Only "half of the machine" used when using single-ended on a fully balanced DAC?

[Rantenti]

I have a Gustard X26 Pro DAC feeding a Topping A90. Both are said to be fully balanced from input to output. I am using XLR connection currently, but I need to lend the XLR cables to another system for a while, and I'm now using RCA cables between the DAC and Preamp.

1. For such fully balanced DAC/ preamps, is only "half of the circuitry" working (i.e. I'm wasting half of the machine) when I used single-ended connection?

2. If I have BOTH XLR and RCA cables installed between the X26 Pro and the Topping A90 (the Topping A90 has an RCA/ XLR input selector switch), is there any signal / current running in the RCA cable when I selected XLR input (and vice versa)?

Thank you very much!

 

[EdTice]

I have a Gustard X26 Pro DAC feeding a Topping A90. Both are said to be fully balanced from input to output. I am using XLR connection currently, but I need to lend the XLR cables to another system for a while, and I'm now using RCA cables between the DAC and Preamp.

1. For such fully balanced DAC/ preamps, is only "half of the circuitry" working (i.e. I'm wasting half of the machine) when I used single-ended connection?


Thank you very much!

I can't answer your second question and I'm not sure I can answer the first one in general since people can come up with all kinds of DAC circuitry and I guarantee that by even attempting to answer this I will learn something new.

It is very likely the the DAC does exactly the same thing regardless of what output connector you use. When the signal actually gets to an amplifier, a balanced connection need to get converted to unbalanced because that's how amplifiers work.

Some amplifiers (or preamps) just take half of the balanced connection to get unbalanced and run it through the same circuitry. Others will treat them differently. But that reason is because some RCA inputs are driven by consumer devices with crappy voltage (i.e. 0.75volts). i.e. the Behringer A800 applies 40dB of gain to RCA and 32 to XLR so you can reach peak output for either input. My guess is that it's done through an additional gain stage for RCA but I don't have a circuit diagram for the amp and never cared.

But it would be fairly wasteful for a DAC to have two different gain stages one for balanced and one for unbalanced. So my guess is that if you use unbalanced, unless you introduce a ground loop issue, there's essentially no change to what the DAC does.

I don't know the A90 at all but per the specs it can get to 25 Vpp at gain=high regardless of input used. Given that the amp can reach full output regardless of the input, if the DAC can put out the specified 2.9volts of input sensitivity unbalanced, nothing is really "wasted."

Does this help or did I just add confusion?

 

[devopsprodude]

With the A90, you get about half the power when using unbalanced headphones vs balanced. But it still provides gobs of power to unbalanced headphones with very low distortion. So you're not really missing anything.

 

[Vincent Kars]

Both are said to be fully balanced from input to output.

Are they?
https://shenzhenaudio.com/products/...5-0-xoms-xu216-high-performance-audio-decoder says they have both SE and BAL out but that is not equal to having the entire circuit e.g. from USB input to analog out being duplicated.
Most of the time a balanced out is a matter of a SE circuit where at the XLR out you have the hot, a balun inverting this signal (the cold) and a ground.

 

[devopsprodude]

It's worth referring to https://www.audiosciencereview.com/...topping-a90-headphone-amplifier-review.13592/ to see difference in power between unbalanced and balanced with the A90. It will transparently power just about any headphone, balanced or unbalanced with very low distortion.

 

[EdTice]

Are they?
https://shenzhenaudio.com/products/...5-0-xoms-xu216-high-performance-audio-decoder says they have both SE and BAL out but that is not equal to having the entire circuit e.g. from USB input to analog out being duplicated.
Most of the time a balanced out is a matter of a SE circuit where at the XLR out you have the hot, a balun inverting this signal (the cold) and a ground.

In order to be truly "balanced" wouldn't the DAC have to output four channels instead of stereo?

 

[Rantenti]

There are DAC/ preamps, usually top of the line of that company, which claim to be "TRUE and FULL balanced", while the same companies also produce mid-tier "balanced" DAC/ preamps, which may share some similar components as the top model, but are significantly cheaper, due to not being "TRUE AND FULLY balanced" all the way from input to processing to output (thus cost-saving). When opened, these mid-tier machines have much less components on the board compared with the top-tier ones, I suspect the difference is between "fully" and "not really fully" balanced.

So what I postulate is, for the mid-tier machines, there isn't a great difference how the signal goes through no matter you use RCA/ XLR, so you get similar quality for RCA/ XLR. But for the top tier ones, if I use RCA, I'm wasting the potential of the "fully balanced circuitry" and it's just like downgrading myself to the middle-of-the-road models and wasting half my money, if half of the circuitry remains idle with nothing going through. Then I'd better go and buy some XLR cables and make sure every component of the machine is working hard Doesn't feel so good to have fully paid four workers but only give work to two.

 

[radix]

In order to be truly "balanced" wouldn't the DAC have to output four channels instead of stereo?

Balanced audio is 4 signals. Each of left and right have a + phase and a - phase. So, 4 signals. The - phase is not the same as ground.

 

[charleski]

In order to be truly "balanced" wouldn't the DAC have to output four channels instead of stereo?

The Gustard uses two ES9038PRO chips. Each one has 8 output channels, I assume these are mixed down to produce balanced stereo with reduced distortion.

 

[Rantenti]

The Gustard uses two ES9038PRO chips. Each one has 8 output channels, I assume these are mixed down on produce balanced stereo with reduced distortion.

So I presume with balanced input, each chip uses 4 output channels to work with the + and 4 others to work with the -ve. If using RCA input, then only the 4 channels working the +ve signal will be used and the remaining four channels per chip will be idle?

 

[charleski]

So I presume with balanced input, each chip uses 4 output channels to work with the + and 4 others to work with the -ve. If using RCA input, then only the 4 channels working the +ve signal will be used and the remaining four channels per chip will be idle?

This would certainly be the simplest way of doing it. It's also possible to convert to single-ended by using a small line transformer, in which case both output polarities would be used to form the final signal. But frankly the SINAD from the RCA outputs is superb at 113dB and well below the audible threshold so I really wouldn't worry about it.

 

[ZolaIII]

All depends on implemented typology. There are design that will only use two opamp's instead of one for balanced, ones with two DAC's which depending on implementation can use and both DAC's in unbalanced mode of operating and one's that will use more channels DAC IC's in parallel mode.
Main thing is as long as it performs good and you don't have problems with it you really shouldn't care at all.

 

[KSTR]

1. For such fully balanced DAC/ preamps, is only "half of the circuitry" working (i.e. I'm wasting half of the machine) when I used single-ended connection?

No, as any competent design that claims to be "fully balanced from input to output" still must implement re-referencing the input signal pins' difference to the local circuit GND. That's the whole idea about going balanced, we don't want to propagate common-mode signals beyond this initial subtractor (the output of which then can either be balanced or not).

By this, you can feed it with balanced or unbalanced signals to the balanced input, the output remains the same. Actually, the transfer can still be balanced with the proper adapter at the source end but for short connections going unbalanced is OK.

The A90 has RCA input so by that alone it is clear it has an internal re-balancer to make the RCA signal symmetric for balanced drive.

 

[Rantenti]

No, as any competent design that claims to be "fully balanced from input to output" still must implement re-referencing the input signal pins' difference to the local circuit GND. That's the whole idea about going balanced, we don't want to propagate common-mode signals beyond this initial subtractor (the output of which then can either be balanced or not).

By this, you can feed it with balanced or unbalanced signals to the balanced input, the output remains the same. Actually, the transfer can still be balanced with the proper adapter at the source end but for short connections going unbalanced is OK.

The A90 has RCA input so by that alone it is clear it has an internal re-balancer to make the RCA signal symmetric for balanced drive.

This is reassuring for the part from the DAC output to the Preamp! I would expect a worse SNR for RCA input but it shouldn't be audible for the high-performance A90.

As for the DAC, I can expect only half the capability of the DAC chip (and subsequent circuitry) used should I choose to output to an RCA, what I lose common mode noise rejection as with XLR, but the D-A decoding performance and analogue signal quality should be the same as that in the +ve channel of the XLR output. Please correct me if I'm wrong.

Thanks.

 

[EdTice]

Balanced audio is 4 signals. Each of left and right have a + phase and a - phase. So, 4 signals. The - phase is not the same as ground.

Your post seems to contradict mine but I think we are saying the same thing although maybe I am just using the terminology wrong? A DAC that can put out two signals can't be truly balanced. The balanced outs are just a post-DAC circuit. But if the DAC can put out four signals, one could build a truly balanced system. Others have pointed out that the DACs in this device seem to be able to put out more signals than that!

 

[radix]

Your post seems to contradict mine but I think we are saying the same thing although maybe I am just using the terminology wrong? A DAC that can put out two signals can't be truly balanced. The balanced outs are just a post-DAC circuit. But if the DAC can put out four signals, one could build a truly balanced system. Others have pointed out that the DACs in this device seem to be able to put out more signals than that!

Many DAC chips do put out balanced analog. The ESS ES9038Q2M used in some Topping products puts out a balanced analog signal -- i.e. 4 signals. Thus, it is possible (and likely) that the analog side is fully balanced.

What is done in the DAC chip? I'm not sure. Do they decode the digital audio to a single phase analog then balance it? Do they decode straight to balanced? I'm not sure, but I don't think it really matters. The THD+N and SINAD numbers for the Topping products go well beyond inaudible to the human ear. The DAC chips are 32-bit (with some operations done in 64-bit), so things like digital volume control should have very little added noise.

 

[Adaboy4z]

I just purchased the S.M.S.L A0200, D0200 and H0200 in December. I also acquired a used Loxjie P20 out of curiosity. I wondered the same thing as the OP. Would I be wasting the XLR balanced inputs on the P20 by using RCA as I ran out of XLR connections. From what I gather in the discussion the answer is no...

 

[EdTice]

Many DAC chips do put out balanced analog. The ESS ES9038Q2M used in some Topping products puts out a balanced analog signal -- i.e. 4 signals. Thus, it is possible (and likely) that the analog side is fully balanced.

What is done in the DAC chip? I'm not sure. Do they decode the digital audio to a single phase analog then balance it? Do they decode straight to balanced? I'm not sure, but I don't think it really matters. The THD+N and SINAD numbers for the Topping products go well beyond inaudible to the human ear. The DAC chips are 32-bit (with some operations done in 64-bit), so things like digital volume control should have very little added noise.

I don't know this DAC so I can't say either. Some DACs claim that they use the four channels to do +v and -v for L and R and so what you get out of the DCA is balanced which would seem to be better as long as there aren't any timing issues/jitter between the two analog outputs. Otherwise balancing with analog circuitry would seem to be more accurate. But this really the real of curiosity here as I think we've established for the OP that using whatever cables happen to be handy is just fine and that the two outputs are really there for convenience

 

[kongwee]

Normally -6dB drop using RCA if both end are truly differential.

 

[EdTice]

Normally -6dB drop using RCA if both end are truly differential.

And in that case you really *are* only using half of the machine. And probably no audible difference. So suddenly you want half your money back!