Balanced outputs - how to find out?

[ʝɵʝɵ]

Recently I have been rearranging some of my gear and was thinking about balanced audio - all my synths are connected through the typical dual mono output jacks, and I tried reading up the technical specifications in their respective manuals to figure out if those outputs are balanced or unbalanced, but there was no mention of this. I'd like to take advantage of balanced audio to get rid of the little line noise and hum that I have, but I'd rather not stick a DI box behind each and every synth unless it's absolutely necessary.

So, is there an easy way of testing if the output is balanced? I suppose I could send a mail to most vendors but I doubt they would give me answers about 20-30 year old gear.

If not, here's the list of synths I'm interested in figuring out if they have balanced output jacks or not:
- Roland D-50
- Roland XV-5080
- Roland U-220
- Korg X5DR
- Korg Triton Rack
- Yamaha AN1x
- Yamaha CS6R
- Yamaha TG77
- Kawai K4

Cheers!

[madtheory]

If it doesn't say, then it's unbal. That's the most common, as it saves money and most users don't need bal. I can confirm from experience that the D50, all of the Yamahas (but I don't know the CS6R) and the K4 are unbalanced line level. Don't the manuals show the TS connection for the jack in the specs? Or you could check the service manuals?

I'd bet good money though, that every single one is unbalanced.

What kind of noise? Hum, buzz, hiss or crackle? Each one is a symptom with a fairly specific cause.

Personally I would prefer not to use a DI box to balance the signal, because it includes attenuation to mic level, then you add gain again with a mic pre to bring it back to line level. It will increase the noise floor a little, and colour the sound. Not by much, but if you're bothered by unbal, I think that will definitely bother you once you know it's a thing (as you do now LOL). I would use a "balun" box- something like the ART DTI or the Alice Match Pack is preferable. The Alice is a lot cleaner, and sometimes you'll find them in a job lot discarded from an old radio or TV install. Otherwise, expensive, as it's a lovely piece of engineering. Such a device doesn't attenuate, just balances.

Notwithstanding that, I've found over the years that if you have hum on the line (50Hz), it's more effective to investigate if there's EMI near the cable (bad PSUs, especially outboard ones, and the electric cable behind the plasterboard or under the floor are common suspects), and tie them up to avoid such sources. Take a bit of work, but it avoids the extra colouration of a balancing device.

If it's a quiet buzz (50Hz plus some harmonics) you are trying to solve, you should star connect all the plugboards (rather than putting them in series) clean all the earth pins and ensure screws are tight on the cables in all plugs. A little voltage difference from one end of the line to the other can be caused by corrosion and poor connections, because they make a voltage divider.

If it's a really bzzz buzz (i.e. 50Hz plus harmonics, high ones), then you have a ground loop. That means faffing about with checking schematics, lifting the screen at one end of a cable, labelling it a such, or possibly putting a HPF kludge on the screen. In those cases, I prefer a balun device because life's too short

[ʝɵʝɵ]

Hi madtheory,
thank you for your detailed and helpful (as so often ) reply.

That's the most common, as it saves money and most users don't need bal.

Which is kinda sad given that most of these devices were expensive high-quality studio equipment used by professionals.

Don't the manuals show the TS connection for the jack in the specs? Or you could check the service manuals?

The technical specs are text-only, I suppose I could see the typical "example setup" illustrations but I would assume that if they show plugs, it's just for illustrative purposes and not to show the exact stereo/mono plug to use. I could check service manuals I guess.

What kind of noise? Hum, buzz, hiss or crackle? Each one is a symptom with a fairly specific cause.

There is a very low hum, it's not audible at normal amplification. Also some weird noise at around 2 kHz + harmonics which is more pronounced than the hum. I will try to record a noise profile tonight to demonstrate the kind of noise I hear. At normal amplification it's not much of an issue but since I often record oneshot samples from my synths, I want those samples to be as clear as possible.

Can you tell if is the Behringer HD400 comparable to the ART DTI, or is it something completely different? If so, I could get that new cheaply, but I also just found an ART DTI on eBay for around 30€ so that's a comparable price.

[ʝɵʝɵ]

Hmm... maybe the synths are the least of the problems.

I captured some noise before and after turning the mixer on: mixer off/on comparison
This is obviously heavily amplified but you can hear the main noise that annoys me, it's 1 kHz + overtones, and it's already there when the mixer is turned off. The mixer itself only seems to add a little bit of hum which is way more quiet than this 1 kHz tone. Of course it could be coming from the computer itself but given the quality of the ESI Juli@ I would hope this is not the case (Edit: The noise level is much lower without any cables connected, so I guess it's not the card itself). However, if the noise is being picked up somewhere between the mixer and the sound card, I should be able to get rid of it with one of those little boxes since the Juli@ has balanced inputs, right?

I also figured out another interesting thing - I used to use my gear with a laptop, and when I powered the ESI M8U XL MIDI interface through the laptop's USB, I got some terrible high-frequency USB whine (probably around 5 kHz or so). The solution was to use an external power supply (which is optional - and I was thankful that ESI thought of this optional feature!), and the noise was much better.
Now with my desktop PC I realized that the noise is actually worse with the extra power supply. Even if a synth isn't even turned on, it may still pick up this noise (and muting the respective synth on the mixer gets rid of the noise, so it's not the mixer that is at fault here). I have now disconnected the power supply, and that already gets rid of a 14 kHz peak (not present in the noise profile above).

[madtheory]

Which is kinda sad given that most of these devices were expensive high-quality studio equipment used by professionals.

I think it can be argued that it is a sensible design decision. Balancing does not improve signal quality (as some people seem to think) it just greatly reduces the possibility of picking up noise, as you know. The circuit is more complex so if you want to have the same bandwidth and noise specs as the unbalanced version, so the cost goes up a lot, probably x4 (although with your experience with the cards I think you're in a better position to put a figure on that!). Have you seen the price of a panel mount XLR versus a PCB mount TS jack socket?? Plus, in a lot of cases, the quality could actually be a little worse because they might use a cheaper op-amp to save money.

I haven't used the Behringer. I've used the ART on the output of an LXP15II (notoriously prone to ground loops) and for synths in a PA system. I didn't hear a change in sound. The same transformer is in their passive DI. That thing definitely has a sound, but it is nice. Adds warmth and softening in comparison to a BSS active DI, which is super clean and amazing. This was with passive bass guitars, which have a weak signal so are more prone to being altered. I do know that the original Behringer active DI (a silver clone of the BSS) is really terrible, so that would put me off. But that was a while ago, their stuff is better now, and the Midas team is onboard... but it's hard to make a good transformer. Maybe Behringer have cloned ART's trafo, or maybe they bought the factory in China that makes it!

Hmm... maybe the synths are the least of the problems.

Yes, it sounds like there is EMI coming from a digital device. Does it change as you move the mouse? There are isolators for USB signals as well, but the USB 3.0 seem to be very expensive and there are a lot that are snake oil products for hifi. I don't know much about them, but electrically it would an opto-isolotor I think. Maybe you could build one? There is a market for it, obviously! But if that solved the problem, you would not need to compromise your audio lines with a cheap transformer. Hmm, that's a good advertising tag line...

The analogue parts of computers these days are often really bad, with ground loops and poor screening, and even bad board layout where EMI and RFI passes to adjacent components. It could even be a faulty neon bulb in a plugboard. The only way to check is to power everything down, then power things up to see when the noise comes back. But if it's the computer, that's really annoying!

I guess the solution might be to find the one audio line that is picking up the EMI, and balun that?

[ʝɵʝɵ]

Great news - just adding the HD400 between the mixer and audio interface got rid of the interference. The SNR is now significantly higher. Just recording silence from the mixer gives me about -100dB. Of course as soon as I enable some of its channels they add more noise, but it's a lot more manageable now.

[madtheory]

about -100dB.

Result! About as good as it gets.Does it change the sound?

[ʝɵʝɵ]

I can't notice any colourization, and the levels also seem to be the same as before.

Tomorrow or Monday I'll also have an ART DTI arriving - I actually have two mixers daisy-chained and there is still some hum between those two (but thankfully no EMI), so I will have the DTI hopefully salvage that connection as well. Or maybe I'll swap them and see if there is a difference... we'll see!

[ʝɵʝɵ]

Got the ART DTI now as well. Compared to the Behringer, the noise floor is definitely less flat - the treble is greatly boosted - but I think the impact on the rest of the signal is marginal, if there is one at all. Still, I think I'd rather go with the Behringer from what I have seen so far.

[madtheory]

Wow I wonder why that is?

[ʝɵʝɵ]

Good question. Maybe there's some circuitry related to compensating for some drop-off in the treble frequencies? My first thought was noise shaping like it's used for dithering, but I guess that doesn't really apply in the analog domain - and in particular the noise sounds louder than it should with this boost.

Interestingly there are some people complaining about the Behringer box, but from my observation its frequency response is as flat as it should be. I played some samples through the XV-5080, compared them with the original WAV files and the results are as identical as they can be with some analog circuits inbetween.

[madtheory]

There doesn't appear to be any circuitry in the audio path apart from the transformer (in the ART there's a capacitor to ground for the output side, but that's an LPF and not in the audio path as such, I think?). They both seem to have a basically flat response up to 30kHz or so. So maybe you're hearing LF harmonics (if the level going in is a little on the high side), or the phase shift? Some interesting measurements in this review:
https://www.soundonsound.com/reviews/art-dti

[ʝɵʝɵ]

Hmm, no idea. My unit is used but since the rest of the sound is okay, I doubt it's defective. My levels are not high. They are always below 0dB on the mixers when playing something, and obviously when just monitoring the noise they are at around -100dB.
Here is a direct comparison of the same signal chain with the ART in the first 20 seconds and then replaced with the Behringer for the other 20 seconds. I think you can tell that the white noise level is pretty much the same, and the only real difference is the buzzing coming from the ART DTI.

All of this was first done using regular unbalanced cables (2x mono lines). I made another comparison connecting balanced cables to the Behringer's output, but the noise floor was exactly the same as with unbalanced cables. So to anyone reading this thread: just the box itself should be enough.

[madtheory]

AudioXPress magazine has an article this month about capacitor testing, by Ethan Winer. Free sub here:
https://simplecirc.com/subscribe/sample ... cTnbESp_8g


He also tested the Behringer unit you have. It has a distortion profile in the audible band. He compared it to a Triad audio transformer (made in China).

The distortion is 60dB down which is low but it could certianly become audible in certain situations. If you could set up a blind ABX test for yourself (perhaps with the help of a friend to do the switching) it would be interesting to see if you could tell consistently which transformer was in the signal path. You probably could!

[ʝɵʝɵ]

Interesting article. I'll see if I can do some more extensive tests when I find the time - as said, when I played some samples through my XV-5080, they sounded pretty much identical to the WAV file on the PC, but I also didn't try to play them at extraordinarily low or high volumes. I always try to keep the output of my mixer at around 0dB anyway.