Balanced vs Unbalanced?

But during my test, I notice that it is not the case. I don’t know if its my system act differently or not.

The problem is that your gear doesn't support the balanced line standard, also known as AES48. This means that the interconnect cables will play a role in the system sound at the very least. It may well be that the balanced signal is going through additional circuits that the single-ended signal isn't.


To get everything out of balanced operation (otherwise, why do it?) you have to follow the standards. Otherwise the results are highly variable! Single ended connections don't have much in the way of standards. High end audio pretty much ignores AES48, perhaps because its harder to do. One of the most difficult issues is that in a balanced connection, the ground connection is ignored- there's no signal return currents in the shield of the interconnect cable. This is the aspect of balanced operation that is most commonly violated in high end audio.

Herman, Yes, that's true. I've found in practice though that it does not seem to come up. Also, again the CCS circuit plays a huge role in this- even if there is a mismatch in the tubes, you don't seem to get any extra noise.

BTW, any tube amp with a single-ended input can be modified to have a true balanced input, often without any change to the way it operates with a single-ended input, and without using an input transformer. IOW its possible to have a balanced front end and an SET as well, with all the benefits.

Alright, there is a fair amount of misinformation in this thread, and I thought since we essentially introduced balanced line operation to high end audio, I thought I should correct the misinformation herein and explain some of the whys.

Balanced operation exists for the sole reason of reducing/eliminating artifact in the interconnect cable. It will do so regardless of how long or short(!) the cable is. If it is set up right, it will always outperform single ended cables.

Balanced operation does not require twice as many components!! That is a very common myth. This is true even if the the preamp or amp is fully balanced. Now some balanced cirsuits will require twice as many components, and you will find that they also do not perform as well. The best balanced circuitry will also be differential. Differential circuits do not double components.

You can also run balanced operation without balanced components and realize all the benefits of the cables. This done via the use of transformers, as any transformer can convert from balanced to single-ended or vice versa, simply due to how its hooked up. I don't like transformers myself, but in cases where its either long interconnects or long speaker cables, the transformers and long interconnects will win out easily.

When I said that balanced operation is devised to eliminate cable artifact, I meant it- the cost of the cable becomes unimportant. However, to accomplish this, the balanced system has to be **low Impedance** (600 ohms is the standard, IOW what is driving the cable should be able to drive 600 ohms without distortion or loss of low frequencies). It is this last bit that has 90% of high end audio products falling flat on their collective faces, and is the reason why there is controversy about balanced at all. IOW if you don't embrace the standard, your balanced setup will be no better than single-ended and possibly worse.

Now if the components involved are internally balanced, you will realize two benefits: lower noise per stage of gain, and distortion cancellation with each stage of gain. So balanced preamps and amps can take advantage of that by using less gain stages as a result. This means that they will be more transparent, as reducing distortion reveals detail. I can expound more on that if you like.

Phono cartridges are balanced sources so LP can be run from needle to speaker fully balanced. The same is true of microphones and tape heads. CDs may or may not be balanced, but there is still advantage to using balanced operation with them due to the advantages of the interconnect itself. This is **not** to say that if your CDP is single ended that you can just stick a balanced cable on it and it will get better, it means that if you use it with a balanced preamp, the preamp will handle the signal fine and it will be delivered in better condition to the power amp.

I am certain I have forgotten a number of points but I expect they will come out in the responses to this post. The bottom line though is that balanced operation is a pathway to greater resolution and lower noise than is possible with single ended, but only if the standard is embraced.

I did forget something- a request. There are those of you here that disagree with my prior post and will have talking points. Please keep them if you can to one talking point at a time, so that we can be clear that they get responded to in a coherent fashion. Thanks!

Herman, just for the record, I like SETs a lot myself and have built a number of type 45 and 2A3 based amps.

But before we go into that,

One unmentioned difficulty is that for everything to work as designed you have to have a circuit capable of producing 2 signals that are exactly the same except for their polarity. I agree that using a differential circuit makes it easier but there are still a lot of things that can cause an imbalance. If there were not the CMRRs achieved would be much higher than they are.

The above comment is not exactly correct. It only applies to the use of balanced circuits **that are not differential**. IOW, if you have an imbalance in a differential circuit, it will be corrected in the next stage. In addition, a differential circuit will absolutely insure that the inverted and non-inverted phases of the signal will be absolutely 180 degrees out of phase but otherwise identical, even if their amplitudes are slightly different. So it does not have to be as exact as you describe in order for the CMRR to be retained.

I have found in the execution of differential circuits that the constant current source used in the circuit is a big deal. Most CCS circuits only have a single device (tube or transistor) and so are ineffective. They have to have at least 2 stages to work right. My point here is that doing balanced right is like anything else: a matter of execution. We get CMRRs well in excess of 100 db, even without perfectly matching signals.

Although its a different subject, you did mention the multiple tube issue in OTLs. That is no different from using multiple tubes or transistors in other amplifiers, and just like in other amps (including SETs) there are ways to make those tubes work quite well together! I agree that SETs have a lot of linearity, especially in that first watt where everything is so critical. It might interest you to know that the only other type of tube circuit that has the same behavior is an OTL- if designed right they have no crossover or notch issues, so like an SET the distortion vanishes as power goes down. Unlike an SET, they don't make nearly as much distortion at higher power, and so offer less coloration and smoother sound. It is easy to demonstrate. PM me.

Good source + triodes + good horn speakers = bliss

This is a formula that has worked well for me for a very long time :) I add to it: balanced operation- this ends the sonic role that interconnects traditionally play.

Rleff, it really depends on the speaker rather than the technology. Its the 'all else being equal' part that prevents an answer beyond that. Either one can be awesome or suck really bad.

Herman, look at it this way: a differential amplifier does not care if it is getting a single-ended or balanced input signal. It will act exactly the same in either case. So if the input signal has two opposing inputs that have slightly different amplitudes, the differential amplifier will still not care.

IOW, 1 volt at one input while the other is at ground is the same as two 0.5V at both inputs. Or 0.25V and one input and 0.75 at the other. The differential amp does not care- it just amplifies what is different between its two inputs, regardless of differing amplitudes.

In the meantime, the CMRR is not dependent on the signal, its dependent on the differential amplifier, and it gets an awful lot of that from how effective its Constant Current Source (CCS) is.

The bottom line is you can have an imbalance and it will work out fine, so you don't need loop feedback in the circuit to insure perfect balance.

BTW in general this issue is really poorly understood, so that was an excellent question!

OK Herman, yeah maybe a little, or maybe more than that :) But really, its pretty easy. If you drive the cathode of the input tube rather than the grid you have what is called a grounded-grid circuit. You need a coupling cap to do it. Turns out that the value of that cap is the same if that cap were to be used as a cathode bypass cap, in the case of normally driving the grid.

So, why not drive both the grid and the cathode in opposition? In practice, this works quite easily. You install an XLR on the amp, and attach the cathode bypass cap (if there is one, if not you have to supply one) to pin 3 of the XLR rather than ground. Pin 1 of the XLR goes to ground, and pin 2 goes to the RCA input. Now if you drive the grid only, single-ended, you will want to put a jumper between pin 1 and 3 if the cap was a bypass cap. If not, you don't need the jumper.

The only trick is that the grounded-grid input is pretty low impedance. You might want to have an auxiliary resistor that plugs into the RCA input that sets the normal input (pin 2) to the same value. This is not a problem if your preamp supports the balanced line standard.

I have this hookup running right now on a set of DIY amps I built, which are a set of push-pull amps using 45s for output tubes (class A), and a single-ended driver circuit similar to a Dyna ST-70, only all-triode. I didn't bother with the extra resistor, as this input behaves a lot like a differential input. I guarantee you that if you tie the two inputs together and try to inject a signal, you will find there is a substantial CMRR.

Koegz, in a nutshell, that is the whole point of balanced line operation. The system was created to eliminate interconnect cable problems.

There is a standard that has to be conformed to in order for this to be true. If the conformance is not there, you will still hear differences in the cable.

The primary hanging point with most high end gear is the 600 ohm termination standard. What this means is that the equipment should be able to drive a load of 600 ohms (which might be at the input of the amplifier) without difficulty (for example loss of bass). For tubes this usually means an output transformer, or very large coupling caps, 100uf or more. Most high end manufacturers don't want to put in such a large capacitor at the output of the preamp, as it will color the sound and for the same reason aren't usually too keen about an output transformer either. That's why conformance to the standard is rare.