50 ohm vs 75 ohm vs 125 ohm coaxial cable

[Chrise36]

As i was searching for the difference between 50 ohm and 75 ohm coaxials i read that besides attenuation capacitance plays a big role in transferring digital signals and that 125 ohm cables have the least capacitance. I was wondering what the attenuation for these three cables is for a small length cable like 1m and if it would be better to use higher ohm than 75 ohm cabling. Maybe Amir could test something like this

 

[daftcombo]

I've always read to use 75 ohm coaxial cables, like CANARE ones. That's what I have at home.

Another interesting question would be about the max length of those cables before they get too much interferences.

 

[sergeauckland]

Unbalanced digital audio and analogue video use 75 ohm impedance so it's best to use that impedance cable. In the case of video, it matters even for short cables, digital audio is much less critical. For 5-10 metres or thereabouts, I don't bother to make sure the cable is 75 ohms, I use whatever's of the right length, and it seems to work fine. I've seen digital audio literally sent over wet string with no errors.

RF transmission generally uses 50 ohm cables, so again for lowest loss, (SWR) it's best to stick to 50 ohm cables. I haven't come across 125 ohm cables so can't comment.

S.

 

[Chrise36]

Unbalanced digital audio and analogue video use 75 ohm impedance so it's best to use that impedance cable. In the case of video, it matters even for short cables, digital audio is much less critical. For 5-10 metres or thereabouts, I don't bother to make sure the cable is 75 ohms, I use whatever's of the right length, and it seems to work fine. I've seen digital audio literally sent over wet string with no errors.

RF transmission generally uses 50 ohm cables, so again for lowest loss, (SWR) it's best to stick to 50 ohm cables. I haven't come across 125 ohm cables so can't comment.

S.

So you agree even a cheap short 50 ohm cable is enough for e.g a dac at home but are there any tests to show this?

 

[sergeauckland]

So you agree even a cheap short 50 ohm cable is enough for e.g a dac at home but are there any tests to show this?

I do agree that for short cables, impedance doesn't seem to matter. However, why would you would choose to use a 50 ohm cable, when RG59 nor RG6 cable sold as satellite or TV antenna cable costs pennies?

I haven't seen any tests published on this, as the magazines all bang on about how important (expensive) cables are, but my own tests indicate that any old bit of coax will do fine for a couple of metres. I even managed to get digital audio through my body, holding the ends of the cable, although that was a bit tenuous...I used an AES-EBU output into a fairly sensitive S-PDIF input, as the voltage drop across me was significant. What astounded me was that when it worked at all, it worked fine.

Some may recall Canford Audio showing their AES-EBU transmission system working over two lengths of wet string. It was wet because it went through a fish tank with real fish swimming about.

There's an awful lot of rot talked about the need for specific cables for digital audio. Analogue video does need carefully matched impedances, as any reflections due to mismatching are very visible as ghosting on the picture. This isn't the case with digital audio that as long as the eyeheight is sufficient for the receiver to lock, audio decoding will be perfect. It takes a massive mismatch for the eyeheight to be such that a receiver won't lock.

S.

 

[Chrise36]

L

I've always read to use 75 ohm coaxial cables, like CANARE ones. That's what I have at home.

Another interesting question would be about the max length of those cables before they get too much interferences.

Y

I do agree that for short cables, impedance doesn't seem to matter. However, why would you would choose to use a 50 ohm cable, when RG59 nor RG6 cable sold as satellite or TV antenna cable costs pennies?

I haven't seen any tests published on this, as the magazines all bang on about how important (expensive) cables are, but my own tests indicate that any old bit of coax will do fine for a couple of metres. I even managed to get digital audio through my body, holding the ends of the cable, although that was a bit tenuous...I used an AES-EBU output into a fairly sensitive S-PDIF input, as the voltage drop across me was significant. What astounded me was that when it worked at all, it worked fine.

Some may recall Canford Audio showing their AES-EBU transmission system working over two lengths of wet string. It was wet because it went through a fish tank with real fish swimming about.

There's an awful lot of rot talked about the need for specific cables for digital audio. Analogue video does need carefully matched impedances, as any reflections due to mismatching are very visible as ghosting on the picture. This isn't the case with digital audio that as long as the eyeheight is sufficient for the receiver to lock, audio decoding will be perfect. It takes a massive mismatch for the eyeheight to be such that a receiver won't lock.

S.

Your research is phenomenal! You could have sold your patent and make millions. The thing is with 3.5mm coaxials there are no 75 ohm terminals so you should use a 50 ohm cable and be good with it. The other thing is are there true 75 ohm rca terminals? How does this work with coaxial because bnc is more understandable.

 

[SIY]

So you agree even a cheap short 50 ohm cable is enough for e.g a dac at home but are there any tests to show this?

FWIW, I used a simple cheap RCA to RCA cable (one of the throwaways packed with something or other that we bought) with several ADC/DACs and had zero issues. Under a couple meters seems to be no problem.

 

[sergeauckland]

LY

Your research is phenomenal! You could have sold your patent and make millions. The thing is with 3.5mm coaxials there are no 75 ohm terminals so you should use a 50 ohm cable and be good with it. The other thing is are there true 75 ohm rca terminals? How does this work with coaxial because bnc is more understandable.

Frankly, my dear, I don't give a damn.

It just doesn't matter. RCA connectors (or 3.5mm coaxials) are what's usually used for S-PDIF, so that's what you get on most, even seriously (stupidly!) expensive cables.

BNCs are better, insofar as they can be proper 75 ohms, but then they get used with a BNC-RCA adapter

Unless your digital cables are very long, and by that I mean over 10 metres (that's the longest I've ever used at home) anything will do fine. There may be some DACs that have such poor receivers that they might just be upset by non 75 ohm cables, but none I've ever used have ever been bothered by the cable. Try it, if it works, it'll work fine. The worse that can happen is that it won't work at all. That's one of the nice things about digital. It either works, in which case it works well, or it doesn't work, in which case you'll get drop-outs and mutes and weird noises. There's NO WAY it can ever affect sound quality except to be muted.

S

 

[Chrise36]

Frankly, my dear, I don't give a damn.

It just doesn't matter. RCA connectors (or 3.5mm coaxials) are what's usually used for S-PDIF, so that's what you get on most, even seriously (stupidly!) expensive cables.

BNCs are better, insofar as they can be proper 75 ohms, but then they get used with a BNC-RCA adapter

Unless your digital cables are very long, and by that I mean over 10 metres (that's the longest I've ever used at home) anything will do fine. There may be some DACs that have such poor receivers that they might just be upset by non 75 ohm cables, but none I've ever used have ever been bothered by the cable. Try it, if it works, it'll work fine. The worse that can happen is that it won't work at all. That's one of the nice things about digital. It either works, in which case it works well, or it doesn't work, in which case you'll get drop-outs and mutes and weird noises. There's NO WAY it can ever affect sound quality except to be muted.

S

I agree with you and i think it is even better to use straight 50 ohm instead of 50 ohm terminals with 75 ohm cable if you cant find straight 75 ohm cable and terminals because of so called reflections or "mirroring" . I have not found a logical explanation about 75 ohm rca how do companies claim this ? Is there a resistor somewhere? And i am asking about capacitance because i think it could be more important than impedance if we (Amir) could measure and prove this.

 

[scott wurcer]

BNCs are better, insofar as they can be proper 75 ohms, but then they get used with a BNC-RCA adapter

I would think a cm of mismatch might matter to 10GHz backplanes but not audio. BTW I would think real men would find some ancient hermaphroditic ones like GR connectors, when mated they form a perfect air dielectric coaxial line.

 

[sergeauckland]

I agree with you and i think it is even better to use straight 50 ohm instead of 50 ohm terminals with 75 ohm cable if you cant find straight 75 ohm cable and terminals because of so called reflections or "mirroring" . I have not found a logical explanation about 75 ohm rca how do companies claim this ? Is there a resistor somewhere? And i am asking about capacitance because i think it could be more important than impedance if we (Amir) could measure and prove this.

The 'Impedance' of a cable is an RF phenomenon. If you measure a cable at DC, it has an open circuit i.e. infinite ohms impedance. At RF. i.e somewhere over 1Mhz, the impedance measured becomes the 'characteristic' impedance. That's why impedance matters at very high frequencies, but is of no consequence for audio frequencies. Digital audio is 'just' into the frequency range where it matters, which is why it matters for long cable lengths, but doesn't for domestic lengths.

S.

 

[Chrise36]

So a dsd recording might have issues but pcm is not affected

 

[sergeauckland]

So a dsd recording might have issues but pcm is not affected

Maybe, don't know about DSD, but then, just as with PCM, try it. It'll either work or it won't. There's no magic about DSD, it's just digital.

S.

 

[Speedskater]

Wait, it's not about the cable, it's about the transmission line system impedance!
Which is:
output stage >> cable >> input stage, and they all have to have the same impedance for best results.

If you have a well matched system then the extreme length of the cable is dictated by things other than the Radio Frequency Characteristic Impedance of the system.

 

[SIY]

I have not found a logical explanation about 75 ohm rca how do companies claim this ?

The 75 (or 50) ohms is not resistance, it's what's called "characteristic impedance," and since it has the units of ohms, it can be confusing if you don't understand EM theory to a physics level. The capacitance is one element of it, but there's several others as well (if you're curious, look up "telegrapher's equation," which came about because of lines hundreds or thousands of miles long). In the limit of relatively short interconnects (meters to tens of meters), you can approximate it as the square root of the ratio of inductance to capacitance. It is NOT a resistance or what we usually think of as an impedance.

Bottom line is that it's important at very high frequencies (tens to thousands of megahertz), unimportant at audio, and only very mildly important for digital audio transmission. If there's a severe enough mismatch, you'll get dropouts, pops, and the like. If someone claims "smoother bass" or "more extended imaging" or stuff like that, you can then know whom you can safely ignore.

 

[DonH56]

As i was searching for the difference between 50 ohm and 75 ohm coaxials i read that besides attenuation capacitance plays a big role in transferring digital signals and that 125 ohm cables have the least capacitance. I was wondering what the attenuation for these three cables is for a small length cable like 1m and if it would be better to use higher ohm than 75 ohm cabling. Maybe Amir could test something like this

For analog audio it is irrelevant. I used 93-ohm coax for my phono cables way back because it was lower capacitance and I had a bunch of it lying around.

For digital and video signals where frequencies are higher it is best to match the load and source impedance. A 125-ohm cable adds discontinuities at both ends and can cause problems depending upon the length of the cable. The mismatch will cause the source and load to "see" the capacitance where it is not "seen" in a perfectly-matched system. There are online tutorials, and I think one or two in the technical area of ASR, talking about mismatches and reflections and what they can do to the signal.

Loss is generally independent of the impedance. Ideally in a matched system there is no loss; loss comes from the resistance of the wire. If the 125-ohm cable has a smaller conductor than the 75-ohm cable it will have higher loss, but that won't matter for a short (e.g. 1 m) cable.

Basic impedance equation: Zo = sqrt[ (R+jwL) / (G + jwC) ]
where RLCG are resistance, inductance, capacitance, conductance; j = sgrt(-1), w = freq (radians/s)

G is usually neglected. When R is small the impedance is just sqrt(L/C); R is the lossy part of the line.

HTH - Don

edit: I got distracted so others have already answered...

 

[sergeauckland]

Wait, it's not about the cable, it's about the transmission line system impedance!
Which is:
output stage >> cable >> input stage, and they all have to have the same impedance for best results.

If you have a well matched system then the extreme length of the cable is dictated by things other than the Radio Frequency Characteristic Impedance of the system.

No they don't!!!!

For long cables perhaps, but what matters for 'normal' domestic lengths is the eyeheight. With any normal cable and a length of 10 metres or less (possibly more, never tried it), it'll work fine. All it has to do is to transfer the data. There is NO audio being transferred, only numbers and as long as those numbers get transferred correctly, the decoded audio will be fine.

As SIY said above, any suggestion that subjective audio parameters will be affected is just bunkum.

S.

 

[SIY]

As SIY said above, any suggestion that subjective audio parameters will be affected is just bunkum.

I don't think @Speedskater implied anything different. And since we've (virtually) known each other for years, I know that he's extremely knowledgeable and very aware of the transmission issues and non-issues.

 

[Eirikur]

I've used 50ohm BNC network coax for a length of ~15 meters without any problems for years.
Layout: SPDIF "jumpers" on PC motherboard to BNC female -> 15m cable -> BNCtoRCA -> amplifier

 

[Blumlein 88]

Some good info here.
https://www.rane.com/note149.html

RCA SPDIF is listed in a chart as 10 meters max. I've run the stuff at least twice that far with no problems using RG59 or RG6.

Balanced AES/EBU is the cheap way to go far as you can use regular XLR microphone cable of the proper spec. Or if you have proper BNC on boths ends RG6 will work to 1000 meters.

As others have said, at a few feet, almost anything will work that has wire in it.