Hi,
Given the same hardware, balanced will be 3dB noisier.
The external noise sensitivity is not really enhanced by balanced connections, IF the single ended circuit is laid out correctly.
Maybe less than 5%? I can think of two or 3 at the most.
The so-called 5-Pin Mini-Din on many (but not all) tone-arms is basically balanced, but the external wiring is usually not balanced (terminated with RCA).
It is a minimal in this application, for good reasons as well.
Ciao T
Given the same hardware, balanced will be 3dB noisier.
The external noise sensitivity is not really enhanced by balanced connections, IF the single ended circuit is laid out correctly.
Maybe less than 5%? I can think of two or 3 at the most.
The so-called 5-Pin Mini-Din on many (but not all) tone-arms is basically balanced, but the external wiring is usually not balanced (terminated with RCA).
It is a minimal in this application, for good reasons as well.
Ciao T
Thanks Thorsten for your input, it saves me a lot of trouble.
I essentially agree with Thorsten.
For many decades I have designed both balanced and unbalanced designs, BUT for phono stages, I generally prefer unbalanced. This is the reason: For a given input noise, without transformers, it takes 1/4 as many input devices. Since my favorite input devices are hard to buy these days, and expensive when you can find them, this presents a problem.
IF a fairly short run of phono cable is used, less than 10ft or 3 meters, then there is little or no inherent advantage to balanced input. It only requires common sense placement of external noise sources by locating them away from the phono cable.
Today, I am making 2 balanced input phono stages, and I have already have designed and use 2 unbalanced input phono stages.
Balanced is mostly a marketing 'advantage' and to be 'all inclusive' in the marketplace, it is a necessary accessory. After all, some phono playback systems will today be balanced in their wiring and optimum grounding.
However there is an added cost of about $200 OEM per channel, either way we do it.
Either we use a QUALITY transformer, the best that money can buy, off the shelf, or we use 4 times as many equivalent jfets that can cost real money, is necessary. As it stands today, I am using 4 times as many Vendetta Research quality jfets, and not using a transformer, in order to get the 10 ohms or better equivalent noise, without a transformer.
An alternative approach would be to use a single jfet or differential jfet pair (to lower inherent distortion) and a $200+ transformer. This would give balanced or unbalanced and the same, or even slightly lower input noise. For example, 5 ohms equivalent noise could be possible, if carefully designed. Transformers, however, have their own limitations, and their wire resistance can't be reduced to 0, and this will set the noise threshold.
Some will argue for transformers, but I have decided not to use them in my latest designs. They just are not completely defined, at least not well enough to trust in their ultimate sound quality for the project that I am working on, at least. SY will argue differently, but he only has to please himself. I have to please the 'world'. '-)
I essentially agree with Thorsten.
For many decades I have designed both balanced and unbalanced designs, BUT for phono stages, I generally prefer unbalanced. This is the reason: For a given input noise, without transformers, it takes 1/4 as many input devices. Since my favorite input devices are hard to buy these days, and expensive when you can find them, this presents a problem.
IF a fairly short run of phono cable is used, less than 10ft or 3 meters, then there is little or no inherent advantage to balanced input. It only requires common sense placement of external noise sources by locating them away from the phono cable.
Today, I am making 2 balanced input phono stages, and I have already have designed and use 2 unbalanced input phono stages.
Balanced is mostly a marketing 'advantage' and to be 'all inclusive' in the marketplace, it is a necessary accessory. After all, some phono playback systems will today be balanced in their wiring and optimum grounding.
However there is an added cost of about $200 OEM per channel, either way we do it.
Either we use a QUALITY transformer, the best that money can buy, off the shelf, or we use 4 times as many equivalent jfets that can cost real money, is necessary. As it stands today, I am using 4 times as many Vendetta Research quality jfets, and not using a transformer, in order to get the 10 ohms or better equivalent noise, without a transformer.
An alternative approach would be to use a single jfet or differential jfet pair (to lower inherent distortion) and a $200+ transformer. This would give balanced or unbalanced and the same, or even slightly lower input noise. For example, 5 ohms equivalent noise could be possible, if carefully designed. Transformers, however, have their own limitations, and their wire resistance can't be reduced to 0, and this will set the noise threshold.
Some will argue for transformers, but I have decided not to use them in my latest designs. They just are not completely defined, at least not well enough to trust in their ultimate sound quality for the project that I am working on, at least. SY will argue differently, but he only has to please himself. I have to please the 'world'. '-)
If I had to "please the world" (meaning beyond the tiny number of my readers and the somewhat larger but still tiny number of people paying multikilobucks for a phono stage), I would take an approach entirely different than the one I used or the one you use. You wouldn't be happy with it unless you were listening without peeking. 
At low signal levels and with Mu-metal cores the loss is practically immeasurable. Advantages in this case far outweigh disadvantages.
John
Yes, SY, I do NOT subject myself to double-blind listening tests, unnecessarily. I am not good at them, and mostly they don't show any differences, anyway. I just trust myself and my associates to listen and decide. This has worked for me for many decades. I would have to quit audio design, IF I listened to the 'double blinders'. What would be the point of trying so hard 'to get it right'? There are others who are better at mass production and 'cost engineering' than me. I would let them do it.
One of the headaches of this audio business is trying to just get the gain modules 'right'. Why can't there be the 'perfect' IC that we could drop in almost anywhere? I have been looking for this IC for the last 45 years. Perhaps, the very first one that I tried will do the job with little alteration. I bet that I could make a uA709 sound about as good as a typical jfet input IC, or a moderately noisy transistor input. Might be worth a try. Some progress, huh?
One of the headaches of this audio business is trying to just get the gain modules 'right'. Why can't there be the 'perfect' IC that we could drop in almost anywhere? I have been looking for this IC for the last 45 years. Perhaps, the very first one that I tried will do the job with little alteration. I bet that I could make a uA709 sound about as good as a typical jfet input IC, or a moderately noisy transistor input. Might be worth a try. Some progress, huh?
Well, they do when the differences are actually there...
Asking for a "perfect" IC that you can "drop in almost anywhere" is like asking for a perfect resistor or bolt or transistor or washer that you can "drop in almost anywhere." A more relevant question is, are there specific ICs optimized for different amplification functions that do perform (when tested by ear, no peeking) essentially perfectly in their intended application? I'm using one in my latest mike preamp which I have to admit thoroughly outperforms the best tube solutions I can do (still looking for a 1.4nV/rt Hz tube), and I'm not exactly a slouch at tube design.
The reasoning behind balanced lines is that it is a noise reduction mechanism for cable introduced noise. The noise reduction from balancing is primarily for low frequency "Magnetically" coupled fields. These are usually from the power line.
So in a turntable you often have a power cord for the motor which is a bit of distance away from the output cabling. You also often have some fairly decent shielding from the arm itself. Then there is the extra ground scheme often used to make sure that there is no potential difference between chassis.
So if the power line noise is at or below the thermal noise there is a disadvantage to using balanced lines. RF pickup is not really an issue unless you have really tightly twisted wires and a problem with the preamp's front end.
It really is an issue capable of clean design based decisions. All of the possible problem issues can be calculated and measured! Since the goal is clear... less noise, unless someone has tested a representative sample of turntables and found a significant number have induced magnetic fields that are significant then it may be worth reconsidering.
Of course if some turntables have a lot of power line noise then it is the turntables problem. Same thing if it is the pickup cartridge picking up the power line noise.
So in a turntable you often have a power cord for the motor which is a bit of distance away from the output cabling. You also often have some fairly decent shielding from the arm itself. Then there is the extra ground scheme often used to make sure that there is no potential difference between chassis.
So if the power line noise is at or below the thermal noise there is a disadvantage to using balanced lines. RF pickup is not really an issue unless you have really tightly twisted wires and a problem with the preamp's front end.
It really is an issue capable of clean design based decisions. All of the possible problem issues can be calculated and measured! Since the goal is clear... less noise, unless someone has tested a representative sample of turntables and found a significant number have induced magnetic fields that are significant then it may be worth reconsidering.
Of course if some turntables have a lot of power line noise then it is the turntables problem. Same thing if it is the pickup cartridge picking up the power line noise.
Last edited:
You're very fortunate to be in an environment free of all kinds of high frequency noise. And to have the room to avoid the "behind the equipment rack" rats' nest. The last is particularly a problem for those of us who multiamp and have to keep the amps in one central spot.
Exactly the opposite of my experience. For me the transformer beats the extra gain stage in those criteria. Don't tell John Curl, though.
John
I agree jlsem, IF you eliminate a gain stage with your QUALITY transformer, go for it. Many do. I don't use any more stages whether I use a transformer or not, so it is somewhat redundant to use a transformer. The problem with input transformers is: That the input impedance is traded for gain, so you wind up with a lowish input impedance and this 'might' not be the optimum source impedance for the phono cartridge.
Hi,
Actually, balanced lines do no such things.
They only suppress COMMON MODE noise, any differential mode noise is as bad as with an SE Circuit.
In the typical 1-2m Cable run of sensible cable that minimises differential noise (twisted pair, coaxial) from the tonearm the common mode noise pickup that balanced lines could deal with is minimal. So the fact that we have different impedances to ground at the two different conductors is is of very little, if any consequence.
Ciao T
Actually, balanced lines do no such things.
They only suppress COMMON MODE noise, any differential mode noise is as bad as with an SE Circuit.
In the typical 1-2m Cable run of sensible cable that minimises differential noise (twisted pair, coaxial) from the tonearm the common mode noise pickup that balanced lines could deal with is minimal. So the fact that we have different impedances to ground at the two different conductors is is of very little, if any consequence.
Ciao T
I know only of one standard issue, one where it's a list option (VPI black box), a few others for which it's special order.
Somewhat amusing, the companies that manufacture balanced phono stages do SE-out turntables, the one with a balanced-out TT manufactures single-ended phono stages only (and pretty mediocre ones, imo)
Hi Scott,
You're certainly right that there will be a 3dB noise hit in most cases, but if the noise in the end is still very low, there may be a net gain in sound quality depending on what balanced brings to the table. This could be an example of where going for the absolute lowest noise at the expense of not doing balanced could be a mistake. I'm not sure how many phono preamp designers have chosen the balanced input route, but I know that Boulder has on both its MM and MC inputs.
For example, in my opinion 1nV/rt Hz is plenty low enough noise for all but the very lowest-output MCs. That input noise level is certainly achievable with a balanced input design (and with the use of only N-channel JFETs).
There would also seem to be a "what's good for the goose is good for the gander" argument as well. If a given designer puts in balanced interconnect capability in the pursuit of higher sound quality, would it not be consistent to do the same for the phono inputs?
Cheers,
Bob
Bob,
This is indeed so.
However, my personal experience has so far been that the best optimised "balanced" circuitry never sounded as good as the amount of effort applied to a single ended circuit. I have couse also found the same with looped feedback across multiple stages and zero feedback.
On the other hand, I personally think the RCA Connector (especially most "heavy metal" audiophile types) sound very bad, next to most XLR connectors (note that in these tests all else was equal, only connectors changed).
So from where I personally stand the pursuit of the highest subjective sound quality uses local feedback only single ended circuitry but XLR connectors where-ever possible (or better).
Ciao T
This is indeed so.
However, my personal experience has so far been that the best optimised "balanced" circuitry never sounded as good as the amount of effort applied to a single ended circuit. I have couse also found the same with looped feedback across multiple stages and zero feedback.
On the other hand, I personally think the RCA Connector (especially most "heavy metal" audiophile types) sound very bad, next to most XLR connectors (note that in these tests all else was equal, only connectors changed).
So from where I personally stand the pursuit of the highest subjective sound quality uses local feedback only single ended circuitry but XLR connectors where-ever possible (or better).
Ciao T
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.