Should I buy XLR cables?

Now you've mixed in differential topology with balanced interconnect. Those are often confused.

Floating ground from a chassis is hazardous and a terrible idea. Never, ever should safety be sacrificed for the idea of perhaps greater fidelity. ARGH!

CJ

 

If you need to ask your new here, XLR.

 

I was referring to gear designed with a floating ground (like 85% of Japanese amps), not the use of a "cheater plug."

 

I know that XLR has some advantages in certain situations, but I don't know if it will make any difference on 1 meter long cables, and I already have really good coax (Blue Jeans LC-1) from my previous system.

 

Interesting. It seems you've already made up your mind. A 3' Blue Jeans Belden 1800F single cable is $28. Not a huge expense to satisfy your curiosity .

Balanced Audio Cables at Blue Jeans Cable

 

The problem with unbalanced is this. When two pieces of audio equipment, each connected to mains power, are connected via an unbalanced cable, a small current will flow at the power line frequency and harmonics in the shield. Because the shield has electrical resistance that current generates a voltage, which is added directly to the signal.

This is the source of a "hum loop".

Now with a simple system - two pieces of equipment, plugged into the same power strip, and a short cable run with a low resistance shield, and line level, there is probably no problem.

But as soon as you have a bunch of gear - preamp with several inputs, power amp - or worse, home theatre with a record deck too - all hum and buzz bets are off.

The key thing with unbalanced is that the cable shield is part of the signal connection.

Lets look at it in terms of resolution. Modern digital sources have a dynamic range of 120dB (20 bit). Which means that system noise has to be lower than 1uV to enjoy the full dynamic range of your source. Now, measured in very careful laboratory conditions, an unbalanced connection on a single piece of audio gear can certainly achieve that, and only comprises hiss, but no hum elements. But as part of an overall system - not a snowball's chance. All bets are off on the hum and buzz front.

The beauty with a balanced connection is that it is immune to hum loops. That is because hum currents flowing in the shield induce the same voltage in both signal wires. And the balanced input stage, which is only sensitive to the difference between the two voltages subtracts any hum currents. The key number is called Common Mode Rejection Ratio (CMRR) which describes the ability of a balanced input stage to reject hum currents in the shield.

The devil is in the detail of course - particularly how the cable shield on pin 1 of an XLR is connected to chassis. Known as the "pin 1 problem" which you can read all about by googling the term. And there is an ongoing debate about whether the shield should be connected at both ends, or one end only, but that mainly impacts RF issues rather than hum and buzz.

I run balanced throughout. Including my phono stage. And it really is plug and play. Quiet as the grave, even from the record deck. Hiss only with an ear pressed to the speakers.

And that is why in professional applications - recording studios, rock concert sound reinforcement etc, balanced is always used. And it has nothing to do with cabling length - it is all to do with being able to plug a bunch of gear together and not worrying about hum and buzz.

 

So you're saying that without xlr and with a complete system plugged in, that you'll always have hum and buzz ? Then why am I not hearing it ? (I'm still playing around with xlr and rca between amp/preamp, but hum and buzz is not what I'm hearing)

 

This doesn't affect me because both my amp and dac are powered by ungrounded power bricks, but why do you guys think that lifting earth ground on some, NOT ALL components would be dangerous? In case of electrical fault on one of the components, both XLR and RCA shields should drain enough current to grounded component to trip ground fault safety device.

 

@Steve Hoffman answered this question himself in this thread

RCA vs XLR findings/questions...

RCA vs XLR findings/questions...

 

Yeesh. OK - lets say you have one piece of gear safety grounded and have lifted the ground of everything else, relying on the interconnect cables to act as a safety ground.

Then a fault occurs - a transformer short to ground, an insulation fault - whatever. These do happen. Now such a fault can cause instantaneous currents in the hundreds of amps range for the second or so that it takes the main panel breaker to trip. And that current will be carried through the RCA connectors and the cable shield. The cable will ignite, the RCA connectors will burn or weld to the chassis socket, and the internal wiring will burn through. And the chassis of the faulting equipment will go live until the breaker trips.

To avoid any doubt - under no circumstances disconnect safety grounds (or use cheater plugs) on equipment to get around hum and buzz problems. It could kill you or burn your house down. Or both. To even suggest this is both dumb and dangerous. And if someone does that based on your advice and something nasty happens, you will be personally liable and get your ass sued. And the judge won't give a rats ass that it cured the buzz.

 

No - I'm not saying that. All I'm saying is that wiring a system up entirely with single ended is a lottery. You might be OK, but you might not, in ways that run the gamut from hearing a buzz if you put your ear to the speaker, to an all pervading nuisance that spoils your listening.

But here is the thing - the system noise should be just a hiss. If it isn't and buzz dominates, that is undesirable.

As far as hiss is concerned, single ended actually wins. With a properly designed input stage you can get to -120dBu. A naively designed balanced input stage will manage around 95dBu - 25dB noisier than single ended. The THAT balanced chips manage -105dBu, but have traded the noise for superb CMRR. You can design a really low noise balanced input stage which delivers -120dBu, but it is not straightforward.

And there is absolutely nothing wrong with using a transformer to do balanced (or single-ended) isolation, so you break hum loops without the stoopid idea of floating chassis safety ground. Jensen make a whole load of transformer isolation stuff for precisely this use.

And mixing desks, through which all your music passes during the recording process often have have hundreds of transformers doing precisely that - giving galvanic isolation for the dozens of inputs and outputs.

 

1st that was hypothetical question, I'm not giving anyone advice.

I live in EU so electrical safety standards might be different from the rest of the world, but here earth ground is used instead of electrical ground to prevent precisely what you described.
Earth ground has higher resistance than electrical ground and it can't sink a lot of the current. Ground fault device isn't the same thing as circuit breaker, ground fault device will trip if it detects just a few milliamps difference between live and neutral (current flowing to the ground).
But like I said I live in EU were nominal voltage is 230v, so safety standards might be higher than places that use 100-120v systems. if 230v live created a short when it come in contact with grounded chassis it wouldn't be good for anyone.

 

Well, I'm from the UK, which is at least for the next couple of months in the EU. And although the trip time at rated LN imbalance current of 30mA is 200ms I still would not rely on the feeble shield of signal cable to protect me. Safety ground is for precisely that purpose.

All wiring standards I know of connect neutral and safety ground at the point of power entry - in the UK at the meter tails. It is that connection that is essential for electrical safety during a live/chassis ground fault, and has to be of a gauge that can carry a full fault current while the RCB trips. An RCA shield and connectors do not fulfill that purpose - it is a totally uncontrolled connection. It is even worse with a balanced connection, because a cable might have its shield connected one end only, or the equipment might connect the balanced shield via a capacitor or capacitor/resistor network. Or back to back diodes.

Any of those scenarios breaks all electrical codes and has the potential for significant harm to the user and his equipment.

 

Yes. You may not notice a difference, but the design reduces noise, so on principle you should go that way if you can

 

XLR connectors, which I generally like, because for most of my life, I have fairly well disliked RCA connectors, and would have thought that they would have gone out of favor long before now.

In pro-sound applications, XLR connectors are used because they lock into place, making a secure connection and they have a better rejection of noise when you are running a long wire.

In pro-sound, the amplifier is placed close to the speaker and the long wire run is the interconnect, which is why superior noise rejection is important.

If this was important in a short interconnect, asy a meter or two, then all RCA type connections in high end audio, would simply cease to exist.

This is the whole matter in a single sentence. IF.

Most designs that offer XLR connectors and RCA connectors simply have both sets of connectors wired in parallel.

This is NOT the same as if BOTH devices have truly dual differential circuitry. Which is really what XLR connections are designed for in the first place.

Also, in the pro-audio world, these XLR interconnects are made out of plain old ordinary copper wire.

 

Exactly, in many designs, the wiring of the XLR jacks is incorrect for eliminating ground loops - and the jacks merely exist for the sake of sales/convenience. This is what some don't realize. They read a couple threads and watch a couple videos and suddenly they're experts. I've gotten 60 cycle hum out of amps that were hooked to nothing but the power outlet and speakers.

The other issue with pro Audio is there's all sorts of randomness encountered because of different venues, performer's gear, etc. XLR just reduces the likelihood of noise. In a domestic setting, the noise floor differences are most likely to be completely inaudible. The only time I've had noise issues with RCA cables, the cause was connection integrity - cold solder joints and such. IME, a quiet hi-fi system starts with a quality amp and how well it jives with the house mains. Get that right and everything else is cake.

 

I get why would you use balanced interconnects, especially with microphones (phantom power, less vibration transfer to the mic, noise rejection, more flexible cables, and so on...), and I get why would you use balanced power amps (simple way to eliminate distortion on signal ground). But the idea of "fully balanced" line level components like DACs or preamps seems like absurd waste of money to me, in the same way that directional speaker cables, and interconnects with battery in them do. But I might be completely wrong on this.

 

I have to go from balanced to RCA's in my triamp system except the Crown amp running the low freq boxes. On the RCA end of IC's I have to tie the ground and - wires together. And, my e-crossover has the cheesy opamp quasi balanced circuitry as well. Not optimal. Also I run around 11 ft line level patch cords to my F5 Turbo amp running the high efficiency BMS compression drivers (113 dB/watt).

System is quiet as a mouse. I'm sure this varies by conditions but I surely could get by without XLR's here, it's not needed for noise control. OP can experiment but either way will work IME.

 

i recently switched from single ended to balanced cables for the run from my DAC to my preamp. Since I live in the city fairly close to the city center in fact... i decided i needed the shielding... as others have noted here XLR's run a few db hotter than single ended. If you live somewhere where there's a lot a rf and emi noise (ie a busy city) then i would say yes...

The other consideration is if your gear is actually balanced or if they are just using XLR connections on a single ended run. In other words both my PREAMP and DAC have true balanced connections... if you're running from a balanced connection into one that's single ended might as well just stick with RCA's.

 

The "has to be balanced internally" to matter is just plain ol' Internet lore. Over and over, like in this thread, it's said it has to be "fully balanced end to end" to matter. This is BS!

Interconnection and how the circuit topology is handled internally are seperate issues. With a lack of understanding of how noise rejection and common mode noise rejection occurs, I can see how this story line developed and perpetuates. But it just ain't true!

For one thing, to pull off internally a differential circuit is difficult and expensive. Stereophile at one time had an article talking about this.

Internet lore holds that you have to have the two sides, to the wiring AND the electronics, end to end or somehow you've perpetrated fraud. Actually, that's fraud!

CJ

 

My gear (RME ADI-2 DAC & THX AAA 789) has real xlr. But even though they are right next to plasma tv I don't get any audible noise with 1 meter long Blue jeans LC-1.

 

I think that @ghostofzuul is referring to gear that is treating xlr like single ended connetor (only hot phase and shield are connected, and cold phase is left floating)

 

Sorry but you're just plain wrong here. How the amp/components are wired internally does matter. Some manufacturers simply put XLR jacks on what is really a single-ended output/input. Anyhow, real ground loops are pretty rare in domestic settings unless a ground is poorly terminated. I've come across quite a few RCA cables with cold solder joints over the years. The most common causes of noise are garbage transformers and poor internal circuit layouts.

 

Just by using common sense, I'll have to respectfully disagree, I think that @CoolJazz is right on the money here. But than again my background is in digital electronic, so I might be completely wrong here. Besides this thread is about interconnects, not component topology. And XLR is just a method for getting audio from one place to another, it has nothing to do with topology of components it's connecting, and XLR existed before "full balanced" was a thing. As a matter of fact unshielded balanced was used before coax was invented for transporting audio and telephones.

 

@Helom you edited your post after I quoted you.
What @CoolJazz is referring to are components with fully differential topology, not "fake xlr"