Advice on shielded cat 5e installation

[Speedster]

Hi guys. Our church recently got a new sound mixing desk. One of its features is that it uses a LAN cable to run signal from the stage to the mixing desk. We use a local school hall so have to set it all up each week. There is currently a UTP cat 5e cable running from stage to the room where the mixing desk is, but this cable doesn't allow the device on stage to sync with the mixing desk. I've read up about the specs for the desk and they clearly require shielded cable (although unshielded works over short runs). I popped over to a local cable shop and bought 50m for us to install. My plan is to run the cable through the roof and down behind the stage, terminating in a standard Cat5e wall mount as the cable shop said there is no special wall mount for a shielded cable. My question is what happens to the ground when one terminates the cable in a wall mount? How important is the ground to the integrity of the connection (given that the device does work on UTP over short distances)?

 

[Gnome]

For a device to actually be compliant it needs to allow UTP (unshielded twisted pair). So either the mixing desk sucks or your LAN cable is damaged, or more commonly, the end terminations are bad.

Anyway since you already just went ahead and bought STP cable may as well work with what you have.

This should do as a primer:
https://www.lanshack.com/cat6a.aspx

Where the cable goes to the switch run a ground wire from a plug onto the switch/router housing (to ground it).

the cable shop said there is no special wall mount for a shielded cable.

The STP jacks I have are made of metal and are different and the information online says otherwise. But everywhere you break the ground you need to connect it to a ground conductor.

There is very conclusive research paper showing ground loops are not a source of interference for networking equipment so multiple ground points are not supposed to be a problem.

 

[chrisc]

Just ground the shield at one side only. You need to determine which side is better

 

[Speedster]

For a device to actually be compliant it needs to allow UTP (unshielded twisted pair). So either the mixing desk sucks or your LAN cable is damaged, or more commonly, the end terminations are bad.

Anyway since you already just went ahead and bought STP cable may as well work with what you have.

This should do:
https://www.lanshack.com/cat6a.aspx

The manufacturer's spec quite clearly states it should be shielded cable (note that although they use standard LAN hardware, it is a different protocol called AES50). I've seen on a number of audio forums (through google search) other users have had the same problem which was resolved by using shielded cable.

 

[Speedster]

Just ground the shield at one side only. You need to determine which side is better

My question is, how would I ground it? Cat5e wall mounts don't AFAIK have a grounding option.

 

[Gnome]

The manufacturer's spec quite clearly states it should be shielded cable (note that although they use standard LAN hardware, it is a different protocol called AES50). I've seen on a number of audio forums (through google search) other users have had the same problem which was resolved by using shielded cable.

Ouch. (btw I added a bit more information to my post)

Doesn't really make sense to me, that is the whole point of using a digital communication network standard (ie> ethernet) but some people obviously suck at engineering and took shortcuts ¯\_(ツ)_/¯

 

[Speedster]

Ouch. (btw I added a bit more information to my post)

Doesn't really make sense to me, that is the whole point of using a digital communication network standard (ie> ethernet) but some people obviously suck at engineering and took shortcuts ¯\_(ツ)_/¯

I'm not sure shortcut would be the correct description. It's an industry standard protocol designed to run over the same hardware (ie cat5e / 6 cable) but carries uncompressed individual sound channels, rather than data (as per standard ethernet).

 

[Gnome]

I'm not sure shortcut would be the correct description. It's an industry standard protocol designed to run over the same hardware (ie cat5e / 6 cable) but carries uncompressed individual sound channels, rather than data (as per standard ethernet).

(very much TL;DR post). I thank you for interesting journey of audio standards I just had

Just read through the standard (having implemented a TCP/IP stack in C for Arduino all the way down the ethernet frames I got curious). I must say there doesn't seem to be any reason AES50 cannot work on standard cable. It is down to the individual chip used for AES50 if it is required. (so lower quality chips I guess will struggle more).

The lowest level component of a network is called the MAC (media access control). It translates data into electrical signals that devices on the network can understand.

Ethernet has something called a ethernet frame which looks like this: https://en.wikipedia.org/wiki/Ethernet_frame#Structure

In this case AES50 changes the out the frame payload. The preamble is still used to synchronize, so there is really no reason to force STP.
AES50 is used for low latency.

There are standards like AES51 that uses standard ethernet but carries the data on the ethernet frame. So it would be possible to implement using standard ethernet components.

I'm very curious what sort of latency implications it would have to use a standard ethernet frame VS layer 1 implementation like AES50. Latency is really the only advantage AES50 could have, because in every other way it is just ethernet. (and actually the stated goal of it is low latency).

Interesting read tho.

EDIT: I should add, AES50 chips are "specialized" and thus almost certainly not up to the same standard as Ethernet. Another case of re-inventing the wheel IMO that lead to an inferior product due to lack of competition. Gigabit ethernet MACs are cheap as chips these days, if they used standard ethernet they could have been in that world now.

 

[Speedster]

(very much TL;DR post). I thank you for interesting journey of audio standards I just had

Just read through the standard (having implemented a TCP/IP stack in C for Arduino all the way down the ethernet frames I got curious). I must say there doesn't seem to be any reason AES50 cannot work on standard cable. It is down to the individual chip used for AES50 if it is required. (so lower quality chips I guess will struggle more).

The lowest level component of a network is called the MAC (media access control). It translates data into electrical signals that devices on the network can understand.

Ethernet has something called a ethernet frame which looks like this: https://en.wikipedia.org/wiki/Ethernet_frame#Structure

In this case AES50 changes the out the frame payload. The preamble is still used to synchronize, so there is really no reason to force STP.
AES50 is used for low latency.

There are standards like AES51 that uses standard ethernet but carries the data on the ethernet frame. So it would be possible to implement using standard ethernet components.

I'm very curious what sort of latency implications it would have to use a standard ethernet frame VS layer 1 implementation like AES50. Latency is really the only advantage AES50 could have, because in every other way it is just ethernet. (and actually the stated goal of it is low latency).

Interesting read tho.

Thanks, that helps. From my (very uneducated) reading, the requirement has to with some interference UTP picks up over the length of spans these cables run. The original spec from the manufacter for our desk (Behringer) didn't require STP, but that's since been updated due to the number of user picking up errors.

Back to my issue - what are the potential problems of terminating the STP cable in standard, plastic, Cat5e wall units? Is there some sort of charge that builds up on the shield if it isn't grounded, or why is the grounding necessary?

EDIT: When testing with a 5m UTP cable, it all worked fine. However, when we plugged into the approx 40m cable run through the roof the units wouldn't establish a handshake.

 

[sajunky]

Shielded cables gives more problems if installed incorrectly. I will cover it later wwhen you pass the first sync tests.

AES50 is not using standard Ethernet MAC frames, therefore you can't use Ethernet switches. It is end-to-end connection using Ethernet Cat5 cable with a proper pin-out termination according to the documentation. You should not expect presence of automatic MDI/MDI-X switching. I expect it will be some crossover pin-out, it is why a standard straight T568A or T568B cable do not work.
[EDIT] I read a short cable is working. Is it straight or crossover cable?

For this reason I suggest to check documentation first, then [-]test the equipment using a short UTP cable[/-].

 

[Speedster]

Shielded cables gives more problems if installed incorrectly. I will cover it later wwhen you pass the first sync tests.

AES50 is not using standard Ethernet MAC frames, therefore you can't use Ethernet switches. It is end-to-end connection using Ethernet Cat5 cable with a proper pin-out termination according to the documentation. You should not expect presence of automatic MDI/MDI-X switching. I expect it will be some crossover pin-out, it is why a standard straight T568A or T568B cable do not work.
[EDIT] I read a short cable is working. Is it straight or crossover cable?

For this reason I suggest to check documentation first, then test the equipment using a short UTP cable.

As mentioned above, I've tested with a shortish UTP cable with no errors. It's the longer span through the roof that doesn't work. I'm not using any switches, just Cat 5e cable, wall boxes (like this) and RJ-45 connectors.

 

[upup]

Can you perhaps change any speeds over the cable. You can take the screen part, with a little wire to your equipment's body. Does that product not state the maximum length of your cable or a type of cable.

Did you test the cable in the roof, with maybe a laptop and a router.

 

[sajunky]

Adding information about shielded cables. I don't know what cable you already bought. It will be probably 'installation' cable (meaning made of solid core), this one is good. - confirm.

Gounding must be done end-to-end, unless it is specified otherwise. It means all components must carry a ground, including wall boxes. If you don't have shielded wall boxes, connect the cable directly to the equipment. Lack of the ground usually have devastating effect.

 

[sajunky]

As mentioned above, I've tested with a shortish UTP cable with no errors. It's the longer span through the roof that doesn't work. I'm not using any switches, just Cat 5e cable, wall boxes (like this) and RJ-45 connectors.

You didn't check it is straight or crossove cable and which one is required.

The wall box is not good for STP. Connect both cable ends directly to the equipment.

 

[Speedster]

Can you perhaps change any speeds over the cable. You can take the screen part, with a little wire to your equipment's body. Does that product not state the maximum length of your cable or a type of cable.

Did you test the cable in the roof, with maybe a laptop and a router.

The cable in the roof works for standard ethernet functions, as well as for the previous generation audio transfer. The product states a maximum length of 100m and "recommend using a shielded CAT 5e cable with Ethercon connectors wired to the T568B standard. Unlike other digital protocols, AES50 (SuperMac) uses all the pins within the cable making cabling quite important."

Ethercon is a ruggedized and lockable RJ45 connector system, optimized for pro audio, video and lighting network applications. It is fully compatible with RJ45 (i.e. one can use a standard RJ45 too, although the Ethercon won't fit into a standard RJ45 plug).

 

[Speedster]

You didn't check it is straight or crossove cable and which one is required.

The wall box is not good for STP. Connect both cable ends directly to the equipment.

Straight, not crossover. What would be the effect of not grounding the shield? Since UTP works over short runs, the ground obviously isn't needed for the devices to work, but only to minimise interference on the cabling. Would not grounding the cable nullify the effect of the shield?

 

[sajunky]

Straight, not crossover. What would be the effect of not grounding the shield? Since UTP works over short runs, the ground obviously isn't needed for the devices to work, but only to minimise interference on the cabling. Would not grounding the cable nullify the effect of the shield?

Read, I wrote: devastating effect, trust me.
Broken shield in STP cable is worse than no shield at all.

 

[Speedster]

...snip...
EDIT: I should add, AES50 chips are "specialized" and thus almost certainly not up to the same standard as Ethernet. Another case of re-inventing the wheel IMO that lead to an inferior product due to lack of competition. Gigabit ethernet MACs are cheap as chips these days, if they used standard ethernet they could have been in that world now.

I've been reading up a bit as well the last hour or so. Found this very insightful, particularly regarding the issues with standard ethernet in these live audio applications. Remember that you're working on a close to instantaneous delivery of the audio, there is no time to process anything. In other words, if a drummer strikes a drum, he (and the rest of the band) needs to hear the audio instantly otherwise the band will be playing out of sync and that'll just be really awkward.

Or, as mentioned in the link above,

The delay that a vocalist experiences between singing a note and hearing it in their earpieces cannot be perceptibly greater than the time it takes for the sound to be
physically transmitted from their mouth to their ears

 

[Speedster]

Read, I wrote: devastating effect, trust me.
Broken shield in STP cable is worse than no shield at all.

Please explain?

 

[sajunky]

Please explain?

Well, it is technical. A shield adds electrical capacitance to the wires - it is a large conducting surface in a close proximity to the wires. It gives negative effect on the signal (signal loss), but grounding isolate wires from the external noise, so shielding generally improve signal in the final effect.

Ungrounded (floating) shield acts as an antenna, in both directions. It means that noise goes in, noise goes out as usual in UTP, but it is multipled by the antenna gain (in simple terms). So, you have a bigger noise introduced to the wires, combined with higher signal loss by the capacitance of the shield.