Why do RCA cables have a 75 Ohm impedance?

[FriedChicken]

I don’t get it.

does this mean cables of different lengths need to be made of different materials?

where does the impedance even come from?

 

[xaviescacs]

I believe digital RCA cables have this Impedance to reject low voltage signals, noise, loosely speaking. The impedance of a coaxial cable depends on its materials and geometry.

All analog cables are made to have the less impedance resistance possible, I guess.

We'll have to wait for an educated member to show up to know the precise answer though.

 

[Katji]

There was something about it a couple days ago...

.............

Different cable structures for RCA interconnects
Suffice to say you need a 75 ohm coaxial cable (often called "video" cable) for the best results in terms of an RCA cable. The shield works fine as the neutral wire, since it should actually be called the earth or ground wire. It is therefore not necessary to shield the neutral wire, but you are...
Yesterday at 10:26 PM

https://www.audiosciencereview.com/forum/index.php?search/30890694/&q=75+Ohm&o=date

 

[xaviescacs]

https://www.audiosciencereview.com/...r-rca-interconnects.29589/page-2#post-1097790

 

[Joachim Herbert]

If you care about the details: https://en.wikipedia.org/wiki/Characteristic_impedance

 

[xaviescacs]

If you care about the details: https://en.wikipedia.org/wiki/Characteristic_impedance

To answer this...

does this mean cables of different lengths need to be made of different materials?

the key part being...

"[...] for any transmission line, and for well-functioning transmission lines, with R and G both very small, or ω very high, or all of the above, we get Z_0 ~ sqrt(L/C)"

which explains why at high frequencies the length of the cable doesn't matter.

 

[Berwhale]

If you care about the details: https://en.wikipedia.org/wiki/Characteristic_impedance

Reading that reminded me of the nightmare that was 10BASE2 Ethernet cabling.

 

[FriedChicken]

Reason I'm asking is because I will need to hook a pre-amp to an amplifier, and I'm not sure if I need anything premium or otherwise, or if I should buy cables with specific properties.

I started reading about coaxial cables, and saw that most of the information is relevant to video. I'm also confronted with potentially buying "premium" RCA cables vs those redubiquitous red/white RCA cables. I'm trying to work with distortions as low as 0.001% THD + N, so that's what I'm aiming for.

 

[EB1000]

Char impedance has nothing to do with the lumped impedance of the cable. If a cable (two wires) is treated like a transmission line, it means that the cable is caring a signal (sine wave for example) whos frequency is high enough so the signal's wavelength is significantly smaller that the cable's length (at least x10 times smaller). This makes the cable a distributed system, in which Kirchhoff laws no longer holds (they need to be replaced with a more generalized set of equations, called the Maxwell equations, which are not only function of time only like KVL/KCL. but functions of both distance and time.

The char imp, Zo is a ratio between the magnitudes of the traveling voltage and current waves in the same direction (voltage or electric field. current or magnetic field). To calculate the cable's lumped impedance (not distributed), we assume at the sending end we have V1, I1, and V2 and I2 at the receiving end. The we can take the voltage drop of the cable V2-V1 and divide it by the average current (I1+I2)/2.

Note that the cable has two types of impedance elements. Series elements like copper resistance R and wire inductance L, and parallel elements, like capacitance C (wire to shield), and leakage conduction G (between lead wire and shield due to dielectric contamination)

So, the longer the cable's length, the greater the series impedances R and wL, and the lower the parallel ones, 1/wC and 1/G. So, the total impedance tend to remain stable along the length, which in a way, explains the need for the char impedance - a quantity that is not affected by length of the cable.

One more thing, if the cable length is infinite, then Zo will be the same as the actual impedance along the cable...

 

[FriedChicken]

Char impedance has nothing to do with the lumped impedance of the cable. If a cable (two wires) is treated like a transmission line, it means that the cable is caring a signal (sine wave for example) whos frequency is high enough so the signal's wavelength is significantly smaller that the cable's length (at least x10 times smaller). This makes the cable a distributed system, in which Kirchhoff laws no longer holds (they need to be replaced with a more generalized set of equations, called the Maxwell equations, which are not only function of time only like KVL/KCL. but functions of both distance and time.

The char imp, Zo is a ratio between the magnitudes of the traveling voltage and current waves in the same direction (voltage or electric field. current or magnetic field). To calculate the cable's lumped impedance (not distributed), we assume at the sending end we have V1, I1, and V2 and I2 at the receiving end. The we can take the voltage drop of the cable V2-V1 and divide it by the average current (I1+I2)/2.

Note that the cable has two types of impedance elements. Series elements like copper resistance R and wire inductance L, and parallel elements, like capacitance C (wire to shield), and leakage conduction G (between lead wire and shield due to dielectric contamination)

So, the longer the cable's length, the greater the series impedances R and wL, and the lower the parallel ones, 1/wC and 1/G. So, the total impedance tend to remain stable along the length, which in a way, explains the need for the char impedance - a quantity that is not affected by length of the cable.

One more thing, if the cable length is infinite, then Zo will be the same as the actual impedance along the cable...

What about for very low frequency signals over very short wires?

 

[MakeMineVinyl]

Reason I'm asking is because I will need to hook a pre-amp to an amplifier, and I'm not sure if I need anything premium or otherwise, or if I should buy cables with specific properties.

I started reading about coaxial cables, and saw that most of the information is relevant to video. I'm also confronted with potentially buying "premium" RCA cables vs those redubiquitous red/white RCA cables. I'm trying to work with distortions as low as 0.001% THD + N, so that's what I'm aiming for.

Characteristic impedance is irrelevant to audio frequency transmission. For digital S/PDIF and video, yes it matters.

 

[FriedChicken]

Characteristic impedance is irrelevant to audio frequency transmission. For digital S/PDIF and video, yes it matters.

So I'll be ok with a bundle of cheap/simple cables going from the pre-amp to the amplifier? No interference/crosstalk/whatever?

 

[EB1000]

What about for very low frequency signals over very short wires?

Give an example. Say a 0.5m cable ans 20Hz. The length to wavelength ratio will be almost zero, thus the system is lumped. For short cables to become a problem, you need to be in the high MHz range

 

[Hayabusa]

What about for very low frequency signals over very short wires?

Then it will be not 75 ohm

 

[MakeMineVinyl]

So I'll be ok with a bundle of cheap/simple cables going from the pre-amp to the amplifier? No interference/crosstalk/whatever?

You can use most any simple cable to go from your preamp to power amplifier, or any other analog audio connection for that matter. Characteristic impedance only matters at very high frequencies, well beyond audio, such as RF or digital.

 

[egellings]

For analog audio signals, characteristic impedance of a cable has no bearing at all.

 

[SIY]

What everyone else said. Get a cable with mechanically sound connectors and sufficient flexibility. That's all that really matters.

 

[MRC01]

You can use most any simple cable to go from your preamp to power amplifier, or any other analog audio connection for that matter. Characteristic impedance only matters at very high frequencies, well beyond audio, such as RF or digital.

Yep. I think it would have to be beyond MHz, into the GHz range.

Please fact-check my basic high school physics:

v = F*l
v = speed of propagation (3e8 in vacuum, about 2.7e8 in copper wire)
F = frequency
l = lambda / wavelength

A 20 kHz wave traveling along copper wire has a wavelength of about 13.5 kilometers. Much longer than any interconnect cable. Characteristic impedance doesn't matter.
For the wavelength to be 1/10 of the cable length with a typical 2 meter cable, lambda is 0.2 so the frequency is 1.35e9 which is 1.35 GHz.
Is digital audio via SPDIF even that high?
Digital audio at 192 kHz / 24 bit / stereo is a raw data rate of about 9.2 Mbps (192e3 * 24 * 2).
If each bit is one pulse, that's more than 100x lower than 1.35 GHz.
Sure the format adds some overhead bits to fill data blocks, but the overhead ain't 100:1.

So characteristic impedance definitely doesn't matter at audio frequencies.
Does it matter for digital audio? It seems like a borderline case, maybe even that is not a high enough frequency to matter.

 

[DonR]

Yep. I think it would have to be beyond MHz, into the GHz range.

Please fact-check my basic high school physics:

v = F*l
v = speed of propagation (3e8 in vacuum, about 2.7e8 in copper wire)
F = frequency
l = lambda / wavelength

A 20 kHz wave traveling along copper wire has a wavelength of about 13.5 kilometers. Much longer than any interconnect cable. Characteristic impedance doesn't matter.
For the wavelength to be 1/10 of the cable length with a typical 2 meter cable, lambda is 0.2 so the frequency is 1.35e9 which is 1.35 GHz.
Is digital audio via SPDIF even that high?
Digital audio at 192 kHz / 24 bit / stereo is a raw data rate of about 9.2 Mbps (192e3 * 24 * 2).
If each bit is one pulse, that's more than 100x lower than 1.35 GHz.
Sure the format adds some overhead bits to fill data blocks, but the overhead ain't 100:1.

So characteristic impedance definitely doesn't matter at audio frequencies.
Does it matter for digital audio? It seems like a borderline case, maybe even that is not a high enough frequency to matter.

Luckily, digital signals are far more robust to cable impedance than analog signals. A misshapen 1 is still a 1 as long as it's not too out of whack.

 

[MakeMineVinyl]

Yep. I think it would have to be beyond MHz, into the GHz range.

Please fact-check my basic high school physics:

v = F*l
v = speed of propagation (3e8 in vacuum, about 2.7e8 in copper wire)
F = frequency
l = lambda / wavelength

A 20 kHz wave traveling along copper wire has a wavelength of about 13.5 kilometers. Much longer than any interconnect cable. Characteristic impedance doesn't matter.
For the wavelength to be 1/10 of the cable length with a typical 2 meter cable, lambda is 0.2 so the frequency is 1.35e9 which is 1.35 GHz.
Is digital audio via SPDIF even that high?
Digital audio at 192 kHz / 24 bit / stereo is a raw data rate of about 9.2 Mbps (192e3 * 24 * 2).
If each bit is one pulse, that's more than 100x lower than 1.35 GHz.
Sure the format adds some overhead bits to fill data blocks, but the overhead ain't 100:1.

So characteristic impedance definitely doesn't matter at audio frequencies.
Does it matter for digital audio? It seems like a borderline case, maybe even that is not a high enough frequency to matter.

I don't really get into RF type stuff. I'll leave that to the ham radio crowd.