I tell ya, it's gotta be that near-superconductivity stuff. Dat would splain evertin...
Cheers, John
Don't know what this is supposed to mean. Powerflux asked the right question superposition holds in this case and separating signal from noise would have to violate several first principles.
Scott,
I am not sure what you are asking? I replied about a general case not the claims being made.
In a typical signal source and signal load there is in theory a noise resistive noise source at both ends. As you know a 10,000 ohm resistor makes more noise for a given temperature than a 100 ohm unit (Same....). If you connect the two the noise drops to the value you would expect for a 99 ohm resistor. Why?
ES
But it's the same with the signal. If you have a source sending a signal into a 10k resistor and you connect a 100 ohms at the receiving end, you have the signal level you would expect for that source feeding into a 99 ohms resistor. Or do I misread you?
jan didden
In order for the passive device to have a direct effect on thermal noise there must be missing terms in the Johnson-Nyquist equation for thermal noise. Are there memristors in it or something? At this point I'm pretty sure that the device can't change thermal noise, which is a problem because even downstream where the load impedance may be below 8 ohms there is still likely to be more thermal noise than anything from quantum effects that many scientists and engineers can't agree even exists (mainly because they've never had a problem with it, never saw it, even in circuits that must have extrememly low noise in general.) Still, it seems that in order to significantly act on the passed signal it would have to be mainly done through the magentic field over the .025 ohm resistor, because any capacitively coupled effect will have to force large currents into it in order to modify the signal, and especially in signal level circuits, an appropriate amount of current wont even exist. I'm beginnning to wonder whether, if this device has any military significance at all, that it is only to investigate the civilian and particularly civilian layman capability to do any meaningful research or deal with extremely low signal to noise and low level signals. Might be important if you are good at it and need to be sure how good anyone else is.
Last edited:
The directional vs non-directional doesn't make sense i.e. the "signal" could be hiss at the same amplitude.
In your example, by superposition this is the only answer. More interesting, what is the current noise in the loop they form? Again, no you can't tell which resistor the noise is from. In either case there is only one noise.
Yes that is right. Now model a perfect source resistor in parallel with a current noise source and perhaps even a signal source. This travels via perfect conductors to the load resistor which is also in parallel with it's own current noise source. The noise voltage measured across the perfect conductors drops when the source is connected and goes up when it is not. This model indicates that noise power is traveling in the reverse direction from the load resistor to the source resistor. Of course there is also signal and noise power going in the forward direction.
Devices that can tell the difference between forward and reverse power are common in RF work.
What happens when you induce voltage into the output of an amplifier is an interesting issue. Of course the amplifier damps much of it. But it can be an issue. Are the noise levels from an 8 ohm loudspeaker going to be enough to influence any decently designed audio power amplifier? I strongly suspect not.
However there are a fair number of people who do indicate they hear a difference. This is quit believable.
The Bybee website mentions using them in series with both legs of an AC line. Around here when I measure my power line I see less than 100 milliohms of source resistance. Would adding 50 more make a difference? Probably when you understand most amplifiers with capacitor filters draw current from the line at very small conduction angles. If the conduction angle were 15 degrees that could reduce the peak power of a 250 watt class AB stereo amplifier by .28 db or so. Most amplifiers probably less and under some conditions more.
An interesting aside is that when I was making various power line cleaners, if I made the line impedance too low some amplifiers would blow rectifier diodes. So I suspect under some conditions the increase in power line impedance may be able to be perceived.
There are also those who claim an improvement in audio interconnect cables. I have mentioned before that just unplugging and reconnecting a connector makes a difference that can be measured. So there are certainly those who could have inserted theses gizmos and actually gotten an improvement.
Of course any really refined measurement of any passive component will show that resistance never perfectly follows ohms law, capacitance has lots of non linear effects and inductance really can change if you wave a stick at it! So any really finicky measurements will show something.
So my conclusion is that people do hear things, the question is can Sy isolate what that is?
I'm still mainly interested in finding a way to set a hard trap for this thing doing something another Pacific .025 ohm resistor can't. Various ways of having drive connections affect the measurement should pretty much be removed rather than try to figure out all the ways it can be connected so the connection has more of an effect. If all you need is a different connector or some impedance to fix your stereo, you don't need a "Bybee".
Maybe in a few weeks when I get some dough I'll buy a couple. That would take some form of guts on my part. I'd like to see what it does to nanosecond edges and big pulses.
Maybe in a few weeks when I get some dough I'll buy a couple. That would take some form of guts on my part. I'd like to see what it does to nanosecond edges and big pulses.
Last edited:
Hi SY,
I have a spare, but we live too far apart. Next ? I think I have service information on it, as I'm pretty sure you must as well. PM me if I can lend a hand.
It would be so much easier if you couldn't just swing by!
John, you sly fox!!
-Chris
That sucks.
I have a spare, but we live too far apart. Next ? I think I have service information on it, as I'm pretty sure you must as well. PM me if I can lend a hand.
It would be so much easier if you couldn't just swing by!
That means the signals are processed Ferrodynamically. The device is a slight of hand trick to get you looking in the wrong place!
John, you sly fox!!
-Chris
All you need is two devices for test and two identical value resistors. Place them in a classic Wheatstone bridge setup. You will probably need a current limiting resistor on the drive input to protect your signal source amplifier and some very tiny value tweak to balance the bridge. Most likely a trim potentiometer in parallel with one side of the bridge. Now you can sweep the bridge and see twice the difference between your devices under test and the reference resistors.
Done with good shielding, a low noise source and lots of DSP samples and averaging will most likely show some differences. But that does not answer the question, "Are they a useful addition to a system?"
Oh and I forgot on my list of changes that people may hear from installing Bybee's, is there are some amplifiers that have been sold that have an inadequate, under certain conditions, output inductor. Adding .025 or .05 ohms could actually swing them to more stable.
Sy,
Your beer advice was good how about the snack food?
ES
Last edited:
having followed this thread from the start, I could accept that there are things beyond my understanding but ultimately-
I cant accept that once you have distorted a signal, you could separate the distortion from the original sound without using the undistorted 'bluprint' for reference.
Its like having a glass of 100% pure h2-0 and adding a thimble of pure h2-0 that has been taken from the other side of the world and expecting it to be possible to separate the 2 back into there orginal sources by shaking it.
Without tracking the molicules,or intelegent selection, theres no way to distinguish the 2- making the task scientifically impossible- end of.
Surely the only concevable way a passive non intelegent device could filter anything like this is for the distortion to be distiguishable from the master copy without needing to reference the master.
Is there even any test of any kind known that can indicate how much a signal of any sort has been destorted without needing a reference point??
This is a kind of backwards approach I grant you but hopefully you see what I mean.
I cant accept that once you have distorted a signal, you could separate the distortion from the original sound without using the undistorted 'bluprint' for reference.
Its like having a glass of 100% pure h2-0 and adding a thimble of pure h2-0 that has been taken from the other side of the world and expecting it to be possible to separate the 2 back into there orginal sources by shaking it.
Without tracking the molicules,or intelegent selection, theres no way to distinguish the 2- making the task scientifically impossible- end of.
Surely the only concevable way a passive non intelegent device could filter anything like this is for the distortion to be distiguishable from the master copy without needing to reference the master.
Is there even any test of any kind known that can indicate how much a signal of any sort has been destorted without needing a reference point??
This is a kind of backwards approach I grant you but hopefully you see what I mean.
Simon-
Yes, the bridge setup is handy for checking response that varies against a known load. I didn't plan on being very particular about the load resistor against the DUT but figured it pretty practical to start with a one the same as what's used in the Bybee itself. Remove maybe some of a variable. Far as impedance matching to any acceptable performance amplifier that were not purpose built, that is one reason for suggesting the transformer. The other is it is pretty easy to get a very high degree of balance to do a full bridge measurement, but the problem with that, even though it might handle some problems with parasitic load unbalance that might be otherwise severe at very at high frequencies, then you need a differential signal amp with very good specs to take advantage of it, although there could be quite low common mode, and if you did have any, you could check for fixture/transformer problems that might be harder to track without balance. Possibly you would even need balance to get a low enough noise floor to begin with. I don't really know what I'm doing with these kinds of measurement, but I had been envisioning some heavy copper plate to construct a full shield right off the bat. How you'd handle shielding and construction of the transformer would be very critical. The bonus is it doesn't have to be very high power, although large anyway if to pass low frequency. All the well known problems of wideband transformers apply.
Presumeably, assuming advertized action, the thermal noise on the test resistor should measure lower too, but you need something that can see how many picovolts above it's own noise to actually measure that. You could get 100,000 Bybees and just measure the noise coming off of that, if you could miraculously shrink them down so you didn't mostly receive your closest power line and radio station anyway.
Yes, the bridge setup is handy for checking response that varies against a known load. I didn't plan on being very particular about the load resistor against the DUT but figured it pretty practical to start with a one the same as what's used in the Bybee itself. Remove maybe some of a variable. Far as impedance matching to any acceptable performance amplifier that were not purpose built, that is one reason for suggesting the transformer. The other is it is pretty easy to get a very high degree of balance to do a full bridge measurement, but the problem with that, even though it might handle some problems with parasitic load unbalance that might be otherwise severe at very at high frequencies, then you need a differential signal amp with very good specs to take advantage of it, although there could be quite low common mode, and if you did have any, you could check for fixture/transformer problems that might be harder to track without balance. Possibly you would even need balance to get a low enough noise floor to begin with. I don't really know what I'm doing with these kinds of measurement, but I had been envisioning some heavy copper plate to construct a full shield right off the bat. How you'd handle shielding and construction of the transformer would be very critical. The bonus is it doesn't have to be very high power, although large anyway if to pass low frequency. All the well known problems of wideband transformers apply.
Presumeably, assuming advertized action, the thermal noise on the test resistor should measure lower too, but you need something that can see how many picovolts above it's own noise to actually measure that. You could get 100,000 Bybees and just measure the noise coming off of that, if you could miraculously shrink them down so you didn't mostly receive your closest power line and radio station anyway.
Last edited:
I should clarify, that my take on the ad sheets and hearsay is that the device is not supposed to reduce thermal noise, just some noise due to "quantum level effects", or quantum noise, if you will. And maybe there are some other signal distortions, calling noise a distortion, caused by these quantum effects, although I didn't get that there was supposedly such an effect. My take on the matter which led to the test fixture angle was that if Any noise is reduced enough to matter it should be measureable. Now, if it merely changes the signal but not the average level of noise, then you gotta actually look at the signal more carefully to pick apart what's happening. Either way you probably need very high resolution and probably low noise, especially at low test voltages, which is what you'll get if you're picking off a signal from a .025 resistor below a few watts long term. Possibly high power burst tests could be used. Where you'll find something interesting I really have no idea. I don't even know if the device can be destroyed by pulse power below the limitaions of the shunt resistor.
Last edited:
For added "come on now" effect, it's possible that it's not supposed (many Supposeds here) to do anything if the device is placed in an arrangement where it has no quantum effects to modify. The balanced transformer secondary may have a turn or few on it, the resistance of it should be below that of the device. So at that point you may be mainly modifying performance of ustream circuitry. Still, the idea is a difference should be measurable across the bridge if some action is only taking place in one side of it.
I think we need a proper meaurement instument to capture the purifying effect:
Scanning tunneling microscope - Wikipedia, the free encyclopedia
Another group collection? I can donate 10€ for the purchase of the instrument.
We need to see what is the state of the sound atoms and if their quantum has been purified or not. However if the 0.01 nm resolution is not enough the effect may not show up in a blind A/B test.
- Elias
Scanning tunneling microscope - Wikipedia, the free encyclopedia
Another group collection? I can donate 10€ for the purchase of the instrument.
We need to see what is the state of the sound atoms and if their quantum has been purified or not. However if the 0.01 nm resolution is not enough the effect may not show up in a blind A/B test.
- Elias
The input stage has gone a bit noisy- on the 0.1uV scale and a shorted input, the needle nearly pins. Chances are, it's the tantalum cap, but I also got a new input FET. The issue is the time to get into it and do the repairs. As a stopgap, I ordered some National LME49990, which should arrive today- configured for 20-26dB of gain, I should be able to get it above the new noise floor of the 3581A or even be able to measure the noise with my digital voltmeter, scope, and soundcard.
That'll be quite the collection plate. I found a book on STM's for a mere 170 Euros.
- Status
- Not open for further replies.