The ratio of C and L fixes characteristic impedance. The product of C and L fixes velocity ratio. You need the former fixed to avoid signal reflections. You need the latter fixed so all signals arrive at the far end at the same time. Therefore C and L are fixed.
what you show applies to one cable, usually a cable pair or a coax pair (core+screen)
but when you parallel multi strands, the capacitance must change.
Looks like that must be my first experiment. Confirm that a twisted pair is different from a star quad.
but when you parallel multi strands, the capacitance must change.
Looks like that must be my first experiment. Confirm that a twisted pair is different from a star quad.
Last edited:
Characteristic impedanse has no value at audio freq's- only valid for RF and higher speed pulse trains, say,- from 1MHz and upwards...
The proposal (post18) I made was about measuring capacitance of multistrand CAT5 speaker cables.
What has >1MHz got to do with that?
Are you expecting me to use a test frequency of > 1MHz?
What has >1MHz got to do with that?
Are you expecting me to use a test frequency of > 1MHz?
Characteristic impedance calculations demands the use of at least fractions of a wavelength, which for 20kHz is in the order of 10-15 km, ( choose your own Er ).
For lengths in the order of a handful of meters, only the discrete value of capacitance is of importance.... maybe also some inductaance, but this value will be rather smallish...
For lengths in the order of a handful of meters, only the discrete value of capacitance is of importance.... maybe also some inductaance, but this value will be rather smallish...
Not likely as both produce negative charges with teflon producing very little negative charge.
The RF characteristic impedance sets the C and L, which remain valid for audio frequencies even though the characteristic impedance changes due to the dominance of copper R. The RF characteristic impedance usually holds true from some kHz upwards - exactly where depends mainly on cable resistance.
As I said, if the cable has the same defined impedance for each pair and the same velocity for each pair then the C and L must be the same for each pair. This of course assumes normal use, where the pairs carry independent signals. It may be that the velocity varies somewhat over a short length, but averages to the same figure over a long length.
As I said, if the cable has the same defined impedance for each pair and the same velocity for each pair then the C and L must be the same for each pair. This of course assumes normal use, where the pairs carry independent signals. It may be that the velocity varies somewhat over a short length, but averages to the same figure over a long length.
You are right, of course, in the sense that f does not play part in the equations for char. impedance, hence the classical definition of 600 ohms audio lines, but on short lengths like this, it has no meanig as such. 1/10th of a wavelength used to be the classical limit. Common mode conditions and all that is of course still valid.
Cat 5 some specs:
DC RESISTANCE
The resistance of any conductor shall not exceed 28.6 Ohms per 1000 ft at or corrected to
a temperature of 200 C.
DC RESISTANCE UNABALANCE
The resistance unbalance between the two conductors of any pair shall not exceed 5%
when measured at or corrected to a temperature of 200 C.
MUTAL CAPACITANCE
The mutual capacitance of any pair at 1 kHz, measured in accordance with ASTM D 4566 is
not to exceed 5.6 nf per 100 m at or corrected to a temperature of 200 C.
CHARACTERISTIC IMPEDANCE*
The cable shall have a characteristic impedance of 100 Ohms? 15% for frequencies from 1
MHz to 100 MHz.
NEAR END CROSSTALK (NEXT)
Maximum values permitted for NEXT (rounded to the nearest dB) are listed in table 4 :
* PCB traces to the transformer and then to the connector should be laid out with a differential impedance of 100Ohms.
Should read 20 deg C didn't copy very well.
DC RESISTANCE
The resistance of any conductor shall not exceed 28.6 Ohms per 1000 ft at or corrected to
a temperature of 200 C.
DC RESISTANCE UNABALANCE
The resistance unbalance between the two conductors of any pair shall not exceed 5%
when measured at or corrected to a temperature of 200 C.
MUTAL CAPACITANCE
The mutual capacitance of any pair at 1 kHz, measured in accordance with ASTM D 4566 is
not to exceed 5.6 nf per 100 m at or corrected to a temperature of 200 C.
CHARACTERISTIC IMPEDANCE*
The cable shall have a characteristic impedance of 100 Ohms? 15% for frequencies from 1
MHz to 100 MHz.
NEAR END CROSSTALK (NEXT)
Maximum values permitted for NEXT (rounded to the nearest dB) are listed in table 4 :
* PCB traces to the transformer and then to the connector should be laid out with a differential impedance of 100Ohms.
Should read 20 deg C didn't copy very well.
Last edited:
At those impedances and signal levels, probably not an issue.
Actually, the combo is a classic case of things from the opposite ends of the triboelectric series creating charge.
18AWG AC line cord bought at any hardware store (very cheap) will sound just as good as any other wire. Silver is technically lower resistance than any other metal, but the difference is miniscule. The only real issue is the quality of the connectors. I use gold alloy plated banana connectors, and solder the wire at each end, rather than depending on those little screws to stay tight over time (they don't).
Lots of interesting stuff on Cat 5 cables and the like from a AES-PNW meeting.
AES PNW Meeting Report - Everything You Need To Know About Using Ethernet Cable for Portable Audio
AES PNW Meeting Report - Everything You Need To Know About Using Ethernet Cable for Portable Audio
As discussed in another thread.
Yet more Stephen Lampen Belden blog pages.
"Twisting the Night Away - Part 1"
Posted by: Steve Lampen on January 17, 2013
f you've ever played around with Category cables, such as Category 5, 5e, 6 or 6A, you might have noticed that the pairs in these cables are very precisely twisted. And, if you've got good eyes and look closely, you'll notice that the four pairs inside these cables are all twisted differently. The length of the twist is called the "lay length" so each pair has a different lay length. How we twist them, and how we decide what they lay length each pair should have is an interesting story.
Twisting the Night Away - Part 1
Yet more Stephen Lampen Belden blog pages.
"Twisting the Night Away - Part 1"
Posted by: Steve Lampen on January 17, 2013
f you've ever played around with Category cables, such as Category 5, 5e, 6 or 6A, you might have noticed that the pairs in these cables are very precisely twisted. And, if you've got good eyes and look closely, you'll notice that the four pairs inside these cables are all twisted differently. The length of the twist is called the "lay length" so each pair has a different lay length. How we twist them, and how we decide what they lay length each pair should have is an interesting story.
Twisting the Night Away - Part 1
Kevin - how significant an effect could those differences in lay length be expected (or have been modeled / tested) to have on audio frequencies when used at the short lengths we're generally talking about here?
I mean, of all the things to be obsessed over in an audio system, if short lengths / single pair of strands are employed in a low powered FR speaker, should using 2 feet of the brown pair on one channel, and blue on the other be high on that list?
I mean, of all the things to be obsessed over in an audio system, if short lengths / single pair of strands are employed in a low powered FR speaker, should using 2 feet of the brown pair on one channel, and blue on the other be high on that list?
The only issue is measuring the 2 feet. If you measure them to length while twisted, the actual wire length will be different.
dave
By how much before it's audible or even measurable at audio frequencies is my question.
Does that mean that if we cut off soldered wires when swapping out drivers that we need to measure / correct to the mm the remaining wires to each pass-band's driver in all enclosures?
Does that mean that if we cut off soldered wires when swapping out drivers that we need to measure / correct to the mm the remaining wires to each pass-band's driver in all enclosures?
Small and probably not audible, but it is at least tidy to keep them the same colour.
When i am doing that i do make sure each pair is shortened the same.
dave
In general, reasonable differences in length of any type of speaker cable are not audible. And in this case the differences are much shorter the reasonable.
What's a "FR" speaker?
Why would low powered speakers use smaller wires than high powered speakers? It's the end-to-end resistance that matters.
Unless the wires are very thin (high resistance) or very widely spaced (high inductance), matching to with a metre or so will do. That assumes, of course, that the wires understand and obey the laws of physics. If you have wires produced by alien technology or snake-oil merchants then much closer matching may be needed.chrisb said:
- Status
- Not open for further replies.