Difference between coax cable and shielded cable

[tkamiya]

What exactly is a difference between a shielded cable and a coax cable?  They are both coaxial in nature.  Coax is made for a certain impedance and RF attenuation and mode.  I know that....  but what really determines if a cable is a coax or a shielded cable?  (yes, I know coax is a form of shielded cable and vise versa)

I always wanted to know this.

[AG6QR]

Shielded cable is any cable with an outer shield.  Inside the shield may be multiple conductors, and they generally won't be coaxial with each other. In other words, if you slice a three-conductor stereo headphone cable apart, you won't find three concetric circles of three different diameters.  Instead, you'll find two ordinary (non-coaxial) wires surrounded by a conductive shield.

[tkamiya]

I see....  that's very true.

OK, then what's the difference between single center conductor shielded cable and a coax? 

[Berni]

The coax cable is as the name suggests coaxialy symmetrical.

This makes the center conductor evenly spaced to the shield at all points and directions in a cable. Combined with the dielectric material between the two this gives the cable a so called characteristic impedance. That's the 50 Ohm or 75 Ohm or 120 Ohm that you often hear.

In a normal multi conductor shielded cable there is no such even geometry and as such the characteristic impedance of the wires are all different and random.

This impedance is important in RF applications as driving and loading the cable with the same impedance as the cable causes the cables parasitic capacitance and inductance to "disappear" and allows signals from 0 to >1 GHz to be sent for miles along such a cable and still make it out the other end without significant distortion. How and why this works is a topic for a whole new very long thread.

[hagster]

Any cable where the conductor and shield are centered on the same axis is coax. It may of course have a different characteristic impedance depending on the ratio of the diameters and permittivity of the insulator.

However coax is only one way of having a controlled impedance. Many digital systems use twisted pair type wires to achieve differential signaling with a relatively controlled impedance(often 100ohms).

[T3sl4co1l]

I see....  that's very true.

OK, then what's the difference between single center conductor shielded cable and a coax? 

Who even makes such a thing yet does not market it as coax?

You could have one-conductor cable made up with foil screen and/or braid just like a multiconductor cable, but this would only have the effect of cheapening the insulation (e.g., using crummy old PVC instead of low loss PE or PTFE) and worsening the shield resistance or effectiveness, and probably giving poorer impedance (but that's already understood given the lossy PVC insulation).

Related note, I've heard of ferrite-loaded hookup wire, which gives reasonable attenuation (filtering -- absorption) at high frequencies.  It's a distributed ferrite bead.  Really oddball, mil spec I think.  No electric-field shielding (as you get from a foil or braid shield), nor magnetic shielding beyond the surface layer.

Tim

[tooki]

I see....  that's very true.

OK, then what's the difference between single center conductor shielded cable and a coax? 

Who even makes such a thing yet does not market it as coax?

Musical instrument cable, for example. Various audio cables are essentially coax, even if not marketed as such. Correct me if I’m wrong, but “coax” usually implies a solid center conductor. The shielded single conductor cables I’m thinking of use a stranded conductor. I also doubt they use controlled-impedance dielectrics.

Take a look here at the instrument cables and many of the audio cables: https://www.thomann.de/gb/cable_per_meter.html

[T3sl4co1l]

RG-174 is stranded, and an official type.

I've seen audio cables that are just, whatever.  A ground that's not woven, but all the strands wrapped around the same way, so they only provide a little electric shielding and no magnetic shielding (which is adequate for most audio purposes, where ambient electric fields are more offensive than magnetic; just don't run the cables by any transformers, motors or switches..).

Looks like they have that kind, among others, on the link.

The geometry is still coaxial, but of course because the shield is poor, the RF leakage will be high, and the probably PVC insulator will have poor RF losses (that these cables are also used for video is almost dubious; fortunately they're usually short, and analog video is only several MHz).

Tim

[soldar]

I've seen audio cables that are just, whatever.  A ground that's not woven, but all the strands wrapped around the same way, so they only provide a little electric shielding and no magnetic shielding (which is adequate for most audio purposes, where ambient electric fields are more offensive than magnetic; just don't run the cables by any transformers, motors or switches..).

Yup. Microphone cable is shielded but is not "coax" in the sense of built to offer a defined impedance at higher frequencies.

[tooki]

RG-174 is stranded, and an official type.

Hence why I said “usually”, not “always”.

I've seen audio cables that are just, whatever.  A ground that's not woven, but all the strands wrapped around the same way, so they only provide a little electric shielding and no magnetic shielding (which is adequate for most audio purposes, where ambient electric fields are more offensive than magnetic; just don't run the cables by any transformers, motors or switches..).

Looks like they have that kind, among others, on the link.

The geometry is still coaxial, but of course because the shield is poor, the RF leakage will be high, and the probably PVC insulator will have poor RF losses (that these cables are also used for video is almost dubious; fortunately they're usually short, and analog video is only several MHz).

Why would a nonwoven shield provide no EM shielding to speak of? Whether woven or coiled for flexibility, if the shield strands are all touching, wouldn’t they still form a largely continuous shield? It’s not litz wire where each strand is insulated, after all.

But doesn’t your reply pretty much confirm my answer to your question about who would make shielded single conductor without calling it coax? That they aren’t advertising it for good RF properties is hardly a mystery. Of course they’re fine for audio (or even some SD video), that’s what they’re made for!

[tooki]

I've seen audio cables that are just, whatever.  A ground that's not woven, but all the strands wrapped around the same way, so they only provide a little electric shielding and no magnetic shielding (which is adequate for most audio purposes, where ambient electric fields are more offensive than magnetic; just don't run the cables by any transformers, motors or switches..).

Yup. Microphone cable is shielded but is not "coax" in the sense of built to offer a defined impedance at higher frequencies.

Microphone cable is actually not coaxial, it’s a twisted pair with shield. The question was about single-conductor wires with shields. Instrument cables (like guitar cables) fall into that category. So does the stuff used in RCA audio cables and the like.

[T3sl4co1l]

Why would a nonwoven shield provide no EM shielding to speak of? Whether woven or coiled for flexibility, if the shield strands are all touching, wouldn’t they still form a largely continuous shield? It’s not litz wire where each strand is insulated, after all.

I didn't say that, I said E, but poor M.   Strands touching is unlikely; it will happen, but not often enough to be significant.

Even (lazy) braiding doesn't do all that great -- RG-58 is a sieve.

But doesn’t your reply pretty much confirm my answer to your question about who would make shielded single conductor without calling it coax? That they aren’t advertising it for good RF properties is hardly a mystery. Of course they’re fine for audio (or even some SD video), that’s what they’re made for!

It's still coaxial in the sense that the outer conductor is cylindrical and on the same axis as the center conductor.  Just depends how generally the term "coax" is meant.  I was taking the more general, geometric meaning.

Tim

[tooki]

We must not be thinking about the same kind of coiled shield. The ones I’m familiar with have enough strands that frequent contact is inevitable. (Like, obviously not every strand will be in direct contact with every other strand, but enough that cumulatively, a reasonably dense mesh covers every part.)

I understand that you were talking about geometry. You expressed skepticism that anyone would make such a cable with a coaxial construction and not market it as coax. I provided examples of exactly this.

[tkamiya]

Actually, very cheap microphone cables are coaxial.  $10 microphone from RadioShack used to have this shielded cable with one center conductor with very little braid.  I think all pro stuff uses balanced transmission so they are twisted pair in shield.

[T3sl4co1l]

We must not be thinking about the same kind of coiled shield. The ones I’m familiar with have enough strands that frequent contact is inevitable. (Like, obviously not every strand will be in direct contact with every other strand, but enough that cumulatively, a reasonably dense mesh covers every part.)

We're talking quite low impedances to provide adequate magnetic shielding -- lateral and axial resistivity near that of solid copper!

Related subject -- stranded wire does in fact save some AC resistance over solid.  It's not as good as proper litz, but it's surprisingly not that much worse (given you may not have much choice over the strand size, of course).

Which is equivalent to saying: stranded wire is more transparent to magnetic fields than solid wire is.

The same will be true here, and is true of braid as well, though obviously braid does much better at acting solid than the swirl stuff, but again not as good of a shield as solid (semirigid).

Another way to put it: the contact resistance increases the cross-strand impedance, significantly increasing the skin depth in that direction, therefore allowing magnetic fields inside.

I understand that you were talking about geometry. You expressed skepticism that anyone would make such a cable with a coaxial construction and not market it as coax. I provided examples of exactly this.

Yeah, foil cable came to mind first, didn't think of the spiral stuff at the time.

Tim

[Berni]

Well in terms of the shield braiding there is a ton of difference between nice high quality cables and cheap crap.

Some are nice thick tightly wound ones on the nice cables or sometimes there is essentially a few strands of it wrapped around with gaping holes in between, some include a foil under the braid some not etc... There are also the double shielded triaxial cables for when you really want to shield things well and proper.

But yeah for single conductor shielded cables id say the only reason its not called a coax cable is that they are not built according to any coax cable standard that prescribes the dimensions, materials and of course the all important characteristic impedance in Ohms. If a cable can't be trusted to be 50 Ohm (or whatever else you need) then you can't use it as RF cable. But you could certainly use real RF coax as a single conductor shielded cable if you wanted.

Another extreme example of such cables are the cables used to connect the 3 phase motor drive to the BLDC motor in electric vehicles. They are huge 50mm2 copper and up wires that are surrounded by a metal shield to keep the wires radiating out all the motor controller switching noise and causing interference.

[tkamiya]

I wonder if anyone with VNA would be willing to characterize a good quality shielded 1 conductor cable....?  Then compare it against say RG174.  Unfortunately, this is well outside of my capability.

[Berni]

I wonder if anyone with VNA would be willing to characterize a good quality shielded 1 conductor cable....?  Then compare it against say RG174.  Unfortunately, this is well outside of my capability.

You probably wouldn't see much interesting. As long as its good even cable you would get a RF cable with a weird impedance of like 63.8 Ohm. It might have more loss or less loss than RG174 depending on what type of dielectric is used inside.

The weird impedance would make it difficult to use for RF applications because most RF things are designed to have certain commonly used impedance such as 50 Ohm or 75 Ohm. So in order to drive the cable you would need a impedance matching transformer that takes 50 Ohm and boosts the voltage slightly to drive the 63.8 Ohm cable and on the receiving end you need another one of those transformer but connected in reverse so that it drops the impedance from 63.8 Ohm back down to 50 Ohm. If you don't do impedance matching and connect it directly then RF power will bounce back when it enters the cable so not all of the RF power will even make it into it, and once it reaches the end of the cable and hits 50 Ohm again more RF power will reflect back instead of going into the load. This reflection can happen multiple times making the signal bounce around, getting in phase in some spots and adding or out of phase and subtracting, as a result making a cable with a very wibbly wobly freqency response.

[vk6zgo]

I see....  that's very true.

OK, then what's the difference between single center conductor shielded cable and a coax? 

Who even makes such a thing yet does not market it as coax?

You could have one-conductor cable made up with foil screen and/or braid just like a multiconductor cable, but this would only have the effect of cheapening the insulation (e.g., using crummy old PVC instead of low loss PE or PTFE) and worsening the shield resistance or effectiveness, and probably giving poorer impedance (but that's already understood given the lossy PVC insulation).

Related note, I've heard of ferrite-loaded hookup wire, which gives reasonable attenuation (filtering -- absorption) at high frequencies.  It's a distributed ferrite bead.  Really oddball, mil spec I think.  No electric-field shielding (as you get from a foil or braid shield), nor magnetic shielding beyond the surface layer.

Tim

Single conductor shielded cable used to be common, but maybe, not so much now.
That "single" conductor is not concentric with the shield, as in coaxial cable, & the cable's impedance is not well specified..

In many cases, if twin conductor shielded cable is available, people will use it, instead, either ignoring one conductor, or paralleling the two.

Alternatively, they may use thin coax, just because it is widely available & durable, without actually needing the special characteristics coaxial cable brings.

The connectors commonly used with,"single ended" audio are not the most rugged, either..

An example:-
When I worked at the TV Studio, we had some "standalone" TV projectors, which we used to make available, free of charge, along with a VCR, to various community groups, as part of our commitment to being "good corporate citizens".

These needed one BNC to BNC video cable, & two RCA to RCA audio cables.
The former was easily found, but the latter was not normally used in the Studio, where audio was normally
balanced, not "single ended".
We, perforce, had to "make them up".

The store had twin audio cable, & some "Consumer grade" RCAs, so we would do so.
One problem was that the "twin" cable was quite a bit more  bulky than the cable the RCAs were intended for, & after one or two outings, the latter would fail, & need to be replaced.
They also needed to be specially stored, so they wouldn't be lost, something which happened regularly, so we would need to make new ones, all for no profit to the company!

Eventually, we'd "had enough", so we fitted BNC to RCA adaptors to the VCR & projector.
Now we only needed to provide three BNC to BNC coax cables.

[Berni]

Nothing wrong with replacing RCA with BNC. The RCA cables used for composite video are 75 Ohm and the BNC cables found in a TV studio are also going to be the 75 Ohm kind.

There is actually a problematic combination of 75 Ohm and 50 Ohm BNC connectors. I forgot what way around it is but i think plugging a 75 Ohm male into a 50 Ohm female BNC can damage one of the connectors because some of the dielectric inside them gets in the way.

[tautech]

Coax, shielded, who cares ?
Such generalization should be chastised here on EEVblog ! 
It's their intended application that really matters and I'll just drop this here for you all to pick at:



Source PDF attached.

[radiolistener]

The difference is insulator material and insulator geometry. Coax cable has specific low loss insulator material which size is stable along entire cable. It leads to stable relation between inductance and capacitance (impedance) along the cable length.

Usual shielded cable may use more cheap insulation and may not keep insulator geometry along cable length. It leads to variable relation between inductance and capacitance (impedance) which is very bad for high frequency signals. Such impedance variation leads to signal reflections and standing wave in the cable.

The RF energy is flowing in the insulator between central conductor and braid. This is why insulator properties and geometry are very important for coax cable.

[tautech]

Well isn't coax by definition a shielded cable ? Shielded by braid.

Then of course there's coax that's foil shielded and braid shielded too. 

[tooki]

Nothing wrong with replacing RCA with BNC. The RCA cables used for composite video are 75 Ohm and the BNC cables found in a TV studio are also going to be the 75 Ohm kind.

There is actually a problematic combination of 75 Ohm and 50 Ohm BNC connectors. I forgot what way around it is but i think plugging a 75 Ohm male into a 50 Ohm female BNC can damage one of the connectors because some of the dielectric inside them gets in the way.

Indeed, I am quite confident that most people don’t even realize there are different BNC connectors for 50 and 75 ohms. I certainly didn’t until I read about it here on the forums, even though I had a teeensy tiny bit of experience with both kinds, having used 75 ohm ones in broadcast video and on some 90s high end computer displays, and 50 ohm on old ThinNet Ethernet. But that was always with premade cables, and I knew enough to not try to feed Ethernet right into the RGBHV inputs of a display! :p

[Berni]

Actually i got it wrong. Its mechanically fine to mix 50 and 75 Ohm BNC connectors as the only difference is that the 75 Ohm have less dielectric inside.

Its the 75 Ohm N connectors that are dangerous because the center pin has slightly different dimensions and as a result plugging the larger 50 Ohm male into a 75 Ohm female can result in connector damage.