Balanced in phono stages preamp?

Hi jh, It sounds like we're not talking about the thing. How about if I create a bias voltage for the input section that is set at something other than ground? I could set it a -1 volt for example. You really want to eat those resistors??

-And since phono cartridges are balanced, it is actually surprising that balanced phono sections did not appear sooner! When we built the first one in 1989, we thought that someone had done it before us, as it seemed the logical thing to do. But apparently the idea of balanced lines was/is foreign to high end audio, which is a shame because they have been so successful for over 50 years in the recording industry.

Balanced lines were created with the intent of eliminating cable problems and differences but the high end community has preferred to throw money at single-ended cables instead. While the techniques has proven successful, it has hardly been economical :)

Balanced operation also gives you lower noise both in the cable and in the circuitry. A differential amplifier, one of the basic building blocks of a balanced phono circuit, has *6db* less noise than its equivalent single-ended circuit! In practice you get a little less than that, but over successive stages of gain this means a lot- 10 db less noise for example in only 2 stages of gain. This allows for greater simplicity (less stages of gain) which translates to being closer to the music.

All right, everyone, a little lesson about why all phono cartridges are balanced and how balanced operation works. Cartridges are balanced as they have 2 wires per channel, neither of which has to be tied to ground for it to work. Proof? invert the phase of one channel of your cartridge in your system and see if it hums (it doesn't).

Hagtech, the ground is the tone arm itself. Have you ever wondered why the phono is the only hookup in your system that requires an extra wire for grounding (else it hums)? This is the result of trying to operate a balanced source in single-ended mode. Other examples include tape heads on tape machines, the light pickup on a laser head, most microphones and nearly anything with a transformer-coupled output.

You don't need a 'center-tapped transformer' to 'force' the cartridge to be balanced. Any differential input will accept the cartridge without any such work! I've been doing this now for nearly 20 years with my cartridge and we've been building phono sections like this for 18 years. It works.

All of your recordings were made in the balanced domain. All of them. They only get converted to single-ended in playback- if you allow that to happen. For a long time, you had no choice- now you do. The benefits are lower noise with less gain stages (blacker background lower distortion wider bandwidth) from the phono preamp, and less buzz/humm/RF from the pickup wiring. There are no tradeoffs if executed properly- often the signal path is simpler!

Herman, single-ended/balanced has nothing to do with sound in space, quite simply the analogy falls apart and becomes a logical fallacy. SE/balanced has everything to do with how recordings are made and played back: that is where we need to be focused.

That is simply not true. It is not required. A cartridge connected to a single ended input sometimes will hum with the ground connected and sometimes without. Sometimes the ground connection makes no difference.

This is consistent with balanced operation: you may not always get hum when converting to SE. However there is a reason why nearly every turntable manufacturer has a 5 wire system rather than 4, to prevent ground loops (IOW the ground is not mixed with the signal- this implies that the minus output of the cartridge is actually the inverting output, and in balanced systems the inverting signal is often denoted by a minus sign).

Even BSRs and Garrards from the 60s and 70s are set up that way. As a result they can all be run balanced since the ground is not ground looped with the signal.

While it may be true that balanced equipment might have been used in the recording process, the groove in the disc is not balanced.

By that measure, neither is it single-ended! Once the sound is mechanically encoded, you have the same conundrum that you have with actual sound- it is neither SE or balanced- it simply is. It is the way we handle the sound, once it exists as an electrical signal, that makes all the difference (no pun intended :)

Hi jh, the way differential amplifiers work is that ground is ignored. Our preamps are fully differential from input to output and so the ground is merely a shield- no currents are passed through it. This helps prevent ground loops.

As I mentioned, differential amps ignore ground. They have two inputs: non-inverting (positive) and inverting (negative). The differential amplifier will do nothing if the signal is the same on both inputs; it will only amplify what is *different* between the two inputs.

Thus there is nothing disingenuous about the tone arm being ground- it *is* at ground and is providing a shield as the signal leaves the cartridge.

Balanced inputs that are not differential are a different story- ground can play a crucial role. Differential amplifiers (a subset of balanced amplifiers) consequently offer higher performance, less complexity and lower noise.

Herman, There is a tendency to try to create interpretations for all sorts of experiences that we have in the world. In the case of the SE/balanced issue, the interpretation that I am referring to is to identify sounds within the realm of SE or balanced when it is neither. Its OK to question this- I did myself for a while until I got that I had to give up the interpretive story. Then I was able to realize that sound simply is what it is and is neither SE or balanced but a series of pressure variances in the atmosphere.

It is when the sound arrives at a microphone that it suddenly appears in the realm of electronics and is subject to being SE or balanced. All dynamic pickups are balanced as are capacitive pickups (although their preamps are often SE) and all the output transformers on microphones are balanced and on and on.

There is merit to the argument of keeping what started balanced in that domain, but sound itself has nothing to do with this. To give you some perspective, SE signals travel in a cable with a single conductor and a ground. The ground connection is the 'return' circuit- where is the return circuit for sound itself? There is none- it is an entirely different phenomena from that of electronics.

On an LP the sound is encoded in grooves due to a 3rd phenomena: a mechanical process and is again neither SE or balanced. However FWIW the cutterhead is a balanced load insofar as the amp is concerned: like all other balanced devices reversing the connections does nothing except invert phase (a property of a balanced device).

If it helps at all, take a look at an Ampex 351 schematic and see how the input and output transformers are used to accept or create balanced operation. Its an eye opener: a center tap is not part of the equation...

Hi Jeff_jones, we wondered the same thing so we put the same circuit into a more budget priced preamp to see how it works. The same advantages appeared and the phono section of our MP-3 is quiet with cartridges of 0.2mV, while at the same time having only 2 stages of gain and passively equalized. IOW the approach works on a budget too.

Hi Herman, the point is that to operate balanced differential you *don't* need a common reference point. I don't know how that idea got started but it is mythological insofar as differential amplifiers are concerned.

As I have mentioned before, to accomplish balanced differential operation you only need two connections: the non-inverted signal and the inverted signal. Traveling together, the two signals can be remarkably resistant to noise *even without a ground*.

In fact in some extra noisy industrial environments the ground can actually make things noisier; sometimes it is omitted and the signal travels in a simple twisted pair. I have a friend who works in motion control and sees this a lot. As long as the bias requirements of the input stage are satisfied there are no worries. That's why I was mentioning the ability to use dual floating bias references in the exchange above.

I've seen instrumentation inputs that use ground references (non-differential) but they lack noise performance compared to those that ignore ground (differential). Differential amplifiers have a secondary advantage as they are simpler than otherwise 'balanced' circuits.

Any inductive pickup (phono cartridge, tape head, guitar pickup, dynamic microphone) can be operated balanced due to the non-polar quality of the device itself. In guitar pickups this can be a powerful advantage as hum is a big problem with electric guitars.

In a phono system the ground is only a shield and does not provide a reference- neither is it connected to the cartridge. In my previous post, we altered the biasing strategy of the input stage to prevent the input resistors from being an artificial center tap for the cartridge while satisfying the bias conditions that the input section needs. So there is no ground connection to the phono section nor one to the cartridge, yet the setup works fine, because differential amplifiers and balanced sources do not *need* ground references to work: differential amplifiers only amplify what is *different* at their inputs, in this case, the phono signal.

Hi Hagtech, when you get a chance, take a look at the schematics for the Ampex recording equipment- something tube like the 351 electronics.

What you will see is that the output has an *optional* switched center tap on the output matching transformer. The input, which is also balanced, has no center tap at all. Note that the input is conventional single-ended circuitry with a transformer to effect the balanced input. The ground is handled in a way almost identical to how it is done with a balanced phono setup. Of course, going balanced differential without the transformer offers the possibility of higher performance.

If you look at solid state gear, opamps are often used as a balanced (differential) input; the technique is common.

The point is in all these examples that balanced, even if it is not differential, does not require a center tap. However there is almost always three terminals (the example I gave above with two terminals was in an industrial situation) as you mentioned. In a phono system that third 'terminal' (ground) is shared between both channels.

In any event, the system works. When I did this for the first time in my home system as an experiment, about 1987, I was startled at how straightforward everything behaved. Over the years I've used a variety of interconnects, some shielded, some merely braided and all have operated without a trace of hum. Grabbing the braided cable and squeezing it with your hand made absolutely no difference in the noise floor; the system seemed to all accounts impervious. The technology to operate this way has existed for decades; I am at a loss to sort out why it had not been done sooner!

So in this particular case, the differential circuit *does* care about ground.

Well, no it doesn't. In our preamp we have two resistors because we have to have a 47K input impedance and because the tubes require a grid leak resistance- for biasing. We could have set the unit up with a pair of 1M resistors and put the 47K across the input, from pin 2 to pin 3. It would have worked as well, but it would have been more parts.

When we load the cartridge, the load is applied with a single resistor from pin 2 to pin 3 of the XLR. The ground functions only as shield with no signal currents.

Hi jh, if I provide two independent floating bias supplies for the input of the preamp, I don't have to have the grid resistors go to a common point. This way the cartridge does not have a 'center tap'. Yet it still works exactly the same way- no noise running as a balanced source into our balanced differential input. It does in tact satisfy all the requirements of a balanced source.

Inductors being what they are, you can create a balanced output with a transformer that is driven single-ended. And it works fine- the technique has been used for decades. Cartridges are no different in this regard.

You might consider the implications- I took a look at your website and you have some impressive little goodies!

Hi Herman, the input of the preamp is differential so the cathodes are tied together. The cathodes are also tied to a 2 stage vacuum tube constant current source whose cathode is tied to a resistor that is tied to B- (-250V).

So my bias references are a set of resistors that go from B+ to B- and have a -1 volt output on each one. One side is adjustable to match to the other network. Although this is not how we normally do it, it works as well.

Hi Herman, not if the voltages are kept the same. That's what the adjustment is for (in this demonstration case).

In our production preamps, the grid resistors are tied to ground and for the sake of simplicity, double as the input loading (however when other cartridge loading is used the single resistor for that purpose is tied directly across the output of the cartridge and has no connection to ground). We could have done a variety of scenarios here but I favor simplicity whenever possible.