expensive 75ohm/cat5/or 6

UTP cables such as Cat5  are noisy. Ideally they should not be - but either the twisitng is non perfect, or the shield in Cat6 etc picks up EMI currents due to the imperfecections and sinks the RF current to usually the source. For data networks they are fine upto and beyond 1Gbps, and typically upto 100m (by specification), but these imperfections will casue link errors over long distances at higher speeds.

For digital audio if you can decouple the RF noise it can help. The other big disadvantage for UTP is the switching energy required to modulate all 8 wires (for 1Gbps - 6 wires for 100Mbps). This can cancause noise in the receiver through EMI and the RF power density that is being created.

Now SPDIF into 75ohm has only one conductor modulated, so the overal power density is reduced. If the cable is a good quality radio cable such as RG59 (a few pence a metre) and properly terminated into 75ohm connectors, then if you have quality SPDIF transcievers (that is quite a big if) then indeed a cheap deadly accurate coax cable can be achieved. The drawback is that RG59 into BNC plugs is going to only cost a couple of quid at most to make so no ne is very insentivised to sell it - but you can make it very effectively. Also it doesn't tend to have sexy looking sheath coverings or glitzy termniation.

If you have the correct radio engineering equipment you  can also check the SWR of the lead into 75ohm dummy load at the various sample frequencies you are interested in.  You should see a SWR of 1:1 across the range. This means the lead is practically perfect and is causing no reflections as a consequence of characterisitc impedance imbalances. 

yes but the manfacturing of these cables is unknown and previously not tested in this audio framework .if you disect one the can be poorly constructed

Which cable are you talking about? Mil spec RG59 / RG59U ? If so the characteresitic and construction is defined as part of the standard, so purchase from a reputable certified cable manufacturer which is what I do, as opposed to a no name importer. BTW RG59 and RG 6 are used in the AV industry.

If the cable is not manufaturered correctly it becomes a bit of a car crash for HF radio and higher, first tell tale sign is leaking, the next is severe RF signal attenation, and reflections (ie bad SWR). Alas poor cable design is less easy to detect and pinpoint for the average consumer audio setup with SPDIF etc, especially if covered in bling.

Also poor cable changes its characterisitc impedance when bent or over time.

Simon

Simon,

Do you mean to imply that a suitably motivated individual in the hobby might be able to create RG6 IC's that perform well?  If yes, I need a shopping list!  Sounds like inexpensive fun at least.

Nick

Nick

Exactly!! And thats the spirit - and the very least it is fun and at best is  rewarding...

So - first of all I would practice soldering and pressure fitting BNC connectors to some inexpensive RG59. Do it a few time so you are comfortable, you might need a larger solderng iron.

Then buy the following

1) 2x BNC connectors at 75 ohm characterisitc impedance

2) 1 metre of RG6U or RG6UQ - the latter has quad shielding which has reduced leakage.

3) 4x Ferrite Clamps to go over cable (between 6.86mm and 7.57mm outside diamter) - position 3 at digital source end and 1 at sink. The numbers of chokes is not exact..

Get a multimeter and check for no shorts between centre pin and shield.

If ok attach and listen.

Optional bit if you are festidious:

If you want to check characterisitc impedance - get a 75ohm RF dummy load, and a SWR meter. The meter can be quite expensive and so you might want to borrow from a licenced radio engineer/operator - if you know one, just ensure its 75ohm - a lot of HF/VHF/UHF radio stuff is at CI of 50ohm. Sweep between 8MHz and 30MHz and check for 1:1.

Assuming you have no reflections from the transceivers you might be able to 'tune the sound' of the lead by adding or removing the ferrite clamps. I believe a lot of the so called ' differences ' in SPDIF cables assuming htey are correctly designed to be 75ohm CI from connector/cable/connector is down to the transformer effect between induced earth currents in the shield and the centre conductor and the magentic permeability of the shield or shields. You can impede the RF earth currents by varying extent by adding or removing inductance. You then start to get a feel on how to make the leads give a aparticular effect or sonic signature....  or remove one...

Let us know how you get on!!

Simon

On the other hand just take care selecting the right cable - then you don't need to go to all the hassle of measurIng it to make sure 

James - you are right, but the SWR also confirms you have correctly terminated the cable. If your termination is bad - ie how the shield is linked to the BNC - the characterisitc impedance may change, but be fne at DC and read zero ohms on a multimeter, but as I said only for the festidious.

Very true Simon, I didn't want to put Nick off from rolling his own cables. 

James

Thank you, Simon and James,

The key here looks to be I can use the same crimp tool for BNC, RCA, and F-type terminations.  Liberty appears to be a well-regarded brand.  

Were you joking about the expense of the SWR meter?  My superficial search finds many from $9-15 and only one > $100.  What should one look for?

Nick

Hi Nick, a cheap but useable entry level SWR analyzer is the MFJ259B. It's about $300 or £250 and aimed at the enthusiast engineer. It's designed for 50 ohm impedances, but 75 ohm will read 1:5 on 50meter bridge. The MFJ can also be used to directly read impedances across frequencies. This can also be used to measure how well designed the input to a DAC's SPDIF input is, and whether it rather than the lead is causing reflections. 

I find once you have some  basic test/measurement equipment a lot of the voodoo disappears and you can see what is happening.

Hi Nick, a cheap but useable entry level SWR analyzer is the MFJ259B. It's about $300 or £250

At this price why not just buy a well known well reviewed cable (e.g DC1) and be done with it. Naim engineering and expertise is worth paying for even if it is arguably a bit over the top price wise.

Graham, you can totally do this if you want, fine, the original query was how to explore sone of these aspects as part of the hobby. Some  of us are professional engineers as well as amateur musicians(as well enjoying recorded music). Being an engineer its in your blood not to take things for granted and exolore how things work.  However if you don't get that urge to challange and explore these things, the chances are you are not an engineer - there is nothing wrong with that, we are all different, but either way there is abolustely nothing wrong buying an off the shelf stock product if you want to.

Further I did recommend borrowing an analyzer  if you didn't have it.

Simon

The problem with the DC1, apart from cost, is the length. As Simon's posted it's a doddle to make up your own interconnect using good-quality co-ax cable & either BNC or 75-ohm RCAs & you can have the length you want, in my case 11 metres.

Hi Simon and Tony

Having soldered up speaker cables and interconnects in the past I can see the point you are making, especially when it comes to bespoke cable length but I didn't need any expensive equipment. I have long since chucked my cheap Ham Radio SWR meter away.

If you have the technical know how and access to the necessary equipment then that's great. Good luck and enjoy it.

I go fly fishing and see the attraction in catching trout on home tied flies but flytying is not for me. Each to their own. At least you are not saying that digital cables make no difference and can just be cobbled together.

For me, the wish to use BNC pushes my interest in giving this a try.  Retail BNC cables are not that common, on the shelf.  Being able to solve my own problems immediately is worth the small investment in bulk cable, even if I consider homemade cables a baseline.  I already make my own CAT6 cables.  I wouldn't mind any tips on those, too.  As Tony says, getting the length right is an important benefit.

Even if I do not get the meter, this could be a useful option for me.

Thanks all.

Nick

You might be interested to know SPDIF is only specc'd upto 10 metres in length. I am not saying  anything longer won't work, cleary is does for you, but depending on SPDIF transceivers you might start to get some funnies with some equipment as you increase above 10m.

Simon

Having been warned in the past that long digital cables would adversely affect sound quality I did a few experiments around this. Conclusion? a 16 metre digital co-ax sounds the same to me as a 1.5 Metre one and long optical cables also don't sound any different to me either. Perhaps I've got tin ears but then others have noticed the same in the context of my system. It's only my ears that matter though!

They should be readily available - 75ohm BNC cables are sold for video use ...

Cheers

Phil

All this talk of S/PDIF and BNC is fine, but wasn't Rich46's original point that long, cheap Ethernet cables going from a streamer to a switch, and from there connecting to a NAS, a better option than setups that require the use of shorter, more expensive cables like S/PDIF, USB, Firewire, Thunderbolt, etc.?  

If yes, then I agree.

Simon has pointed out that a long Ethernet cable can act like an antenna, but that effect also seems pretty easily squelched using chokes at the streamer's RJ45 input.    I think the biggest advantage of using Ethernet is reduced jitter, and this was summarized by Steve Nugent of Empirical Audio as follows:

"...Networked audio (Ethernet), both wired and WiFi is a unique case. Because the data is transmitted in packets with flow-control, re-try for errors and buffering at the end-point device, it is not as much of a real-time transfer as USB, S/PDIF or Firewire. The computer transmitting the data packets must still keep-up" the pace to prevent dropouts from occurring, but the real-time nature of the transfer is looser. Unlike with other protocols, there can be dead-times when no data is being transferred. Networking also avoids the use of the audio stack of the computer audio system since it treats all data essentially the same. This avoids kmixer on XP systems and the audio stacks on Mac and PC Vista. Because of the packet-transfer protocol of Ethernet and data buffering at the end-point, the jitter of the clock in the computer is a non-issue. The only clock that is important is the one in the end-point device. Examples of end-point devices are: Squeezebox, Duet and Sonos. This would seem to be the ideal situation, which it certainly is. The only problem that can occur is overloading the network with traffic or WiFi interference, which may cause occasional dropouts. The problem for audiophiles is that the majority of these end-point devices were designed with high-volume manufacturing and low-cost as requirements, with performance taking a lower priority. As a result, the jitter from these devices is higher than it could be. It should be the lowest of all the audio source devices available...."

Interesting perspective for someone who makes his money designing USB devices!  Also a great argument for moving to a stand-alone device like an NDX or NDS, perhaps with an external PS, but *not* with an external dac!  Just an RJ45 ethernet network port for input, and a DIN (or RCA) for analog output.

Hook

Hook

A few mixed concepts here from Mr Nugent, or perhaps over simplifications:

1. Trasnport jitter and sample jitter are two seperate things, I think this has been discussed elsewhere on this forum

2. TCP/IP has two transport modes. TCP and UDP.

3. TCP is a connection oriented protocol with varying windowing or flow control. Typically for audio its timing variation will be slow to the sample data. So its transport jitter would fall into the Wander category - which can be buffered and reclocked, and importantly the source flow is confirmed by the sink.

4. UDP is a connectionaless protocol and effectively sends precise bundles of samples with NO flow control. This is typically used for low latency media, such as IP telephony, true web broadcasting and somtimes video. Jitter here is critical and is carefully engineered in network design to fall into agreed tolerences. This is what the term QoS typically refers to. BTW multicast uses UDP, in this case for device sync rather than media in uPNP.

Therefore I really think its a moot argument trying to compare SPDIF, ethernet or USB from a trasnport jitter point of view, they all exhibit it - but it is decoupled from the sample jitter through buffering and reclocking which Mr Nugent states and I agree with (in the detail there are interdependencies - but again that is usually through electromagentic perturbations etc and again that has been covered elsewhere) - with the one exception that professional SPDIF ( AES) can be specified to act as a master clock, so the transport jitter would become the sample data jitter. But this is quite rare now I believe.

Simon

Simon!

Thank you!  I have been on the sidelines too long re custom cables.  I went to my local wire shop today and dropped $200 on 30' each of RG6 and RG6-Quad cable and an assortment of terminators and strippers and crimp tools.  (Sad to think for a moment of what a search on "strippers"  returns on The Forum. . . )  I slapped together an RG6 with RCA to BNC and put it between my Squeezebox Touch and bare nDAC:  nothing.  Re-made the BNC end:  M A G I C !   Best SQ yet from anything through the nDAC.  Early days yet, of course.  Better than cheap optical from Mac Mini; better than Audioquest RCA to RCA from SBT.  Maybe it is the BNC?  If the DC-1 is better, great!  We'll see when I get one.  I did this in part because my dealer did not have a DC-1 in stock for me today, while wife and child are away.

The shop guys looked at me oddly when I said I was interested in an SWR meter that measured RF leaks.  They knew it was a valid issue, but said it is an infrequent request.

Nick

Nick, great stuff! Did your 'magic' cable use regular RG6 or the quad shield variant?

Try the chokes on the lead, and see if you hear any differences...

BTW the SWR meter measures reflections caused by characteristic impedance mismatches, a SWR bridge allows you to measure impedance and SWR for adjustable frequencies , for RF leakage you would use a sensitive field strength meter or even a SW radio receiver with an S meter if pushed.

The only piece of test-equipment i really would definitely recommend if your are making leads is a multimeter to check for DC shorts and open circuits, and a basic one is only a few dollars. ;-)

Simon

Hi, Simon,

Magic was the plain RG6.  I've now made one of each, but it is too late to test today.  I did buy two chokes to test.  One is in a plastic retainer, the other is a large  ring.  Both are too big to fit to the RG6 without some effort.  I suppose  I might tape them in place, as a start. . .?

Anyway, very gratifying results for first tries.

Thank you!

Nick

The plastic retainer one sounds like the  variant to use,  assuming its only slightly too large you could use BluTak. The ring sounds like a ferrite ring. I use ferrite rings on my cartridge interconnects on my TT to stop radio broadcast break through. You need to carefully wind spaced turns across each side of the ring, best look up on google to see how it's done. You would use for this relatively severe RF current attenuation, but there are other engineering considerations that tend to come into phere here... For RG6 and similar for SPDIF, Ethernet and mains the clamp on variants usually work best IME as we are usually talking low level RF currents.

You are welcome and good luck!!

For the benefit of other readers, the results I am getting from my beginner's-try at making my own digital cables are so good I would hesitate to ask my dealer to order me a DC-1.  I am not sure how much better it could be or needs to be.  I will report when I compare the two.  (Maybe this is why they don't stock them?)

I realize not everybody wants to do this; but if you are thinking about it, go for it.

Regards,

Nick