Flat Ethernet Cales

[MrJim]

I'm planning to have engineered wooden flooring put down in my house. I'm wondering whether it's feasible to put flat Ethernet cabling underneath it. Has anyone tried this?

[Bagnaj97]

I did this at my last house, but make sure you get cables with a boot/cover over the latch. If the clip/latch snaps you can't easily run another cable!

Ideally you want an RJ45 socket on the wall at each end, then run the cable under the flooring. That way you can use a standard patch cable of appropriate length at each end, or, if you no longer need the cable, you don't have a random loose wire!

[jim]

I've done it under carpet, wasn't too bad. The only "issue", which is really a bit of a non-issue, was that I couldn't get Cat-6 flat cable and settled for Cat-5e. Not that it's hurt anything other than my need for unutilised speed.

[herulach]

If you're getting flooring done properly you'll be having the skirting done also I assume (if not get it done, its only a tiny cost in the scheme of things, and makes the job look much better than beading) which means you could run proper cable behind the skirting, and terminate in a box either at skirting height, or only have a tiny bit of making good to get to a more sensible socket height*

*disclaimer: I don't have any idea about wiring guidelines and whether they impact the routing of data cables

[aidanjt]

I've done it under carpet, wasn't too bad. The only "issue", which is really a bit of a non-issue, was that I couldn't get Cat-6 flat cable and settled for Cat-5e. Not that it's hurt anything other than my need for unutilised speed.

Whoever was selling you flat 'Cat. 5e' was a charlatan. It's the pair twisting specifications which gives the cable its category rating, which is also what makes it necessarily thick and round. Flat ethernet cables are always category nil.

[TheAnimus]

Whoever was selling you flat 'Cat. 5e' was a charlatan. It's the pair twisting specifications which gives the cable its category rating, which is also what makes it necessarily thick and round. Flat Ethernet cables are always category nil.

First clue is in the name. Twisted Pair.

The idea is to balance the signal, by having each 'pair' carry opposite PD.

There is no reason why you can't make the 'height' of the flat cable, one pair high. I mean afteral no cable is truly flat.

What the CATx specifications demand is a certain amount of permissible cross talk. Technically they need not be twisted much at all, so long as they comply with that.

So yeah, you can have CATx certifications regardless of the physical layout of your cables!

[peterb]

Whoever was selling you flat 'Cat. 5e' was a charlatan. It's the pair twisting specifications which gives the cable its category rating, which is also what makes it necessarily thick and round. Flat ethernet cables are always category nil.

No reason why the twisted pairs can't be laid side by side in a rectangular profile instead of bundled into an approximately round profile and still retain the CAT 5 or Cat 5e specification for common mode rejection, cross talk between pairs, attenuation and return loss.

Other cable forms manage a side by side configuration - SCSI ribbon cables (for example) are multiple twisted pairs in a ribbon cross section.

[aidanjt]

First clue is in the name. Twisted Pair.

The idea is to balance the signal, by having each 'pair' carry opposite PD.

There is no reason why you can't make the 'height' of the flat cable, one pair high. I mean afteral no cable is truly flat.

What the CATx specifications demand is a certain amount of permissible cross talk. Technically they need not be twisted much at all, so long as they comply with that.

So yeah, you can have CATx certifications regardless of the physical layout of your cables!

What are you on about? The TIA/EIA-568 specifications are vastly more comprehensive than just permissible crosstalk (crosstalk is a phenomenon which is only one of the causes of attenuation, and isn't the engineering goal in and of itself). Including the diameter of the cable (6.35mm for 5e), which doesn't leave any room for flattening, or being anything other than rounded, for that matter, considering two pairs stacked side by side are snug enough in that jacket outer diameter.

No reason why the twisted pairs can't be laid side by side in a rectangular profile instead of bundled into an approximately round profile and still retain the CAT 5 or Cat 5e specification for common mode rejection, cross talk between pairs, attenuation and return loss.

No reason aside from the cable diameter specification? Just because a semi-flattened UTP cable can carry an ethernet signal, it doesn't make it Cat. 5e standards compliant.

Other cable forms manage a side by side configuration - SCSI ribbon cables (for example) are multiple twisted pairs in a ribbon cross section.

The SCSI standards specifies the ribbon cable. TIA/EIA-568, doesn't.

[peterb]

No reason aside from the cable diameter specification? Just because a semi-flattened UTP cable can carry an ethernet signal, it doesn't make it Cat. 5e standards compliant.

Which was why I spoke about the critical electrical specification. But I accept that from the point of view of form factor, the flat cable may not be CAT5e compliant.

It does make me wonder how many other cat5e cables are compliant to the letter of the TIA standard.

[TheAnimus]

What are you on about? The TIA/EIA-568 specifications are vastly more comprehensive than just permissible crosstalk (crosstalk is a phenomenon which is only one of the causes of attenuation, and isn't the engineering goal in and of itself). Including the diameter of the cable (6.35mm for 5e), which doesn't leave any room for flattening, or being anything other than rounded, for that matter, considering two pairs stacked side by side are snug enough in that jacket outer diameter.

I've not got access to the direct specification as it is paywalled here, but wikipedia articles on Cat5, Cat5e, Cat6, TIA 568. None of them make reference to a phsyical specification, only the electric ones.

This makes sense to me, as ultimately depending on the purity of the conductor, differences in tolerances of the AWG used, it will require a slight different number of turns per pair. I can see nothing that requires each pair to be turned around another pair. In fact I'd have thought that sheilding each pair from other pairs would have provided more isolation (even if it isn't as needed due to being balanced TP).

The wikis look like they could use a cleanup. For insance:
http://en.wikipedia.org/wiki/Category_5_cable

Mentions that it is spec'd to be 24AWG. Yet the sources do no such thing, they merely state than an implementation claiming by the manufacturer to be compliant is 24AWG. Further down the page it suggests it's commonly 24-26AWG.

This is why I thought that the only physical constraints on the cabling was actually at the connectors! (which I'd also echo advice to OP, put a box up, don't just use an end, much neater!)

[peterb]

I've not got access to the direct specification as it is paywalled here, but wikipedia articles on Cat5, Cat5e, Cat6, TIA 568. None of them make reference to a phsyical specification, only the electric ones.

Well, they do, sort of (I've highlighted the 'sort of' bit)

4.3 Horizontal cable 4.3.1 General
Covered herein are the requirements for balanced twisted-pair cables used in horizontal cabling. The cable shall consist of 22 AWG to 24 AWG thermoplastic insulated solid conductors that are formed into four individually twisted-pairs and enclosed by a thermoplastic jacket. The cable shall meet all of the mechanical requirements of ANSI/ICEA S-80-576 applicable to four-pair inside wiring cable for plenum or general cabling within a building.
NOTE – Additional requirements for 100 Ω ScTP cables are located in annex K. 4.3.2 Applicability
Horizontal cables shall consist of four balanced twisted-pairs of minimum 24 AWG thermoplastic insulated solid conductors enclosed by a thermoplastic jacket. Bundled and hybrid cables may be used for horizontal cabling in accordance with clause 4.3.6. Four-pair horizontal cables containing conductor diameters larger than 24 AWG, up to and including 22 AWG, that meet or exceed the requirements of this Standard may also be used.
6
4.3.3 Mechanical
In addition to the applicable requirements of ANSI/ICEA S-90-661-1994, the physical design of horizontal cables shall meet the requirements of clauses 4.3.3.1 to 4.3.3.6.
4.3.3.1 Insulated conductor
The diameter of the insulated conductor shall be 1.22 mm (0.048 in) maximum.
4.3.3.2 Pair assembly
The cable shall be restricted to four twisted-pair conductors. The pair twist lengths shall be chosen to ensure compliance with the transmission requirements of this Standard.
4.3.3.3 Color codes
The color code shall be as shown in table 1.
Table 1 - Color codes for 4-pair horizontal cables
TIA/EIA -568-B . 2


Conductor identification


Color code


Abbreviation


Pair 1 Pair 2
Pair 3 Pair 4

White-Blue Blue
White-Orange Orange
White-Green Green
White-Brown Brown
(W-BL) (BL)
(W-O) (O)
(W-G) (G)
(W-BR) (BR)

The wire insulation is white and a colored marking is added for identification. For cables with tightly twisted-pairs [all pairs less than 38 mm (1.5 in) per twist] the mate conductor may serve as the marking for the white conductor. A white marking is optional.
4.3.3.4 Cable diameter
The diameter of the completed cable shall be less than 6.35 mm (0.25 in).
4.3.3.5 Breaking strength
The ultimate breaking strength of the cable, measured in accordance with ASTM D4565, shall be 400 N (90 lbf) minimum.
4.3.3.6 Bending radius
Twisted-pair cables shall withstand a bend radius of 25.4 mm (1 in) at a temperature of -20 °C ± 1 °C, without jacket or insulation cracking, when tested in accordance with ASTM D4565, Wire and Cable Bending Test. For certain applications (e.g., pre-cabling buildings in cold climate), the use of cables with a lower temperature bending performance of -30 °C ± 1 °C should be considered.

Taken from http://www.csd.uoc.gr/~hy435/materia...IA-568-B.2.pdf

But that refers to horizontal cable installations in commercial buildings. The TIA/EIA 568 standard is a suite of documents - not all are quickly found. So yes, a cable for a Cat5 installation in a commercial building where the cables are horizontal shall have a diameter <6.35 mm, which implies a round form factor.

And a round form factor for a properly constructed cable (which specifies the twist pitch in each twisted pair) is more likely to meet the Cat5 electrical standards over long lengths (and the standard does specify minimum lengths over which those specs shal be met) than other form factors - so in that sense, it is more likely that a round form factor will be compliant to the cat5 standard in its entirety.

(And note that only solid conductors meet the standard for installation within this standard, but stranded conductor cables, used for patching and drop cables will meet the electrical performance standards)

Nice thing about standards, so many to close from.

But to get back to the specifics of flat cables...

The flat cable referred to is pre-terminated, so for that length of cable it may meet cat5 electrical performance, while not meeting the cat5 mechanical specification. So in that sense, the cables are not Cat5 compliant, and flat cables would not meet the standard for installation in commercial buildings (which is not an issue here)

TIA/568-C.2 discusses cables in greater depth, but I haven't been able to find that online.

[jim]

Whoever was selling you flat 'Cat. 5e' was a charlatan. It's the pair twisting specifications which gives the cable its category rating, which is also what makes it necessarily thick and round. Flat ethernet cables are always category nil.

I see I've managed to spark a debate whilst I've been away, but with regard to your first question, Scan.

Fwiw, they still sell it: http://www.scan.co.uk/products/20m-c...r-under-carpet)

[aidanjt]

I see I've managed to spark a debate whilst I've been away, but with regard to your first question, Scan.

Fwiw, they still sell it: http://www.scan.co.uk/products/20m-c...r-under-carpet)

Naughty scan, doubly so since it's even untwisted flat.

[peterb]

I see I've managed to spark a debate whilst I've been away, but with regard to your first question, Scan.

Fwiw, they still sell it: http://www.scan.co.uk/products/20m-c...r-under-carpet)

I think aidnjt's point was that flat cable cannot be compliant with Cat5 standards because the standard specifies the form factor (i.e., round!) and so should not be sold as Cat5. (Although 'charlatan' is a bit OTT!) However, while he may be right, the only standard I have found refers to the mechanical specification for commercial building specification. That doesn't mean the standard for generic cat5 cable doesn't specify the mechanical form factor, just that no reference has been linked or quoted.

[MrJim]

Thanks for all the replies guys - all interesting information

As far as physically laying the cable is concerned, what would you say is the best way to keep it in place? I was thinking of just gaffer taping it to the floorboards, or are there any specialised clips for the job? Also, I'm not sure about how well it would route around corners; a straight run should be easy, but I can't see how the flat profile will bend?

[peterb]

Yes, tape would work.

The bend radius along the flat axis will be larges so you either need to put a half twist in it (not too tight) or route it carefully minimise bends, or use the minimum length necessary to go under the carpet and use cat5 couplers at each end to conventional round cable. Couplers do introduce a bit of loss, but probably not enough over a short run to make a difference.

Make sure the floor installer doesn't nail through the cable - and while the cable is flat, if it is going under the wooden floor, which is being laid on another flat surface, you will still need a recess. If the floor is being laid over an underfloor cavity, could you run standard cable in before the floor is laid? Flat cable is usually used under carpets.