Balanced vs RCA between amp/preamp

 You suggest to use transformer isolation and you deny the way to use it. 

What is your opinion?

Audio Research reply directly that I should use transformer coupled  to connect RCA pre with their XLR only power amplifier.

@runwell If you are using a single-ended preamp, using a transformer might be the only way to drive certain power amps with a balanced input. ARC is often one of them. I'm not familiar with their current lineup, but it was only 5-6 years ago that some of their gear was only offered with a balanced input, but that input had a poor CMRR (Common Mode Rejection Ratio) according to a personal friend of mine who used to work there (Kalvin Dahl). Apparently it was only about 20dB. This meant that if you tried to drive it with a single-ended source, the distortion of the amp was very high and the output power was much lower. So a transformer was the only game in town if you had a single-ended preamp with that amp (which otherwise got good reviews).

If you have an ARC amp and a single-ended preamp, from what I've seen of their circuits I would advise the use of a transformer like the Jensen to do the conversion from single-ended to balanced, so I agree with ARC's advice above.

I don't like using transformers in the signal path which is why I designed a direct-coupled circuit that supports AES48. But transformers have their place; most of the classical LPs or CDs of recordings from the Golden Age of stereo (1958 to 1963) have several transformers in the recording signal path. Jensen and Lundahl arguably make some of the best transformers for this purpose.

Based on what you've told me so far, I would use the transformers! 

" The problem with this approach" Do you mean Transformer coupled approach?

@runwell No. If you are dealing with a ground loop, the transformer can fix that, but take @ieales' post seriously.

So finally it goes down to the transformer isolation,it seems the usual way even the XLR output from the pre,they usually get the rca signal first and use the transformer to get the inverse signal and consist the XLR sigjnal. 

If the preamp has both RCA and XLR outputs its quite rare to see a transformer used for the XLR output.

In high end audio in particular, in a situation like this the RCA is one of the phases of the XLR output (which has an inverting and non-inverting output). The problem with this approach is that ground is being used to complete the circuit, and this leaves the circuit vulnerable to ground loops. One of the goals of the balanced line system is to eliminate ground loops.

    I just want to have a solution between RCA pre and XLR power amp, it seems the transformer coupled is the simple way to go.

It is. A set of Jensen ISOMAX line transformers should do the job. They make one that has RCA inputs and XLR outs.

https://www.jensen-transformers.com/home-theater/audiophile/

They make them to order too- so you can get RCA in and XLR out although you don't see that on their site.

@runwell Yes. We patented the use of a Circlotron output (which is direct-coupled and balanced) with a servo control to prevent DC Offsets from appearing at the output. The Circlotron is a push-pull circuit that allows for entirely symmetrical operation using a pair of tubes to drive the output. In our case we use 6SN7s for that task.

A Circlotron circuit employs floating B+ power supplies and so can have one side grounded or in the case of balanced output, both sides floating with equal impedance to ground, while at the same time not using the ground for anything to do with the signal. In this way it allows a direct-coupled tube output to support the balanced standard, AES48.

the sound of the interconnect won't be a thing

Malarkey!

Assuming balanced circuitry and connections are properly implemented, which far too often they are not, they can reduce common mode noise. FULL STOP.

For balanced interconnects to be immune to cable, they would have to be immune to the laws of physics. Balanced lines [and connectors] have impedance and impedance controls frequency response.

Regarding the full stop, internally in a true balanced circuit in addition to improving CMRR, you can also improve power supply immunity, and since CMRR can be high, you get insensitivity to ground noise.

Actually, the way the interconnects become "immune to cable" (sic) is through two mechanisms, all according to the laws of physics. The first is that the balanced system is low impedance- at 1000 ohms (or 600) the impedances are so low that the capacitance and inductance of the cable are swamped and become negligible- and thus inaudible. The second is that the signal is not sent as a ground return through the shield. If you violate either of these tenants, the cable becomes audible. This is easy to prove through both measurement and audition.

differential amps tend to be lower noise we don't need as much gain stages to get the job done

If number of stages is a criteria, use a transformer. Differential amps per se are all over the map in terms of gain vs noise. The devil is in the details.

Transformers limit bandwidth and make distortion. Without them the circuit is noticeably more transparent. We've been doing balanced line longer than anyone else in high end audio; don't think we haven't tried transformers, and of course we are known for being transformerless :)

the other reason for balanced lines is elimination of ground loop noise

I'll wager that 6 9's of 'balanced' home audiophile systems do not lift the shield at the destination. Depending on amplifier design, there may be ZERO ground loop improvement over an unbalanced system.

In a balanced line system there is no need to lift the ground at one end of the cable. You are correct about the amplifier design issue; IME/IMO many high end manufacturers don't understand grounding and don't understand how balanced lines work either.

The balanced line system was created to get rid of interconnect cable colorations

The balanced line system was invented by the telephone company to increase noise immunity and power transfer over long distances. Nominally 600Ω, but at 1 frequency ONLY! Low frequency performance is abysmal.The 'telephone sound' is a direct result of balanced cables.

:)  of course noise isn't a cable artifact, but you forgot to mention the loss of high frequencies, which made transcontinental phone calls impossible prior to balanced lines. If loss of highs is not a coloration, what is??  Yes, the phone company did just that, and the idea was quickly adapted by the recording industry as running microphone signals over a long distances was a similar problem. Through the ability to transfer power the system swamps cable capacitance and inductance (both of which cause interconnect cable coloration). This is the low impedance aspect to which I referred earlier. Our balanced preamps feature miniature power amps as their line sections- they can make a little power.

Many balanced devices are bodged to accommodate unbalanced operation and coupled with generally incorrect balanced wiring, they don't stand a snowballs chance of optimum operation in unbalanced mode.

If 4+ decades in electronics has taught me anything, it's that if it can be screwed up it will be:

• inexperienced engineers
• incompetent sales staff
• ignorant consumers

This has been a beef of mine for decades, but just because this is so does not mean that all balanced line products have these issues.

I should have said "Use caution if driving an unbalanced amplifier from the balanced outputs."

We run into that problem too. What most people don't understand is that balanced and single-ended operation is inherently incompatible; you can't have 'quasi balanced' or any such nonsense. Either its balanced, or in the case above, unbalanced. In our preamps, since the output floats with respect to ground (no signal return current in the ground) if you attempt to use only one of the signal pins (ex.: pin 2) and ground (pin 1) you will get a buzz. This is because pin 1 has no direct connection to the return circuit of pin 2 and so pin 2 acts like an antenna. To correct this, pin 3 in this situation has to be also tied to ground (and because our output floats, this can be done without damage) to prevent the buzz. But of course then the output of the preamp is single ended, no longer balanced.

that being said your Amplifier and preamp need to have transformers for the input,outputs to be what is considered a True balanced .

This statement is false. You don’t need transformers, but without them supporting the balanced standard gets a lot trickier. We developed a means that is direct-coupled and yet floats the same way that a transformer winding does.

Many owners of Atmasphere gear (amps to preamps....all balanced) have reported the same cable differences that others have reported on other balanced gear.

Oddly, they’ve not been reporting them to me.

A balanced signal is simply two unbalanced signals, one 180 degrees out of phase (mirror image).

This statement is incorrect. With a proper balanced signal, the non-inverted phase is created with respect to the inverted phase, **not ground** (which is for shielding only). So its not two unbalanced signals- its only one, which floats with respect to ground (IOW, if you wanted, you could run a balanced signal with only two wires by simply omitting the shield)! This fact is poorly understood, but think of a line transformer- it does not use a center tap- one side of the output winding is tied to pin 2 of the XLR, the other side to pin 3. No ground connection at all. **Balanced lines if properly executed ignore ground**.


So it isn’t two single-ended signals- although a lot of manufacturers think it is, and in so doing degrade the performance of the balanced connection (and allow for interconnect cable artifacts and ground loops to creep into the system sound equation). When we built the world’s first balanced line preamps back in 1989, it didn’t occur to us that we could do that without supporting the standard. I was a bit shocked when I started to see other balanced gear that in no way supported the standard, and then exotic interconnects appeared because they were needed to work with such substandard equipment. You’d think that audiophiles would jump at the chance to no longer have to spend big $$$$ on cables and yet still get the best results...


In a nutshell, here is the standard:
1) pin 1 ground, pins 2 and 3 are signal. Pin 2 in the US is non-inverted side of the signal
2) ground is ignored, the signal floats; the pin2 signal is created with respect to pin 3 and vice versa
3) the system is low impedance- if there is an output XLR, its able to drive 1000 to 2000 ohms with ease (the old standard was 600 ohms and our preamps support that)
4) the signal travels in an interconnect consisting of a twisted pair with an independent shield.

I clearly said that the pin lengths on the male connectors of my XLR cables appear to be the same for all of the pins. It is the female connector for which I described a difference, which when mated with a male connector having equal length pins would result in the ground connection being made first, upon insertion, and removed last, upon removal.

Al, for all these decades of doing balanced connections, I’d never noticed that. You have to look quite closely at the female connector!! @millercarbon, my apologies on this point.

Balanced are designed for professional use. That is why they are designed with the ground making contact first- professional use sees a lot of stuff being connected/disconnected while powered up. Making the ground connection first avoids the loud annoying buzz you get with RCA. Which with home use is a non-issue, you simply power off or even just change inputs to avoid the noise. Now if you have 20 roadies running around messing with the 768 connections in your system then balanced makes a lot of sense.

Balanced is locking because pro gear tends to get moved around a lot. So get balanced if you like to throw your system in a truck and set up at friends a lot. If your system stays in the room for years at a time then XLR is a total waste of money.

True balanced utilizes two duplicate sets of circuitry, the outputs of which are then compared in order to reject noise. That's how it works. One of the bedrock fundamental truths of high end audio is simpler is better. Another is better parts cost more and the fewer the parts the more you can spend on each part. So this one is not only wasted but by unnecessary duplication its actually counter productive. Except for professional use where runs long enough to go from your pre-amp out the door around the house and back into the room to your amp are the norm. If that's your system go balanced. Totally. Otherwise, another feature wasted on home use.

What else? Oh yeah. Like the man said half the time balanced doesn't really even mean balanced. It just means you can connect your XLR here. Decoy! Fell for it! Don't be that guy.

A good deal of this is false. First- the ground. The ground connection is not made first (ground is pin 1 of the XLR). The reason you can hot plug an XLR connection is because both signal pins connect at the same time (along with the ground). The input is far less able to pick up noise during the connection- no buzz, most of the time not even a pop.


Balanced operation has several nice benefits for home operation. Audiophiles often pay a lot for cables; if the balanced line system is set up properly (IOW the equipment used supports the Balanced Line standard, a.k.a. AES48) the sound of the interconnect won't be a thing. No having to audition cables- it will just work. So cables don't have to be expensive and you can run them longer distances (this is helpful if you have monoblock amps, you can place them by the speakers and keep your speaker cables short, which improves definition).


True balanced circuits do not have duplicate circuitry. So there aren't twice as many parts and so on. We use differential amplifiers in our circuits and because differential amps tend to be lower noise we don't need as much gain stages to get the job done. As a result with our gear there are only 4 stages of gain from the LOMC phono input to the loudspeakers. That is **a simpler signal path** than most single-ended circuits! Balanced operation can (and usually is) done with transformers, in which case the preamp and amps can be single-ended and it can still work quite well.

Other than immunity to interconnect cable artifacts, the other reason for balanced lines is elimination of ground loop noise. That is a benefit whether the cable is 60 feet or only 6 inches.


Now if the amp has balanced inputs processed by circuitry that hands off the signal to the actual single-ended input of the amp (I've seen this in a few consumer grade amps with balanced inputs) then it may well be that the system will sound better with the single ended inputs, simply because there is a simpler signal path. But if the system is internally balanced there won't be any looking back, it will simply sound better.