What Ethernet cable are you using into you streamer?

In the simplest Ansuz cable it is just a LAN cable with Telegärtner contacts so nothing special about that one in the A and D series they use some other stuff so as you say not easy to dupplicate.

And the sheild is connected in both ends so if you dont know what you are doing you might end up connecting two different earthplanes, one in the audiosystem and one in the network and that might cause a earthloop.

I wonder if that is an older product as currently they only seem to offer the A and D series and another ethenet cable called the X2, however that too has the extra tech gizmos.

Yes the shields connected at both ends could be a problem with earth loops. Bearing in mind dCS’ strictures about using unshielded cables I wonder if this is because there may be no connection to ground from the network receptacle in their designs? If there was then presumably a cable with the shield connected at one end could yield benefits?

I do make a couple of recommendations back in the thread but I guess that these may be considered " normal". I have just ordered some even cheaper “normal” Cat 5e cables ( Belden Catsnake).

What are you connecting?

i need to connect the streamer naim nds to cisco switch 3590 and nas synology

@simon68 if I was making my own cables, I would do this…

https://youtu.be/7mIGpaFrMkI

… but not connect the shield at one end. Just wrap insulation tape at the end of the cable to make sure the shield does not touch the connector.

Hope that helps.

Paul

1 Like

https://youtu.be/7mIGpaFrMkI what cable is he using ?

I looked at the old X and the new X2 and they have what they call NSC and that is just a small ferrite the cable passes through right before the connector. It would work like some sort of common mode noise suppressor. I am a bit sceptic to all these gizmos as there most likely in the network equipment already is a common mode noise filter built in the RJ45 connector along the galvanic isolation.

1 Like

I recall a piece from Gordon Rankin of Wavlength audio ( competitor with dCS in respect of early asymmetric USB computer connections and the designer of the AQ Dragonfly DACs) saying that he had found data errors appearing when cables had ferrites attached. Some years ago now but hopefully there may be a link out there.

1 Like

I didn’t find the Gordon Rankin piece, but I didn’t look that hard either. :wink: I found a number of discussions on the topic, including this one.

@PAR
as much as dcs folks have advised on a spec cable.its still not clear the reason or rationale behind this.
what harm would shielding do to a digital/data signal on a ethernet cable. on the analog side shielding can be good or bad or different in various systems and conditions,.

Shielding is , if you like, an active component. It captures parasitic interference and then gets rid of it by draining it to earth. If the earth drain is not connected then the shield can add noise to the wanted signal through leakage. If the shield is connected at both ends and is thereby is connected to two paths to earth then if the two paths differ in electrical potential you can create an earth loop.

A successful shield depends upon the electrical environment within which it functions. It is the same whether the signal is digital or analogue.

2 Likes

This is not the original post that I referred to but Gordon Rankin’s comment can be found about digital errors using ferrites a couple of postings down:

1 Like

For the very simple reason that the Ethernet magnetics in the dCS products are unshielded.

I can only speculate as to why dCS chose so, but most likely because shielded Ethernet magnetics is significantly more complex from an EMI/ground-loop standpoint for zero benefit (and likely has detrimental effects) in a 2-channel audio system. Even DSD512 streaming only uses a faction of the capacity that the current dCS Ethernet ports are capable of.

A simple properly constructed CAT5e UTP cable will yield 100% bit-perfect streaming into the dCS system (easily proven objectively despite what some imagine they can hear). So, any Ethernet cable of a higher spec is purely to aid in one’s cognitive bias only (which no doubt can lead to actual perceptive distortions :laughing:).

1 Like

i asked for clarification regarding shields and drain from an audio company which makes ethernet cables and below is the response i got.

Only the CAT7 is shielded as the standard requires that. The CAT6 models are not.

The quote about the Earth loop is incorrect. Ground loops are circulating currents from differentials between numerous circuits. A parallel shield and drain are a low-impedance path. There are certainly reasons to make the shield and drain connect to one side only whenever possible, but this is not that reason…

The reason we typically drain only to source or load only (depending on the application), it to allow the induced noise drainage to move to the circuit that will suffer the least damage from said noise. However, the current standards for CAT7 do not allow for this where the shields are concerned. The primary conductors are all directionally tested and oriented to assure that we drain in the most advantageous direction for all connected circuits.

Please understand that writing in bold is shouting at me. I find this offensive.

I also mention earlier:

Anup makes a similar point but better ( as usual):

The comment from the audio company which makes ethernet cables:

This seems to be irrelevant if the connecting drains may not exist at the equipment. And have you read this far in the thread without noticing the comments that Cat.7 is not relevant for audio purposes?

@PAR
my apologies.never meant to even write bold.i just copy pasted something and it turned out bold.fixed it now :slight_smile:

Just because the magnetics inside a dCS is unsheilded (many switches and routers are built this way with the magnetics on the circuitboard) dont mean that a shielded cable dont get any connection in a dCS.

The RJ45 port on dCS still have a metalhousing and any cable with a sheild will connect to that and the metallhousing is connected to the case. Sometimes the housing is connected to the case with resistors/capacitors to form a way for RFI to drain but not form a low impedance way for a groundloop to form. I have a Network Bridge and the metalhousing on the RJ45 inlet have connection to the groundpin in the IEC.

The shield in a cable is often ment to be an extension of the cabinets of the two connected devices so interfirence cant get in for example a XLR connection where the shield connects to the case at each end.

With LAN cables there are other mechanisms in place to both galvanicly seperate each device and also commonmode noise filters (these transformers and filters are in the small magnetic devices placed either in the port directly or on the circuitboard on all networkports according to the IEEE standard.

There are places that need shielded LAN-cables but that is not in a regulare home and in places where this is needed you still have to pay attention how the shield is conncted as in factorys for example there can be large potential differences between places in the network and thus form a path in the network cable for current to flow in the shield.

2 Likes

Thanks. That is helpful. Sticking with UTP cables seems the preferable option at least in terms of simplicity (and in avoiding potential earth loops).

2 Likes

Precisely! :smiley:

In fact, it’s all the more reasons not to use shielded Ethernet cables as the last thing you want is stray signals making their way into the ground plane of the dCS via the RJ45 socket housing when the discrete magnetics are unshielded.

1 Like

Hi

Do you mean a unshielded twisted pair but it can still have foil F/UTP or a cable with no foil wrap and no shield U/UTP.

Just looked on the budget “audio grade” cable I have and it is shielded on each pair. I’m thinking of running the CAT 5 outside around the house 20m to the router to avoid any 240v parallel lines.

Thanks