Balanced vs. Unbalanced - What does it mean?

See this paper regarding interfacing a single-ended preamp output to the balanced input of your power amp. It's a bit technical, but well worth reading. As you'll see on page 2 of this paper, and as JAllen suggested, an adapter cable can be used. But alternatively, for better noise rejection, a QUALITY transformer (such as the ones made by Jensen Transformers) can also be used.

http://www.jensen-transformers.com/an/an003.pdf

You can find lots of adapter cables and adapters here:

http://www.bhphotovideo.com/c/browse/Cables-In-Line-Accessories/ci/1667/N/4294549275

Here is one example:

http://www.bhphotovideo.com/c/product/133649-REG/Comprehensive_XLRP_PP_6ST_EXF_3_Pin_XLR_Male.html#features

More generally, concerning balanced vs. unbalanced, I'll add a few additional thoughts to the excellent comments which have been made.

Besides the noise immunity advantage of balanced interfaces, single-ended (rca) interfaces are far more prone to hum and ground-loop problems than balanced (xlr) interfaces. That is because a single-ended interface uses the shield of the interconnect cable to both carry the signal return current, and to connect the chassis of each of the two components together (which are in turn connected to both signal ground and ac safety ground). So any voltage offset or noise differential between the ac safety grounds of the two pieces of equipment will result in an extraneous current flowing through the shield, which the receiving component cannot distinguish from signal current. Besides sometimes causing problems which are difficult to troubleshoot (as evidenced by many threads here at Audiogon), that can limit your flexibility in providing separate filtering of the ac delivered to different components, and using multiple dedicated ac lines to prevent cross-coupling of, for instance, digital noise into analog components.

Also, it is important to distinguish between components which are "fully balanced," and those which are not. A fully balanced amplifier or preamplifier provides a completely separate signal path internally for each of the two polarities of the input and output of each channel. A component which is not fully balanced just has a differential input and/or output stage, but has a single-ended signal path internally. Your MC402 appears to be fully balanced, as indicated by this statement in the manual:

Exclusive Double Balanced Circuity: Each channel of the MC402 consists of two balanced power amplifiers operating in push-pull with their outputs combined in a McIntosh Autoformer. This double balanced configuration from input to output cancels virtually all distortion.

An amplifier which is not fully balanced would typically cost less, would not provide the (at least partial) cancellation of internally generated distortion which this paragraph refers to, and might actually sound worse with the balanced interface than with an unbalanced one, because of the extra stage which is introduced into the signal path at the input and/or output, to do the single-ended to balanced conversion. It would, however, provide some degree of rejection of noise pickup, which may or may not be important in the particular setup.

Finally, re long cables, it should be kept in mind that regardless of whether the interface is single-ended or balanced, if both the output impedance of the preamp (or other driving component) and the capacitance of the cable (which is proportional to length) are too high, upper treble rolloff will result. That's not a concern under most circumstances, particularly with solid state gear, but if the driving component's output impedance were say upwards of 500 ohms, and the run were say 30 feet or more, I would make a point of avoiding high capacitance cables (e.g., choose cables that are around 20 or 30 picofarads per foot, and definitely avoid those that are around 100 pf/foot or more).

Regards,
-- Al

I should add to the mention of adapters and adapter cables in my previous post that although these are viable alternatives for interfacing a single-ended output to a balanced input (using an rca-male to xlr-male cable or adapter), interfacing in the other direction (balanced output to single-ended input, using an xlr-female to rca-male cable or adapter) should not be done with typical commercially available cables and adapters without careful investigation of whether the output of the particular equipment can tolerate having one of the two opposite-polarity output signals connected to ground.

Adapters and adapter cables used for that purpose will typically ground pin 3 (which on USA equipment carries the negative-going or "cold" signal polarity). Grounding that signal will result in improper operation or damage to the output stages of some equipment.

Regards,
-- Al

Cam -- No, I think that Shadorne was speaking in more general terms, and the issue he cited about overdriving consumer-oriented gear with pro-oriented signals is not applicable to your situation.

It would be a concern if the outputs of a pro-oriented source component, or a pro-oriented equalizer or processor, were fed into the line-level inputs or tape inputs of a consumer-oriented preamplifier.

In your case, though, you are dealing with a preamp-to-power amp interface (meaning higher full-scale voltage than consumer source component line-level), and with both components being consumer-oriented. Your power amp has an input sensitivity of 2 volts in unbalanced mode, or 4 volts in balanced mode. That will certainly not result in clipping or overdriving when being fed by typical components.

In fact, the opposite will occur to a minor degree if you go with the adapter cable approach -- you will have to turn the volume control up by 6db relative to where you set it when using the power amp's unbalanced inputs. That is because instead of feeding the amp with a balanced pair of signals that have equal amplitudes and opposite polarities, you will be substituting ground (0 volts) for one of those polarities, resulting in the difference between them being half as much.

The only negative effect that could conceivably have would be to increase background hiss slightly, but unless something is marginal in your system to begin with that will not be a perceptible effect. You can test that right now, in unbalanced mode, by simply turning your volume control up a little bit from where you normally set it (corresponding to a 6db increase, as best as you can estimate that), with no music playing, and seeing if the hiss level becomes objectionable.

Regards,
-- Al

Bob Reynolds: It may not have been clear from the other posts, but there is zero advantage (other than the XLR being a better connector) of connecting the unbalanced signal from your preamp to the balanced input of your amp; there will be no noise rejection.

TVAD: IMO, there's zero benefit connecting single ended outputs (Linn) to balanced inputs (MC402). Might as well run single ended. But, try it both ways and decide for yourself.

In principle, I don't think that is true, gentlemen, although of course in practice whether it would yield a net improvement, a net degradation, or no net change is dependent on the specific equipment and the specific noise environment.

See Figure 2.1 of the Jensen paper I linked to earlier, and the words immediately under the figure:

http://www.jensen-transformers.com/an/an003.pdf

Note the statement that in this example CMRR will be "only 30db at 60Hz." However, 30db rejection of common mode noise (at least at the power line frequency) is considerably better than no noise rejection.

Basically, by using the adapter cable into a balanced input you are still driving a balanced differential input stage with a cable whose construction provides symmetry between the hot and cold signal lines. Therefore noise pickup in the cable will tend to be closer to equal between those two lines than between the shield and the center conductor of an unbalanced interconnect, allowing the differential input to still subtract out that fraction of the noise which is picked up symmetrically.

And note in Figure 2.1 that currents flowing in the shield of the interconnect due to voltage differentials between the chassis of the two components (caused by offsets in their ac safety grounds) will not be sensed by the differential receiver stage, while they definitely would be by a single-ended input stage.

Shadorne: However, a listen and see kind of approach is exactly how sonic properties become associated with interconnecting wires when the root cause is equipment related or impedance matching or mistaken volume level matching or any of hundreds of possibilities that have nothing to do with the purity of the copper. Creating a new myth every month, week or day.

Very well put! And I'll add that both the op and Mechans specifically seemed to be asking for technical explanations, which I think have now been provided and which, while obviously not being the final determinant of what is best in any given system, can help prevent going down the wrong path and even blowing up equipment (as cautioned against in my second prior post).

Regards,
-- Al

Yes, that sounds right, Bob. An adapter cable built with coax would provide much worse noise rejection than one built with shielded twisted pair, because of both the lack of physical symmetry between the two signal conductors, and because of the need to tie pins 1 and 3 together, to provide a connection between the chassis of the two components through the shield. That in turn would amount to a similar situation as a single-ended interface, in which inter-chassis ground-loop currents flow through a common path with signal return currents.

Which raises a good question about how typical adapters are designed internally. I'm not sure what the answer is.

Best,
-- Al