Shifting.
Please read my post about L versus R channel differences and what happened when I reversed the BQP and reference resistor.
More than on way
If electrons were to leave the cathode at higher velocity I would expect an increase in brightness of the trace.
So, the two traces are not identical when looked at in magnification.
Sure the average levels are similar however the spectral distribution is not identical.
Channel response and noise differences are to be expected and need to be cancelled.
One method would be to compare two stereo plots - one stereo plot with BB in L channel and reference resistor in R channel, and second stereo plot with BB in R channel and reference resistor in L channel.
However this will not work accurately because of differing noise during the two time separated experiments.
See where I'm going ?....one source feeding two signal paths driving an accurate differential input of some sort.
Eric.
If the trace is scanning across the centre line of the screen I dont understand why there should be vertical deflection.
If electrons were to leave the cathode at higher velocity I would expect an increase in brightness of the trace.
Do you mean this statement ?.
If this is the passage you mean, it does naught to address the problem of channel differences.
So, the two traces are not identical when looked at in magnification.
Sure the average levels are similar however the spectral distribution is not identical.
Channel response and noise differences are to be expected and need to be cancelled.
One method would be to compare two stereo plots - one stereo plot with BB in L channel and reference resistor in R channel, and second stereo plot with BB in R channel and reference resistor in L channel.
However this will not work accurately because of differing noise during the two time separated experiments.
See where I'm going ?....one source feeding two signal paths driving an accurate differential input of some sort.
Eric.
Only if you don't understand how a CRT works.
As for the rest, I think you also need to read what I wrote. The traces are identical (within noise limits- this is at -145dB with no signal averaging!) except for a few discrete green blips (also at ridiculously low levels). The blips do not change whether or not the BQP or reference resistor is in that channel. Spectral distribution is identical- I have no idea what you're on about there.
If the trace is scannng across the screen centre line it is because zero or very little voltage is being applied to the vertical deflection plates, so no change in vertical deflection is to be expected.
Correct me and clarify please if I am not understanding what you are meaning.
The plots are not identical because of at least two reasons - channel differences, and differences between the BB and reference resistor.
I know the plots are way down at -145dB, however the test method does not actually accurately differentiate the difference between resistor and BB.
In other words the resistor and BB are not proven to be identical...ie the test is inconclusive.
I reckon a different testing method is required.
Eric.
Actually, it does. What it says is that there's no difference. If you have data to the contrary, please share it with us.
The two traces are not identical, so therefore there is some difference between the two channels as tested.
Is it channel differences or differences between BB and reference resistor that causes the non correlation between the traces, or is it all three (4) causes?.
There are too many variables for the test to be fully conclusive.
Stuart, I am not trying to discredit your work, only to find definitive measurements, and we don't have them quite yet.
Eric.
I copied, pasted and read that passage already.
That passage says that one channel is more susceptible to external noise than the other and nothing more.
What I am saying is that if the two channels are not identical then no testing conclusion can be drawn.
You have said that the two traces are the same - sure the average noise levels are near as dammit, but the fine detail of the two traces are not exactly the same.
So what is causing this fine difference...the channel differences or the the BB ?
Or is it both ?.
Nothing is actually proven here yet because of extraneous variables.
This is not intended as negative criticism, only peer review.
Putting heads together solves questions, bashing does not.
Eric.
Good grief. If a resistor and BQP were both checked on the right channel, looking identical, and a resistor and BQP were both checked on the left channel, looking identical, why is there confusion? It doesn't matter if the noise floor of the left channel is slightly different than the right.
missing the point Eric.
The claims for the Bybee (indeed, the very SLIGHTLY more plausible ones) include that it makes an audible difference.
No data on this has EVER been presented of course, but lets just lay that off to one side for a moment and consider a difference in measurement that is -145db in scale.
In the Australian vernacular, that's three fifths of 5 eighths of effaye. It is, to all intents and purposes, unhearable. And by all, I mean all.
Might there be a component of what you can see in the screen shots that arises from differences in the two components? Absolutely. And in all likelihood, you will get the same quantum of difference between two identically rated but differently manufactured resistors. But the absolute size of the difference is miniscule - infintisimal.
But SY's measurements point toward an inevitable conclusion - the bybee device is in practise an expensive resistor, nothing more.
What do you mean by identical - the average level or the fine detail ?.
A test improvement would be to use one output to drive both devices feeding separate inputs - then at least the source noise would be correlated.
Eric.
Unless the random noise were identical in the two measurements, Eric will claim that they're different. If they're the same, then they're not noise. If the random noise is reduced by signal averaging, then Curl maintains that the functioning of the device is obscured. Simple, isn't it?
- Status
- Not open for further replies.