RCA versus Coax , explain me that ....

[Cbdb2]

Seems to be some confusion. RCA is a connector so is BNC, coax is a cable construction. Most RCA cables are coax but there characteristic Impedance can vary a lot.

 

[Yasser 06]

Not really but we use coax and rca as shortcuts.

What i don't get is why people speak about the cable impedance, for a wire impedance is always per meter and so it shoud be very close to zero (per meter) and so the impedance should be mostly the one of the two connectors. Of course you could design the cable taking into account the impedance of the wire but then your connectors would depend on the cable, and any adjustement to the wire length would change the cable impedance. My understanding is that order of magnitude are such that the wire impedance is neglectable for a cable with reasonable length.

May be you can clarify that.

 

[sergeauckland]

Not really but we use coax and rca as shortcuts.

What i don't get is why people speak about the cable impedance, for a wire impedance is always per meter and so it shoud be very close to zero (per meter) and so the impedance should be mostly the one of the two connectors. Of course you could design the cable taking into account the impedance of the wire but then your connectors would depend on the cable, and any adjustement to the wire length would change the cable impedance. My understanding is that order of magnitude are such that the wire impedance is neglectable for a cable with reasonable length.

May be you can clarify that.

You're confusing the resistance of the cable, which is negligible, with the characteristic impedance at RF, which is typically between 50-100 ohms. Digital audio and analogue TV use 75 ohms cables, RF transmitter outputs are typically 50 ohms.
Characteristic impedance is only relevant above about 1MHz, or for very long cables, lengths over several 10s of kilometers. It has no relevance at audio frequencies or for normal domestic lengths.

S

 

[antennaguru]

Not really but we use coax and rca as shortcuts.

What i don't get is why people speak about the cable impedance, for a wire impedance is always per meter and so it shoud be very close to zero (per meter) and so the impedance should be mostly the one of the two connectors. Of course you could design the cable taking into account the impedance of the wire but then your connectors would depend on the cable, and any adjustement to the wire length would change the cable impedance. My understanding is that order of magnitude are such that the wire impedance is neglectable for a cable with reasonable length.

May be you can clarify that.

You're thinking about DC.

In the world of AC, and even more so as the AC frequency climbs higher, transmission line theory comes into play and the source (a transmitter as such) is designed to deliver its output to a particular characteristic impedance load. The load however is not right at the source, so there is a connection cable in between the source and the load. The characteristic impedance of the connection cable must match that of the load, and any mismatch from that characteristic impedance will cause reflections that will reduce power transfer.

Coaxial cable characteristic impedance is a result of the ratio of the outer diameter of the inner conductor to the inner diameter of the outer conductor, influenced by the dielectric constant of what insulation (dielectric) is between the inner and outer conductors. There is plenty of math that goes along with this but I am keeping it simplified here.

Characteristic impedance is not measured with a Volt-Ohmeter, but is measured with a Vector Network Analyzer, at different scattering parameters. OK, that's enough for now...

 

[radix]

You're confusing the resistance of the cable, which is negligible, with the characteristic impedance at RF, which is typically between 50-100 ohms. Digital audio and analogue TV use 75 ohms cables, RF transmitter outputs are typically 50 ohms.
Characteristic impedance is only relevant above about 1MHz, or for very long cables, lengths over several 10s of kilometers. It has no relevance at audio frequencies or for normal domestic lengths.

S

Digital audio can easily get into the 20 MHz clock rate for high-speed sampling. Even a "tame" 48 kHz sample has a 6.144 MHz biphase clock. I believe each biphase pulse is 20% rise, 20% fall, so maybe 32ns rise/fall time for 48 kHz? I've not seen an FFT of that signal so I don't know its BW, but it must require at least a handful of harmonics, maybe 50 - 100 MHz (that's a wild guess)?

 

[LTig]

Not really but we use coax and rca as shortcuts.

What i don't get is why people speak about the cable impedance, for a wire impedance is always per meter and so it shoud be very close to zero (per meter) and so the impedance should be mostly the one of the two connectors. Of course you could design the cable taking into account the impedance of the wire but then your connectors would depend on the cable, and any adjustement to the wire length would change the cable impedance. My understanding is that order of magnitude are such that the wire impedance is neglectable for a cable with reasonable length.

May be you can clarify that.

Maybe this can help.

 

[antennaguru]

BTW, most people have probably noticed that 75 ohm characteristic impedance coaxial cable is used for low power and receiving systems such as CATV and digital audio, where the loss over distance matters (CATV sends signals by the mile), and 50 ohm characteristic impedance coaxial cable is used for high power transmitting such as police, fire and broadcast transmitters.

The reason why that distinction was made back in the 1940s when coaxial cable first came into popular use was because for its diameter 75 ohm coaxial cable has lower loss compared to the same length/diameter 50 ohm coaxial cable, while 50 ohm coaxial cable handles higher power for the same length/diameter cable. When you look at the cross section of the two impedance coaxial cables you will note the relatively larger diameter center conductor of 50 ohm coaxial cable compared to 75 ohm coaxial cable. The larger center conductor allows for great power handling, but is not as far spaced from the outer conductor making it slightly lossier.

Finally, coaxial cable is only one form of constant impedance transmission line, and we also have micro-strip line on circuit boards and in metal boxes as a flat conductor spaced over a ground plane, used mostly for very high and microwave frequencies. There is also high characteristic impedance parallel conductor transmission line used mostly at low HF (Short Wave) and Medium Wave frequencies.

 

[Audiofire]

Most RCA cables are coax

Only for S/PDIF.

 

[SIY]

Only for S/PDIF.

No. I can't find a single non-coaxial RCA phono cable in my huge pile. Not one. Other than goofy "designer" stuff, they're all coaxial.

 

[Audiofire]

No. I can't find a single non-coaxial RCA phono cable in my huge pile. Not one. Other than goofy "designer" stuff, they're all coaxial.

You write no yet, the dictionary definition begs to differ:
"coaxial: (of a cable or line) transmitting by means of two concentric conductors separated by an insulator." Source: Oxford Languages

Definitely not how a generic RCA cable is made as well as most other RCA cables ever sold. Coaxial cables are specifically made with a characteristic impedance for high bandwidth.

 

[SIY]

You write no yet, the dictionary definition begs to differ:
"coaxial: (of a cable or line) transmitting by means of two concentric conductors separated by an insulator." Source: Oxford Languages

Definitely not how a generic RCA cable is made as well as most other RCA cables ever sold. Coaxial cables are specifically made with a characteristic impedance for high bandwidth.

That is absolutely the way generic RCA cable is constructed.

 

[Tom C]

I do have a handful of super cheap, super old rca that has two conductors side by side, like lamp cord. Think it might have come as a freebie with something else, like a Wii or an AVR. Certainly not common.

 

[Jimbob54]

I had some of these once. Cos What-hifi told me to. (The RCA, not the speaker)
I wouldnt describe their construction as the usual.

 

[Vincent Kars]

What is a coax cable?

"coaxial: (of a cable or line) transmitting by means of two concentric conductors separated by an insulator." Source: Oxford Languages

Not a bad definition but lets try to understand it (I have the feeling not everybody does)

Is this what the average "RCA cable" looks like.
Yes it does, just take any of your "RCA cables" and cut it.

Is it a RCA cable?
No, it is a coax cable but most of the time in the audio world it is terminated with a RCA plug.

Ok, let pull out a multimeter and measure its resistance.
It will be 0 Ohm so what is this nonsense about 75 ohm resistance.

You're confusing the resistance of the cable, which is negligible, with the characteristic impedance at RF, which is typically between 50-100 ohms. Digital audio and analogue TV use 75 ohms cables, RF transmitter outputs are typically 50 ohms.

Spot on. It is not about DC resistance but about characteristic impedance.
Analog audio will contain signals up to 22 kHz or even 192 kHz/2 , whatever sample rate you choose.

SPDIF is 32 bit pakkage containing 24 bit payload, if you run 44.1 kHz, you have this bus running at 5.6 MHz!
Indeed that is why characteristic impedance becomes important.

Coaxial cables are specifically made with a characteristic impedance for high bandwidth.

Obvious they are not.
The average analog cable (RCA) is sold as what it is, a coax without any spec of its characteristic impedance. Just check Amazon Basics or Monoprice.
However you can tune them like
50 Ohm like RG 50: https://www.caledonian-cables.com/product/Coaxial Cables/RG Type Low Loss 50 Ohm/RF 50LTA.htm
75 Ohm for SPDIF
110 Ohm for AES/EBU

Note that a 75 Ohm SPDIF cable terminated with RCA is not the "perfect" cable as RCA has no 75 ohm Impedance.
The right BNC termination does it better: https://www.thewelltemperedcomputer.com/HW/Connect/SPDIF_connect.htm

AES/EBU is balanced and terminated with a XLR, this plug is not impedance matched.

To sum up:
The average "RCA cable" is most of the time a coax cable terminated with a RCA plug.
You can use it safely for analog audio.
Fort short runs, you can even use it for digital signal transmission, this is know as the "quarter wave length"

For SPDIF you better use a cable with a characteristic impedance of 75 Ohm. As this about MHz frequencies only, you can also use it without any problem for analog audio.
Mutatis mutandis AES/EBU.

Bit more detail: https://www.thewelltemperedcomputer.com/KB/DigitalCable.htm

 

[Speedskater]

Any cable that has uniform end-to-end construction,
be it: coax, Shielded Twisted Pair (STP), Unshielded Twisted Pair (UTP), Zip cord or Twin-lead (see post #33)
will have a Radio Frequency Characteristic Impedance.

 

[sam_adams]

The links below do a reasonable job of explaining Characteristic Impedance:

http://www.renardson-audio.com/impedance.html

https://www.empiricalaudio.com/computer-audio/audio-faqs/should-i-match-impedances

You can have some fun calculating the Characteristic Impedance of an RCA/IHF connector:

https://www.everythingrf.com/rf-calculators/coaxial-cable-calculator

But you will need to know what the Dielectric Constant of the insulator is in the connector (good luck guessing what it's made of):

https://omnexus.specialchem.com/polymer-properties/properties/dielectric-constant

While the S/PDIF specification calls for 0.5 V - 0.6 V—with a minimum 0.2 V—if you examine the output of various devices at the COAX output jack, you will find that the signal level can be quite a bit higher, sometimes as high as 7 V into a 75 Ω termination.

 

[LTig]

The average analog cable (RCA) is sold as what it is, a coax without any spec of its characteristic impedance. Just check Amazon Basics or Monoprice.
However you can tune them like
50 Ohm like RG 50: https://www.caledonian-cables.com/product/Coaxial Cables/RG Type Low Loss 50 Ohm/RF 50LTA.htm
75 Ohm for SPDIF
110 Ohm for AES/EBU

I make my analog RCA cables with RG59 which has a characteristic impedance of 75 Ohm so I can use them for both analog and SPDIF connections. No need to check just grab one with the proper lenght ...

 

[mhardy6647]

You write no yet, the dictionary definition begs to differ:
"coaxial: (of a cable or line) transmitting by means of two concentric conductors separated by an insulator." Source: Oxford Languages

Definitely not how a generic RCA cable is made as well as most other RCA cables ever sold. Coaxial cables are specifically made with a characteristic impedance for high bandwidth.

That is absolutely the way generic RCA cable is constructed.

Don't make me start cuttin' into old unbalanced cables and snappin' photos to offer evidence in support of @SIY's assertion!

 

[Speedskater]

a] Most papers on Radio Frequency Characteristic Impedance, use the short form formula that only works well at frequencies from about 1 MHz and up.
b] Transmission line theory only applies to cables that a a good fraction of a wavelength long (and longer). For SPDIF that about 30 feet (10 meters).

 

[Audiofire]

That is absolutely the way generic RCA cable is constructed.

RCA cable I have seen has two copper wires, both separately insulated or there would be a short circuit. Don't know how often they are shielded, but I suppose they are almost always shielded with aluminum foil.

The foil is merely an additional measure for redirecting external signals to ground.