Comparing Perf: Original XLR v. RCA- XLR adapter

Al... I was afraid that would be the answer. Thanks for your help.

Raks -- No, I would not recommend doing that. Your xlr digital cable is most likely a 110 ohm aes/ebu cable, while the oppo's coaxial digital connection is undoubtedly a 75 ohm spdif interface. That impedance mismatch will cause substantial waveform degradation, resulting in greatly increased jitter and perhaps even mis-clocking.

Matching of the "characteristic impedance" of a cable to the output impedance of the source device and the input impedance of the load device is not important at analog audio frequencies, but it is highly important for digital signals.

The following thread may be of interest:

http://forum.audiogon.com/cgi-bin/fr.pl?ddgtl&1253115041

Regards,
-- Al

Do these adapters also work on digital cables? I'm in the market for a newer CD/DVD player but a lot of them... like the OPPO's only have a coaxial digitial connection... and my digital cable is an XLR. Will an adapter alow me to use the digital XLR cable connected to the coaxial connection?

So the issue is narrowed down to, will the same cable (i.e. Kubala Sosna Emotion) in XLR version sound better than Kubala RCA (fitted with XLR adaptors) all else being equal? I have to possibility to try the Kubala XLR version.
Many thanks.

I think that you should check with the manufacturer to see if the rca cables are designed with the same internal configuration as the xlr cables. That may or may not be the case. If it is, try to arrange to have them re-terminated, so that you don't have to use adapters. If it is not the case, I agree with the preceding suggestions that the best course is likely to be investing in balanced cables, although of course it would be preferable to first do a comparison test (which you indicated is possible).

To realize the benefits of a balanced interface, the cable has to have the "hot" and "cold" signal conductors designed symmetrically. Such as a twisted pair of conductors, each of which conducts one of the two signals, with the twisted pair being contained within an overall shield together with a conductor for the ground connection.

If the rca cable is designed such that the signal and return conductor are asymmetrical, such as a coaxial arrangement, noise pickup in the two conductors will not be identical, and will therefore not be rejected (or will only be partially rejected) by the differential receiver stage at the destination end of the cable, even if the cable is re-terminated.

Also, as Q indicated, nearly all adapters tie the "cold" signal (pin 3 on most USA equipment) to ground (pin 1). Not all output driver designs can work properly under that condition (although the majority of high quality designs probably can). See this thread for an example of a problem that was caused by using an xlr-to-rca adapter on an output.

And even if the driver stage can handle having its pin 3 output signal grounded, the resulting output impedance asymmetry with respect to the pin 2 output would severely degrade common mode noise rejection, as has been pointed out.

BTW, I agree with Q that a balanced driver stage (meaning one that has output impedances on both outputs which are the same), connected to a balanced input via a properly designed balanced cable, will result in identical rejection of common mode noise regardless of whether a balanced signal pair is present or not. The effect of having one of the two signals not present (i.e., at zero volts) will be a 6db reduction in signal-to-noise ratio and in sensitivity. And if the destination component is "fully balanced," meaning that it has a balanced internal signal path, some of the advantages of fully balanced design, such as cancellation of some forms of internally generated distortion and reduction of power supply related noise, will not be realized.

Regards,
-- Al

Nsgarch

Everything _q says is wrong.

No, it's not. You're simply repeating one of the most common myths about balanced interfaces because you don't understand how they work.

For those who would like to inform themselves on how balanced interfaces work, this is helpful.

Answers to Common Questions About Audio Tansformers

Everything _q says is wrong.

Buy the new balanced cables and run your system fully balanced.

Hello all
thanks a lot for the thoughtful replies. I'd however like to refocus the thread for more meaningful discussion about my specific problem.

I have just bought a preamp (Jeff Rowland Synery 2i) which happens to have only balanced inputs and outputs (i.e. no RCA inputs or outputs). But my CD player and mono amps (tube amps) have both balanced and unbalanced possibilities.

I already have 2 very good interconnects(Kubala Sosna Emotion)and Audio Note ANVx but both are single ended. Thus the only way for me to be able to use the new preamp is either buying new XLR cables or use my existing RCA cables and fit them with the factory supplied XLR-RCA connectors (came with the Jeff Rowland preamp.

My question then is, is it worth investing big money in buying XLR cables or do I just use the adapters for connecting convenience) and make as if am using only RCA cables.

So the issue is narrowed down to, will the same cable (i.e. Kubala Sosna Emotion) in XLR version sound better than Kubala RCA (fitted with XLR adaptors) all else being equal? I have to possibility to try the Kubala XLR version.
Many thanks.

Nsgarch

1. Balanced cables are only "quieter" if carrying a balanced signal (a plus and a minus referenced to a neutral conductor) in which case noise in the normal (non-inverted) signal cancels with the noise in the inverted signal -- this is known as common mode rejection. So if, via an adaptor, one sends a single ended signal ONLY (in other words, half of a balanced signal) through a balanced cable, common mode rejection doesn't come into play. So there goes the noise rejection advantage.

I'm afraid that's incorrect.

Balanced interfaces and the common-mode rejection that they offer have nothing to do with the signal. A balanced interface rejects common-mode noise even when there is no signal present, and it doesn't matter whether the signal's being sent asymmetrically on one line, or symmetrically between both lines.

What "balanced" refers to is the impedance of each line with respect to ground. The source impedances must be balanced in order for the differential input to be most effective at rejecting common-mode noise. Any imbalance will degrade this rejection.

Lall

I would be grateful for feedback on any potential performance difference between an original Balanced (XLR) high end Interconnect (e.g. Kubala Sosna Emotion) and the same cable originally RCA ended but fitted with high quality XLR adapters to meet the needs of balanced gears (preamps and amps) with no RCA options. The entire system is high end.

Rather depends on the adapters. Many of them tie pin 3 to pin 1, attempting to turn the output or input into an unbalanced one. If you can find an adapter that only ties the RCA to pins 2 and 3 it can work well, especially if the cable is symmetrical and unshielded.

Grant is correct, however if you'd be interested in more detail, I just posted the following on another site:

1. Balanced cables are only "quieter" if carrying a balanced signal (a plus and a minus referenced to a neutral conductor) in which case noise in the normal (non-inverted) signal cancels with the noise in the inverted signal -- this is known as common mode rejection. So if, via an adaptor, one sends a single ended signal ONLY (in other words, half of a balanced signal) through a balanced cable, common mode rejection doesn't come into play. So there goes the noise rejection advantage.

2. Conventional amplifiers (both tube and ss) use what's known as a push-pull circuit, which requires a balanced signal to fuction properly/optimally. If you feed the amp an unbal (single ended) signal (through its RCA inputs), the signal first goes through a splitter/inverter circuit which changes the unbal signal into a balanced signal. If you feed the amp an unbal (single ended) signal through its bal inputs, there is no splitter/inverter between the bal inputs and the amplifier's gain stages, so the amp is literally getting only half a signal! An unbal-to-bal adaptor doesn't do the same thing as a splitter/inverter circuit. It doesn't make a bal signal out of an unbal one. It just enables you to send an unbal signal through a bal cable. So if you're in a pinch for some reason, and only have a balanced interconnect available to connect your preamp's unbalanced outputs to your amp, it's OK to use the adaptor on the preamp end, but you must also use one on the amp end so you can connect to the amp's single-ended inputs sending the signal through the splitter/inverter circuit first.

3. Another detail newbies always miss regarding bal versus unbal interconnections, is shielding. Even though balanced connectors have three pins/sockets, there are really four connections -- the fourth one being the barrel of the connector, which is attached to the cable's shield. When a balanced cable is connected to a chassis socket, the shield is connected to the chassis. The neutral pin (usually pin 1) of a balanced connector, is neutral relative to the plus-and-minus signal conductors (pin 2 and 3) but is NOT the same as the "ground" in an RCA connector. This gets us into some sticky territory when using bal/unbal adaptors because you have to know how the adaptor routes the single-ended's ground: i.e does it connect the (RCA) ground to the balanced adaptor's neutral, or to its barrel, or (quite often) to both?

My suggestion is if a given piece of gear offers only single ended connections, then use single ended interconnects because there's really nothing to be gained using an adaptor connected to a balanced interconnect -- and there's a good chance you will be shooting youself in the foot if you don't 're-convert' to an RCA connector at the amplifier end in order to engage the amplifier's splitter/inverter circuit.
.