Still no numbers on that graph? I get the impression soongsc knows the numbers will show his impedance plot dosnt prove anything and thats why he wont show them. My suggestion would be to nuke that graph and all the posts after it as a huge waste of every bodies time.
Last edited:
Soongsc, I have to agree that, without values, it is difficult to interpret the graph. However, no matter what you show, it will be criticized, that is also a fact.
No. In the audio range inductive reactance per unit length can be similar in size to the conductor resistance. Similarly, for capacitive reactance although that can be less of an issue. The sqrt(L/C) formula is a simplification which only applies at sufficiently high frequencies, typically from HF audio upwards. I think the full formula is sqrt((R+jwL)/(G+jwC)), where w is 2 pi f, R is conductor resistance, G is insulation conductance. At low frequencies (LF audio?) it simplifies to sqrt(R/jwC) - so characteristic impedance is capacitive, not resistive. Note that this is not the same as measured impedance (=1/jwC), which is also capacitive.jneutron said:
There is a lot of confusion out there regarding what characteristic impedance means at extremely low wavelength cables.
What makes it even worse, is it is darn near impossible to verify characteristic impedance at audio frequencies.
For loads which are lower than the cable Z, the load as seen by the amplifier terminals will be inductive by nature.
For loads which are higher than the cable Z, the load as seen by the amplifier terminals will be capacitive by nature.
For loads which match the cable Z as determined by sqr(L/C), the amplifier will see a pure resistance. The inductive storage within the cable will be exactly equal to the capacitive storage.
For non matched cases, the settling time of the system will be compromised.
For matched cases, the settling time of the system is the settling time of the drive voltage only, and the transit time to load is based on prop velocity.
Cheers, John
Last edited:
Cable length is not the issue I was raising, but operating frequency. Characteristic impedance does not depend on length, but it may depend on frequency.
Only at sufficiently high frequencies that sqrt(L/C) is a good approximation.
The general assumptions which you use are consistent with the guys who consider cables as rf devices, and they do not worry about frequencies which are lower than quarter wavelength, nevermind a thousanth.
That is what the rf guys believe. And for their applications, it is indeed correct.
It is entirely applicable at audio frequencies, but not in an obvious way you will find by a google search.
It involves the settling time of the line among other things. Too late in the day to elaborate for now, but if you wish to know more, I'll be happy to splain..tomorrow.. (even graphs, equations, and analysis, go figure...)
Cheers, John
No. Reread my post #763. Sqrt(L/C) is not true for low frequencies, not misapplied as you imply, but not true. Circuit theory shows that the full correct expression involves R and G, as well as L and C. See any decent graduate textbook on transmission lines.
That is why I wait for other people to also show something of their own. Then we have some real comparisons and discussion.
Gentlemen we need not be esoteric when talking about 8 Ohm transmission lines at 10Hz. This has all come around before, yawn indeed.
Last edited:
The current website, characteristic impedance and .00001 wavelength line lengths really are silly put together. Move on, I won't bother you anymore.
I think you hit on it John. Settling time is an important factor. That is why you see the peak in the second graph I posted. But I'd certainly appreciate your views on low frequency impedance extension. Especially below 30Hz.
I could not find this thread anywhere.
I'm not learning anything here, could you direct me to your thread.
Actually rather than completely abandon this thread I do have a blue jeans cable as well as a whole heap of chinese made crap that I could possibly do measurements on, with values on the y axis if people were interested
I also have cables designed for well into GHz region that I could possibly measure also.
I also have cables designed for well into GHz region that I could possibly measure also.
Last edited:
You`re right. As with 75ohms coaxes; they`re not 75ohm but they`re optimiced for a 75ohm circuit, simple as that.
Btw; my all time reference IC is a 75ohm tv-coax from Vivanco; KX-710/910. Two coaxes in parralell for each side., only signal in center core. Just no audible signature. The same coax is for some rason a killer digital-coax too, the only one I`ve heard outperform an optical AT&T.
http://www.diyaudio.com/forums/cons...y-contribution-site-if-anyone-interested.html
Skip the beginning unless you want to read some background info.
Also, dont expect graphs and such...I rely on common sense and ignorance at the same time.
Hi,
another day has passed and still no explanation of what was measured, nothing on how the measurements were done, nor how to put values on the wavy lines.
another day has passed and still no explanation of what was measured, nothing on how the measurements were done, nor how to put values on the wavy lines.
- Status
- Not open for further replies.