SE vs Balanced volume


With my tube amp, when I switch between RCA & XLR, the XLR is louder. With my SS amp, when I switch between RCA & XLR, the volume is the same. The specs for the SS amp state that the amp is fully balanced. Here’s a photo of the innards behind the RCA & XLR inputs of the SS amp. XLR is on the left, RCA on the right, the yellow wires go to the external toggle switch. Why the different results? Thx.
steakster
Which SS and tube amps are you using? Many amps that have XLRs on them are not truly balanced.  Most are still single ended, the XLR connections are just converted back to SE internally.  I think BAT makes fully balanced tube amps.  Fully balance circuitry seems more common in SS amps.  In theory, with a fully balanced design you should get back about 6-10dB in volume.     
Forgot to include in the original post that the tube amp is definitely fully balanced. Can the attached photo help determine if the SS amp is also?
steakster
Here’s a photo of the innards behind the RCA & XLR inputs of the SS amp. XLR is on the left, RCA on the right, the yellow wires go to the external toggle switch. Why the different results? Thx.
Without the circuit, that to me looks like fake balanced with that opamp (INA134?) sitting there, if so then the single ended should sound better, unless your running 10mt interconnects, then the balanced "maybe" a touch quieter for noise but not better in sound.
As you got red green and white for opamp power, and the single whitish one to the right looks like it’s the single ended shielded signal carrier to the input board.

Just looked yes it is a fake balanced, that opamp is a balanced to single ended converter.
http://www.ti.com/lit/ds/sbos071/sbos071.pdf
Best opamp is no opamp, go the single ended input, which bypasses the opamp.


Cheers George
Your SS amp is a single ended design, it only has XLR inputs for convenience. Your tube amp is probably fully differential.

A fully differential design will exhibit 6-10 dB more gain when using the XLR jacks as nycjlee mentions above.
Thx for the info. My preamp definitely is fully balanced and offers both XLR and RCA outs. Re: Sound quality - if it’s better to use the amp RCA in, does it matter which preamp outs? Any benefit to using an IC with XLR out/RCA in?
steakster

Seeing this amp is fake balanced, I would just use the rca out of the preamp (just incase it's also fake output) into the rca in of the amp, for the best sound if your interconnects are <5mt.
No advantage in using balanced out of the preamp to rca in of the amp.

Cheers George
@steakster  I wouldn't do that, I agree with @georgehifi go with the RCAs.  He is spot on. 

Seeing this amp is fake balanced, I would just use the rca out of the preamp (just incase it's also fake output) into the rca in of the amp, for the best sound if your interconnects are <5mt.
No advantage in using balanced out of the preamp to rca in of the amp.
INA134 is a differential amp (instrumentation amp), designed for audio, that provides very high (90dB) Common Mode Noise Rejection.  Electrical noise induced in both input wires is canceled.  In addition signal wires in XLR cable are twisted providing very good electrical noise immunity.  In contrast, single ended front (RCA) provides zero rejection for common mode interference.
@steakster Which solid state amp is it? That would be helpful to others. Cheers,
Spencer
" Forgot to include in the original post that the tube amp is definitely fully balanced. Can the attached photo help determine if the SS amp is also?"

Look at the specs for both pieces. If a component is fully balanced, they list 2 sets of specs. One for SE operation and one for balanced. If 2 set of specs are listed, there should be a 6db difference between the 2. The balanced specs will be the larger number.
INA134 is a differential amp (instrumentation amp), designed for audio, that provides very high (90dB) Common Mode Noise Rejection.
It’s still a faked balanced into what is obviously not balanced amp, and it’s also an unwanted opamp in the signal path, of what must be a single end input amp. In the end it’s which is going to sound better, I’ll lay money on the single ended input.
Like the saying goes the best opamp is no opamp.

Cheers George
As I stated before it provides excellent noise rejection and that is the purpose of XLR connection. Amplifier does not have to be fully balanced to take advantage of inherent XLR noise rejection (and most of them are not). Some manufacturers use input transformers (that add distortions at low frequencies) while others use instrumentation amps. Instrumentation amps are widely used in measurement equipment to provide better noise immunity. As for the use of op-amps - many wonderful amps, preamps and DACs have them, including gear from Jeff Rowland.
Like I said the only time balanced is an advantage is for noise rejection, but is only an advantage over rca with very long interconnects, and there is no sound advantage, if anything it’s a worse sound in this fake balanced connection case above with an opamp in the signal path, the $2 INA134 was never designed for high-end audio use, more like PA audio.

Cheers George
 I agree that if you don't need noise rejection you might use RCA input.

To answer OP question - It is true balanced input stage with gain of +1 and normally you should get 6dB gain. Perhaps there is some gain compensation (divider) after this instrumentation amp?  I would also suspect that your preamp might provide +1/2 and -1/2 instead of +1 and -1 outputs, but it is not likely since you already stated that it sound louder with tube amp.  Again this Instrumentation amp has gain of one and reponds to the signal difference between inputs. If you feed input with +1 and -1 then it will produce +2 (+6dB) on the output.
Given Steakster’s mention that the solid state amp is described as being fully balanced, a possibility that hasn’t yet been considered is that the INA134, rather than converting balanced input signals to an unbalanced signal, might be used to invert the unbalanced input and thereby provide the rest of the amp’s circuitry with a balanced pair of signals when the unbalanced input is used. That would be consistent with his observation that the balanced and unbalanced inputs provide the same gain, as well as with the “fully balanced” description of the amp.

George indicated that the whitish cable appearing at the right side of the photo appears to be a one conductor shielded cable. That could certainly be the case, but it is also possible that it is a two-conductor shielded cable, conducting a balanced pair of signals. For example, something like Belden 8761 (although that particular cable is a different color), which incorporates a pair of 22 gauge conductors and a shield yet has an overall diameter of only 0.175 inches.

And note that one of the suggested applications listed in the INA134 datasheet is "unity-gain inverting amplifier.”

Also, given that three wires are connected to the XLR/RCA switch it is possible that the switch is wired to select whether XLR pin 3 or the output of the INA134 is routed to the subsequent circuitry in the amp. With XLR pin 2 and the RCA center pin being connected together and to the other polarity of the subsequent circuitry.

Just a thought, and a possibility as I said. If what I’m speculating is the case, though, using balanced cables between that amp and the preamp would certainly seem likely to be the best way to go.

In any event, Steakster, do let us know what make and model the amp is, which may allow us to provide further insight.

Regards,
-- Al

"Some manufacturers use input transformers (that add distortions at low frequencies) while others use instrumentation amps.

True but there are ways to reduce the distortions using input transformers.  Depends on the design.

"As for the use of op-amps - many wonderful amps, preamps and DACs have them, including gear from Jeff Rowland."

True again but that depends on what you consider "wonderful"

Happy Listening.
but it is also possible that it is a two-conductor shielded cable, conducting a balanced pair of signals.
It maybe Al but not used as such when the fake balancing opamp only has a single ended output. Only if the opamp was the  INA2134 not a INA134
Cheers George
It maybe Al but not used as such when the fake balancing opamp only has a single ended output. Only if the opamp was the  INA2134 not a INA134

George, note the following statement in my previous post:
Also, given that three wires are connected to the XLR/RCA switch it is possible that the switch is wired to select whether XLR pin 3 or the output of the INA134 is routed to the subsequent circuitry in the amp. With XLR pin 2 and the RCA center pin being connected together and to the other polarity of the subsequent circuitry.
So if my speculation is correct, when the RCA input is selected with the switch one conductor in the whitish cable would be conducting the output signal of the INA134 (with the INA134 being used as an inverter for the RCA input, not as a balanced-to-unbalanced converter), and the cable would be conducting the signal from the RCA connector in the other conductor.  When the XLR input is selected the output of the INA134 would not be used, with the signal from XLR pin 3 being selected by the switch and used instead.

The result, if that speculation is correct, being that even though the amp may be fully balanced it would provide the same gain regardless of whether its RCA input or its XLR input is selected.

Regards,
-- Al
 
I’m not sure why you think this amp is fake. The function of such differential amps is to suppress common mode noise and to convert differential signal into single ended. Most of power amps out there are not fully balanced while using XLR connector. Suppression of common mode noise (the purpose of XLR) is achieved either by differential amp (like INA134) or input transformer.

Al, it is possible that they use this amp to invert signal, but why? What are they going to do with two signals of the same polarity? They cannot short them together while summing them involves resistors (poor CMRR) or another amp. It seems much simpler just to use INA134 as differential amplifier and use SPDT switch to select between its output and RCA input. In addition, summing of the signal when only half of it passes thru inverting op-amp compromises CMRR at higher frequencies.
Al, it is possible that they use this amp to invert signal, but why? What are they going to do with two signals of the same polarity?
Hi Kijanki,

Not sure what you mean by this.  I have not referred to two signals of the same polarity.  What I am describing is a possibility that the amp has a fully balanced signal path, and when its RCA input is selected the signal from the RCA connector together with its inverse (generated by the INA134 used as an inverter) form a balanced pair of signals that is provided to the rest of the circuitry in the amp.
It seems much simpler just to use INA134 as differential amplifier and use SPDT switch to select between its output and RCA input.
But if the amp is fully balanced, as has been claimed, this would not provide the rest of its circuitry with a balanced pair of signals.
In addition, summing of the signal when only half of it passes thru inverting op-amp compromises CMRR at higher frequencies.
I have not referred to any signals being summed.  But yes, if the amp is fully balanced and only one of the two signals provided to the rest of its circuitry goes through an inverter when the RCA input is used, that can be expected to result in some degree of sonic compromise.  But arguably some degree of sonic compromise will result with any fully balanced amp if it is provided with an unbalanced input.

The bottom line is that my hypothesis would explain why the amp may be fully balanced, as has been claimed, while at the same time providing equal gains regardless of whether its RCA or XLR input is used, as the OP has reported finding.

Best regards,
-- Al
 
Al, thanks - now I got it.  I forgot that this power amp is fully balanced.  In that case this amp is in the signal path for RCA input, but not for the XLR.  
In that case this amp is in the signal path for RCA input, but not for the XLR.
Exactly! To be sure it's clear to others, "amp" in this sentence refers to the INA134, not to the entire amplifier.

Best regards,
-- Al
 
If you guys look at the shielded cable soldered connection you'll see it's one live and two connections to the ground plane making it single ended.

https://onedrive.live.com/?authkey=%21AGcHA6Bqk4TpTLo&cid=31799128EF335049&id=31799128EF335049%2127368&parId=31799128EF335049%21117&o=OneUp

Cheers George 
Hi George,

If you are referring to the three solder pads that are above R1, as far as I can tell those aren't necessarily where the cable is connected.  And I could be wrong, but I was thinking they might be for the RCA connector.

Regards,
-- Al
 
That’s could be right also Al, didn’t look at it from that angle, just saw the shielded cable going to that point. In any case I wouldn’t like to have that single ended INA134 in the signal path on my amp/s.
And as I said before if the amp was full balanced they would have used the INA2134 to get balanced output from the opamp even then, yuck!.

Cheers George
George, if amp is fully balanced there is no need to put anything in between, I agree.  To convert single ended to balanced they could use different amp, like OPA1632 (That Rowland uses in Capri), that have balanced input/output.  Grounding one input would keep everything symmetrical, but would require, slightly more expensive DPDT,  instead of SPDT switch.  It appears that the main objective was to keep balanced connection clean/straight (since amp is fully balanced) paying less attention to single ended input.
How can they feed a balanced signal from 1 x INA134 to the amp input board, when it (IN134) only has a single ended output? They would either need two IN134's per channel or 1 x  IN2134.

Cheers George 
Hi George,

Please carefully re-read my various posts above. As I have explained in precise detail multiple times, and has Kijanki has also stated multiple times, **if** the amp is in fact fully balanced, as claimed, and if the rest of my hypothesis is correct, when the amp is provided with a balanced pair of input signals via its XLR connector there would be no need for the INA134 to be in the signal path at all, and setting the XLR/RCA switch to "XLR" would remove it from the signal path. Aside, that is, from the slight loading effect its input impedance would present to the signal on XLR pin 2.

And when the RCA input is used, the balanced pair of signals provided to the amp's input board would be comprised of the input signal on the RCA connector together with the inverse polarity of that signal, generated by the INA134 configured as an inverter.

I’m not sure how to explain that any more clearly than I already have.

Regards,
-- Al

Once again, Al figured out another mystery. Thx for explaining why the volume is the same between the two inputs. Not only do the specs say that the amp is fully balanced, but the dealer confirmed it. Next is to try same brand/model interconnects with two different terminations to hear if there might be any qualitative differences. Up until now, different brands have been in rotation, with the balanced cable as the primary feed. Either way, my music sounds incredible. Thx to all for contributing.
As I have explained in precise detail multiple times
Sorry Al, but you assume the amp itself is fully balanced, where and who said this. Can the OP state what amp make and model it is? So we can stop the assumptions.

PS: I just looked at the OP's posts on his amps they use a  (12ax7 as a phase splitter) still saying they're fully balanced?  

Because what I see is the IN134 used to make a balanced input single ended, just as I’ve seen with a certain Classe models, which is fake balanced.

Cheers George
Sorry Al, but you assume the amp itself is fully balanced, where and who said this.
It was stated in the original post at the start of the thread, and reaffirmed in the OP's post just above.

Regards,
-- Al
 
Read PS: above.

Cheers George
Read PS: above.
The amp in question is the solid state amp, not the tube amp.

Regards,
-- Al
 

Ok then it's now solid state, can the OP still post the make and model of it.? So we can put this to bed.

Cheers George
And also Al while were waiting for the OP's answer, I use two s/s amps with fully balance discrete input stage. This amp has rca inputs and xlr inputs with a switch between the two, no opamps involved. When switched to rca it just put’s pin 3 of the xlr to ground.

Cheers George
I use two s/s amps with fully balance discrete input stage. This amp has rca inputs and xlr inputs with a switch between the two, no opamps involved. When switched to rca it just put’s pin 3 of the xlr to ground.
Yes, that approach is used fairly widely, of course. And since the difference in voltage between the two input lines would be twice as much when XLR pins 2 and 3 are provided with a balanced pair of signals, compared to when pin 3 is grounded, there will be a 6 db difference in the gain of the amp between the two conditions. But in the case of the OP’s solid state amp there is no difference in gain when the XLR input is used (and provided with a balanced pair of signals) compared to when the RCA input is used, which was the reason for his question that started the thread.

Regards,
-- Al


So what your saying then is the RCA input uses the IN134 to provide an extra 6db of gain to keep both inputs the same level, but there’s no feedback network to set the gain, and as the data sheet says it’s fixed gain at unity. Also then wouldn’t the switch have to have to be more than just a simple two pole switch?
If not then and I’m wrong the rca is at a disadvantage having to go through the opamp.

Burr Brown INA134
FEATURES ● SINGLE ● FIXED GAIN = 0dB (1V/V) AUDIO DIFFERENTIAL LINE RECEIVER ● SUMMING AMPLIFIER ● UNITY-GAIN INVERTING AMPLIFIER ●

Cheers George
So what your saying then is the RCA input uses the IN134 to provide an extra 6db of gain to keep both inputs the same level, but there’s no feedback network to set the gain, and as the data sheet says it’s fixed gain at unity. Also then wouldn’t the switch have to have to be more than just a simple two pole switch?
No, that is not what I am saying. Again, please carefully re-read my earlier posts.

My hypothesis, as previously stated multiple times and clearly understood by Kijanki, is an attempt to account for the fact that the amp provides the same gain from its XLR and RCA inputs, while apparently being fully balanced. Under that hypothesis the INA134 is not being used to provide 6 db of gain, it is being used as a unity gain inverting amplifier, which as you indicated is one of the ways it is intended to be used. But that use results in the power amp’s overall signal path having 6 db more gain from its RCA input than it otherwise would, because the signal coming in on the RCA connector together with the inverse of that signal provided at the output of the INA134 enables the input board to be provided with a balanced pair of signals. With the voltage difference between those two signals being essentially the same as the voltage difference between the balanced pair of signals that would be provided to the amp if an XLR connection were used between the preamp and the amp (for the same setting of the volume control).

And under that hypothesis, the switch would be a single pole double throw type, which as I stated earlier multiple times would select between the output of the INA134 and XLR pin 3, depending on whether the switch is set to the RCA position or the XLR position.

And yes, under this hypothesis the RCA input is at a disadvantage sonically, at least theoretically, since the INA134 is in the signal path when the RCA input is used but is not in the signal path when the XLR input is used (aside from the slight loading presented to XLR pin 2 by its input impedance). I and Kijanki both stated that earlier.

Regards,
-- Al

"So what your saying then is the RCA input uses the IN134 to provide an extra 6db of gain to keep both inputs the same level, but there’s no feedback network to set the gain, and as the data sheet says it’s fixed gain at unity."

No. The IN134 is not used to get an extra 6db. If you have a fully balanced component that needs to work with other single ended gear, you need to configure it to do so. Think of it as an adaptor. Also, you can't run a fully balanced component SE and Bal at the same time. The circuitry changes. You can't just use 2 pins of an xlr connector to make a SE connection. The pins do different things. If you've ever seen a component that has a little jumper connecting 2 pins of an xlr cable, that's why.

"Also then wouldn’t the switch have to have to be more than just a simple two pole switch? "

Since there's only 3 signal path/pins for balanced, the only change you can make is a simple 2 way switch. You're going from 3 pins to 2.

"If not then and I’m wrong the rca is at a disadvantage having to go through the opamp."

Correct. The opamp is another component in the single path, so it has to have some effect. In many cases, though, the complete opposite is true. If you have a singled amp but want it to work with balanced cables, the opamp is used on the xlr and not the rca.
Sorry Al. You put your last post up while I was typing mine, so I didn't see it. You're much better at explaining this type of thing than I am, so I'm sure you would have had to clarify some of my posts anyway.
because the signal coming in on the RCA connector together with the inverse of that signal provided at the output of the INA134 enables the input board to be provided with a balanced pair of signals.
That means one phase only going to the "balanced" input board is going through the INA134, as it’s clearly just a single ended output, pin 6 on the data sheet. If that's the way it’s no wonder some think the balanced sound better. This should be called fake single ended. Why sacrifice sound quality of the rca for the sake of the same gain.
http://circuits.datasheetdir.com/164/INA134-pinout.jpg

I can see it using the AD8476.
http://www.mouser.in/images/microsites/low-power-diff-con-fig01.jpg

Cheers George
No problem, Sfall. Explanations stated from a second perspective can only add clarity.
Georgehifi 3-12-2017
Why sacrifice sound quality of the rca for the sake of the same gain.
Some fully balanced architectures will not work properly unless they are provided with a balanced pair of signals. For example, some or all ARC Reference series power amplifiers, which do not provide RCA connectors, are designed such that only half of their balanced signal path will see a signal if their XLR connector is provided with a signal on pin 2 but pin 3 is grounded. In that situation the result would be a drastic reduction in power capability, as well as various adverse sonic consequences. So in this case there may be a more important reason for the approach I’ve envisioned than just adding gain for the RCA input.

As Sfall said, if my hypothesis is correct the use of the INA134 can be thought of as an adapter. And it could be viewed as analogous to the many designs having unbalanced internal signal paths in which a balanced pair of output signals is created by means of an op amp configured as an inverter. Is that an ideal approach? No. But it is a simple and inexpensive way of providing a feature that can be useful in some systems.

Regards,
-- Al

George, it is interesting that you picked AD8476 that I use in my projects in 5V applications to convert single ended output to bidirectional A/D converter  (ADS1271) shifting common mode by 2.5V
@steakster

What is this amp??? That you got rid of Artemis 6c33c monoblocks for, I am intrigued.

http://img.usaudiomart.com/uploads/large/1369416-cr-developments-artemis-mono-blocks.jpg

http://img.usaudiomart.com/uploads/large/1369418-cr-developments-artemis-mono-blocks.jpg

Cheers George