Characteristic Z of cable --- zip or coax .....
If you measure Z of a cable at low audio freqs, the Z is near infinitely high. As you increase freq, the Z will drop until at some freq, the Z stays constant.
As I recall, the freq is at around 100KHz where above that freq, the Z doesnt change. That constant Z is the characteristic Z of the cable.
THx-RNMarsh
If you measure Z of a cable at low audio freqs, the Z is near infinitely high. As you increase freq, the Z will drop until at some freq, the Z stays constant.
As I recall, the freq is at around 100KHz where above that freq, the Z doesnt change. That constant Z is the characteristic Z of the cable.
THx-RNMarsh
What makes it possible to rely on ears without getting into ABX is because the change of 'parameter' usually follows a 'pattern', such as a curve pattern with zero gradient in the middle. Often this change can be controlled with one or two variable that we can '.step'. For example, we can '.step' parameter x from 10 to 40 with 10 increment and the expected result is the change of harmonics domination. When x=10 H2 is the highest and when x=40 H2 is the lowest (but higher order harmonics are highest). If we listen to the 4 prototypes gradually from x=10 to x=40 we will be able to understand how to relate the perceived sound to the numbers. And often our ears can tell as the position of the zero gradient in a gradual change that follow a curve even if we are using non accurate simulation model.
But there is a trap. When we change one variable, we will usually change more than just one parameters in the circuit. As a result, we may draw a wrong conclusion when we think that the change in sound we perceived is caused by the change from H2 domination to H3 domination. IME, this H2 vs H3 is not critical and not a determining factor for better sound.
You have stated your opinion. Nobody has denied you that right. We have a right to tell you that you are mistaken: audio cables can be treated as short transmission lines and they do have frequency-dependent characteristic impedance. The issue being discussed here is whether at mid to low frequencies these Tlines support wave propagation.dadod said:
When I buy RG-58 I would usually be buying it for an RF application, where the characteristic impedance is approximately constant and resistive. You know this so why ask the question?jneutron said:
Your statement is untrue, but fortunately not very important. LS cable does have characteristic impedance. You would have to tell me exactly what you mean by "behaves as transmission line" because some people might take this to mean 'supports wave propagation' (not true at lower frequencies) while others will think it means 'does what a careful analysis of the TL equations tells us it should do' (true at all frequencies).dadod said:
Yes.Chris Hornbeck said:
It is difficult for dim people like me to think of a better definition.johnego said:
No. Linear impedances are sufficient. Failing that, sufficiently low source impedance and sufficiently high load impedance. Assuming voltage signals, of course; you want the opposite way round for current signals.
Not "are"; 'is' because damping factor and output impedance are the same thing, just written down in different ways. The output of an amplifier can be modelled as a non-ideal voltage source: the voltage should be a bigger version of the input voltage, and the output impedance should be whatever the speaker designer wants.
No. That constant Z is the RF characteristic impedance of the cable. At lower frequencies you measure a different value, typically higher, reactive and frequency-dependent. This too is the characteristic impedance at that frequency.RNMarsh said:
"image" doesn't necessarily mean sound stage - if a bell is recorded in mono, replay of that bell does not need two speakers to recreate a realistic, natural internal version of the bell
Yes, the definition of 'similarity' between input & output is very reasonable on the face of it - now let's define what we measure to compare input to output & what level of 'similarity' is acceptable & what isn't.
DF96;5808133 said:
I wanted to make it simple for some DIYers with no theoretical background.
For short cables at audio frequencies there no wave propagation and no reflected waves, so we can use just RLC of the cable to calculate the cable behavior, characteristic impedance does not matter.
Is this good enough for you?
It is not the shortness of a cable which prevents wave propagation. It is the cable parameters. If R dominates over L then you don't get wave propagation.
Two conditions were ignored.
1. Your rise time is long or longer than the entity to be measured.
2. Voltage? If you recall, current is the driver of the speaker.
What is the cable L and C?
Oh, and I have made no mention of oscillation in CB's amp. Just that his dual directional coupler tests appear to be quite reasonable. As others have questioned his test apparatus and in essence claim that he lied about the directionality of his coupler, I have asked about specs on the xfmrs to see about repeating with what others consider accurate.
To prove the premise is inaccurate, it is important to duplicate the test setup, both mine and CB's.
Changing anything (rise time for example), may not show what is expected.
I followed this as well with my cable image test. Set the specific conditions, follow them correctly and accurately, do the correct tests. If no image changes occur during my test, there is no difference as a result of the cable so get over it and move on.
It is about testable and repeatability. A claim must be both, which is what I am doing here. And what I would like to do with the coupler.. Build the correct xfmrs, then send them on to someone who can complete the construction and use the unit. Same as I have offered with my low b-dot cvr.
Jn
Last edited:
ABX will tell. Then you see that huge signal dissimilarity is inaudible and we are debating nuts like 100ns time differences here.
I don't like the gauge, otherwise it' no worse than zip. If one does what I did, I would be afraid of the capacitance.
Jn
I hope you got my points though. I get your points, you get my points, if not clear ask further, otherwise case closed 😀
Because it looks like the supra is also steady at 50 ohms from 10khz and up.
For a test that lasts 10 uSec or so, there is no 10khz content.
Jn
I hope you are not including me in "others". I have never suggested any dishonesty by him. Technical errors, yes, presumably caused by insufficient knowledge.jneutron said:
At 10kHz I make it magnitude 50.6R at angle -36.2 deg. That is 40.8 -j29.9. Sufficiently far from 50+j0 to mean that you cannot measure reflections with a 50R bridge. 10kHz is too low a frequency to get that cable into the 'RF' approximation; maybe 50kHz would be enough? He simply gives a figure of 49.6R; I don't know where he got that from.jneutron said:
I "put up" decades ago.
I have provided all a different viewpoint of a difficult (for most) concept. I provided a view, a testable hypothesis, a test methodology, constraints and confounders for that test, and most importantly, predictions of the outcome.
And in the decade I floated this, PMA is the only one to actually put the effort into proving it correct or incorrect.
Granted, he has yet to follow all the constraints (slew rate), perhaps he will, it's up to him.
I also put up a very specific test of audibility of cable based changes to image with specific methodology and specific predicted outcomes. To wit, if the test shows no audible change, then the cables cannot make a difference.
If anybody has a problem with a different understanding that comes with explanations, tests, predictions, and is fully testable and repeatable, especially with already established readily available equipment, then they are not being scientific nor reasonable.
That goes for the objectivists as well as the subjectivists.
As to your statement "put up or shut up", that is childish.
It is not about being right or wrong, it is about follow through.
Jn
Me either. But he is the first I've seen to actually do a 200 element line with more than just L and C.
So the question remains, will it be possible to get accurate coupler results even with impedances in place of resistances? And to what degree will the coupler output amplitudes be affected by the complex impedance of the cables at low frequencies. The RF people, even HP claim their devices as good (IIRC) well down below 10 kHz, are they missing something?
Within his article, at 10 kHz, he shows the results of the supra reflecting with matched load, open load, and shorted. In all three cases they match exactly the prediction of reflection coefficient. No reflection matched, negative shorted, positive open, and in phase. If the complex impedance of the cable didn't trash his measured outcome, why?
He also shows results with a complex impedance load
Jn
No. You have been, as always, above reproach.
Jn
Sorry I didn't know, I've only seen Pavel's tests here. "Put up or shut up" is a well know phrase I thought you'd be familiar with, it was meant as encouragement 🙄
An R/C only approximation does not support the concept of propagation velocity. An R/C ladder will have current in all the R's at t = 0+ and hence it is super-luminal. Do you have a distributed model that supports a sudden transition in behavior?
If I take say 1000mi of low loss cable TV distribution coax and put in a raised cosine pulse at say 10kHz so all the energy is centered around 10kHz are you saying nothing at all comes out at 8 msec or so later? Propagation by diffusion only would be limited by the drift velocity I assume (I might be wrong).
Last edited:
I am really not sure what you mean when saying that "I should follow all the constraints (slew rate)". Sounds to me that I am supposed not to know what slew rate is and when it is important.
Slew rate, by the definition known to me, is a maximum of the derivative of the signal. For the sine wave, it is the slope at the zero crossing, for exponential (1 - exp(-t/Tau)) it is a tangential, slope from right at zero.
For linear passive circuits, slew rate is unimportant and it is just a property of the signal. For non-linear active circuits, like amplifiers, we may speak about slew rate limit which would be the fastest rate of change that amplifier is able to yield at its output.
If I care or do not care about slew rate can be seen from a measurement of one of my preamplifiers
The same type of measurement I am doing for every of my designs.
- Status
- Not open for further replies.