Even cheaper solution than ART Cleanbox etc.?

[Pangloss]

Hi everyone,

I am experiencing a bit of hum on four channels of my setup, having successfully ironed out the rest of the problems. Part of me wants to know why the hum still persists but the other part of me just wants to throw money at the problem and make it go away.

I have been considering getting a couple of 2-channel transformer hum eliminators like the Behringer Microhd400 or the ART Cleanbox, which look like they could well do the trick.

However, I was wondering whether 4 of these would also do the trick:

Microphone Transformer (XLR-Jack)

My DAW output is balanced XLR and the desk in is unbalanced so in terms of connection it seems fine.

My only question is whether the 500 ohm input would be unsuitable for the line level out of the DAW.

I know I'm only saving a few quid but...

Also, is there likely to be any significant difference in transformer quality and frequency response between the kind of things mentioned above.

Is there any scope for a cheap Behringer unit to destroy my sound here or not?

Thanks for any advice,

Alec

[James Perrett]

The input impedance of that Maplin convertor is a little low for line outputs and I suspect that it may be prone to saturation if you feed it a hot line level output.

Before spending too much money on this, have you tried pseudo balanced cables?

Cheers

James.

[Hugh Robjohns]

Before spending too much money on this, have you tried pseudo balanced cables?

That won't work if the desk inputs are unbalanced -- you need balanced inputs to use pseudo-balanced cables.

I share James' concerns about the impedance and liklihood of saturation with those Maplin jobbies.

Hugh

[Pangloss]

Thanks guys, I suspected as much.

With these Cleanbox/MicroHD style passive transformer boxes, can I expect any noticeable degradation in signal?

At high frequencies?

Is there a minimum sort of spend to be made, even on something passive like this?

[Hugh Robjohns]

If the transformers are decent then you shouldn't have any problems at the high end. It's usually the bass that suffers first because the transformers saturate most easily with LF, and it needs a large transformer to handle high levels of LF.

Plenty of people have had great results with the Cleanbox II so you shouldn't have any problems.

Hugh

[Pangloss]

Thanks Hugh.

[James Perrett]

Before spending too much money on this, have you tried pseudo balanced cables?

That won't work if the desk inputs are unbalanced -- you need balanced inputs to use pseudo-balanced cables.

It depends on what sort of circuit the balanced output uses and highlights one of the shortcomings of modern active balanced outputs. A transformer balanced output would work fine with the cold output connected to ground and the hot output connected to the signal pin of your unbalanced input. A Cleanbox is effectively turning your existing output into an old fashioned transformer balanced output.

Cheers

James.

[Hugh Robjohns]

A transformer balanced output would work fine with the cold output connected to ground and the hot output connected to the signal pin of your unbalanced input.

True enough. A transformer output would also provide the galvanic isolation that is needed here to prevent the ground loop, and you are absolutely right, the CleanBox is serving that function.

In some cases you might be able to get away with connecting the ground of the unbalanced input to the cold side of the (electronically) balanced output -- which is what I guess you meant by the reference to the speduo-balanced cable idea -- but most will get upset if (ab)used in this way.

In my experience, the Cleanbox (or something like it) is by far the easiest and most effective solution.

Hugh

[ramthelinefeed]

Despite having read umpteen articles, I still get awfully confused sometimes about this sort of stuff.

Say I have a synth with an unbalanced line level output (1/4" jack)-

Should I:

- plug the synth into a cleanbox, and use an unbalanced lead to plug the cleanbox into an unbalanced mixer line input?

- plug the synth into a cleanbox, and use a balanced lead to plug the cleanbox into a balanced mixer line input

- plug the synth into DI box, and use a balanced lead to plug the DI box into a balanced mixer mic input?

And if it's the latter, should the DI box be active or passive?

Or will they all work fine?
Or will some combinations sound better than others?
And if I use a behringer box will it automatically sound rubbish whatever I do?
etc etc

[Hugh Robjohns]

- plug the synth into a cleanbox, and use an unbalanced lead to plug the cleanbox into an unbalanced mixer line input?

First, you should try plugging the unbalanced output into the unbalanced input. If it sounds okay, then carry on that way. The Cleanbox (and its ilk) are only necessary if you suffer ground loop (hum) problems.

But, if you do suffer hum when connecting the unbalanced output to an unbalanced input, then the chances are the problem is caused by a ground loop, and the solution is to break the ground loop in the signal path. The transformer in the Cleanbox (or any line-level capable equivalent) provides that 'galvanic isolation' and removes the ground loop while still passing the audio signal.

You can happily connect it using unbalanced cables from the unbalanced output to the transformer, and another unbalanced cable from the transformer to the unbalanced input.

- plug the synth into a cleanbox, and use a balanced lead to plug the cleanbox into a balanced mixer line input

If you mixer has electronically balanced inputs, that would work too. (If it has transformer balanced inputs, then that transformer will already provide the means of galvanic isolation of course).

If you have an electronically balanced input you can dispense with the Cleanbox and make use of the differential and ground-free properties of the balanced input, by using the often-mentioned pseudo-balanced cable solution.

The pseudo-balanced cable arrangement also breaks the ground loop, but it does it by connecting the unbalanced output's ground to the balanced input's cold terminal (instead of the input device's ground). The balanced input cold terminal is electrically separated from the balanced input's ground, and hence there can be no ground loop, and no hum.

The unbalanced output signal voltage is developed between the hot wire and the unbalanced device's ground. The balanced input sees that signal voltage being presented between its hot and cold terminals, and thus recieves the full signal as required -- but without sharing grounds and hence without a ground loop (hum).

- plug the synth into DI box, and use a balanced lead to plug the DI box into a balanced mixer mic input?

That would work too, and once again the transformer in the DI box provides galvanic isolation between the two system's grounds to prevent ground loops.

However, in this case, the line level output from the keyboard is usually attenuated to (a hot) mic level, and then you have to use the mixer's mic amp to bring the level back up again. As a result, this solution will introduce a little more noise than any of the previous solutions.

And if it's the latter, should the DI box be active or passive?

Doesn't normally matter for keyboards as their outputs are usually electronically buffered with a very low source impedance. The only advantage of active DIs is that they present a very high input impedance, which is required to maintain the tone and sustain of electric guitars and basses. I use passive DIs with all of my keyboards...

And if I use a behringer box will it automatically sound rubbish whatever I do

Depends on your definition of rubbish and your expectations. Personally, I wouldn't say it would sound rubbish, but I would expect it to be more fussy about signal levels and to suffer from a little more distortion and a reduced bandwidth compared to the class leading professional DI boxes on the market. So it will certainly sound different, and arguably less accurate. I would also expect the B product to prove less reliable and long-lived in the long term....

Hugh

[ramthelinefeed]

thanks Hugh! I think I'll print that out and keep it

[Pangloss]

Hi James,

I see what you are referring too now by pseudo-balanced.

I have actually already tried lifting the earth so that you have bal hot -> unbal signal and bal cold -> unbal ground but that didn't work.

I think, as you identified, the balancing on my outputs is not achieved by transformers and is perhaps not ideal.

I will try something like the cleanbox but it has got me wondering that if there is already hum being introduced into the signal from just the hot and cold outputs this must have entered the signal some way upstream and simply transformer-coupling this signal to the inputs will still transmit the same noise. I hope not.

Anyway, thanks for the help - I'll let you know.

[Pangloss]

Hi Feline/Hugh,

On the subject of pseudo-balancing, or more generally, connenting an unbalanced source to a balanced destination, I have also been confused at times by the large amount of conflicting information.

The simplest psuedo balancing case seems to be connecting source ground -> destination cold and source signal -> destination hot.

However, if we put that aside for the moment there seem to be two different candidate solutions, both using the same ingredients and I would be interested to know which is best.

Ingredients:

Balanced Cable (screened pair)
1 resistor (100 - 400 ohm)
1 mono jack plug
1 stereo jack plug

1) Pseudo Balanced Solution

Modification of that described above as such:

Balanced Output Connector: mono jack plug
Unbalanced Input Connector: stereo jack plug

Unbal Signal -> (via red wire) Bal Hot
Unbal Ground -> (via white wire) Bal Cold
Unbal Ground connected via resistor at unbal end, via screen to the Bal Ground. (Resistor can be bypassed by small capacitor to shunt RF for good measure).

2) True Balanced Solution

Described recently by Hugh as being close to true impedance matching (I suppose the closeness depending on the matching of the resistor value). Done as such:

Balanced Output Connector: mono jack plug
Unbalanced Input Connector: stereo jack plug

Unbal Signal -> (via red wire) Bal Hot
Unbal Ground connected via resistor at unbal end, (via white wire) to the Bal Cold.
Screen connected to Bal Ground and disconnected and Unbal end.

I have read descriptions of how and why each should work but I don't know what is the ideal method.

Until now I have used the first method but never bothered with shunt capacitor due to space inside the jack barrel being limited enough anyway.

Are there any definitive thoughts on this?

Thanks,

Alec

[Hugh Robjohns]

I have actually already tried lifting the earth so that you have bal hot -> unbal signal and bal cold -> unbal ground but that didn't work.

Not surprised. Most electronically balanced outputs develop the signal between hot and ground and (inverted) Cold and ground. By not connecting the balanced device's ground to anything, there was no reference to extract the signal against. Also, some output devices get very upset if they are grounded.

I will try something like the cleanbox but it has got me wondering that if there is already hum being introduced into the signal from just the hot and cold outputs this must have entered the signal some way upstream and simply transformer-coupling this signal to the inputs will still transmit the same noise. I hope not.

It is possible -- but if you can identify where the hum is getting in, the Cleanbox will probably help resolve the problem at that point!

hugh

[Hugh Robjohns]

The simplest psuedo balancing case seems to be connecting source ground -> destination cold and source signal -> destination hot.

Yep. That gets the signal across, and it is the only way to do it. The confusion comes over what to do with the balanced end's cable screen.

1) Pseudo Balanced Solution

Modification of that described above as such:

UNBalanced Output Connector: mono jack plug
balanced Input Connector: stereo jack plug

Unbal Ground connected via resistor at unbal end, via screen to the Bal Ground. (Resistor can be bypassed by small capacitor to shunt RF for good measure).

Yep. That works. The idea is to maintain the cable screen, but to separate the unbalanced and balanced equipment grounds with the resistor which is sufficiently large to prevent significant ground current flow, while still maintaining a nominal ground voltage. The shunt capacitor improves the impedance for RF signals, and thus improves the screening performance, but it's not essential in most installations.

2) True Balanced Solution

UNBalanced Output Connector: mono jack plug
balanced Input Connector: stereo jack plug

Unbal Signal -> (via red wire) Bal Hot
Unbal Ground connected via resistor at unbal end, (via white wire) to the Bal Cold.

Yes, this will work as an impedance balanced output... but ONLY if that resistor has the exact same impedance as the balanced output driver.

The advantage of this solution is that the interference cancelling properties of the balanced input are maintained. The previous solution presents different hot and cold impedances, and so the interference rejection properties of the balanced input no longer work as intended. However, in most cases, this isn't important because there isn't much to cause interference anyway and you are working with line level signals which aren't very prone to interference. It's a bodge, but a workable one in most practical situations.

Screen connected to Bal Ground and disconnected and Unbal end.

This is never a good idea, because with the far end floating, the cable screen effectively acts like a radio aerial, injecting all sorts of RF rubbish straight into the balanced device's ground. Again, you can get away with it in many situations but it's not advised. Connecting the unbalanced end to the unbalanced device ground via a resistor of a 100 ohms or so (and shunted with a small capacitor if possible) is a far better solution. Exactly as before.

I have read descriptions of how and why each should work but I don't know what is the ideal method.

Option 2 is ideal, but difficult to do for most because of the unknown resistor value. Option 1 is easier and works perfectly well 95% of the time.

Until now I have used the first method but never bothered with shunt capacitor due to space inside the jack barrel being limited enough anyway.

Fair enough. The capacitor is the cake icing, and only really needed if you live under a BBC Radio 2 transmitter!

hugh

[ramthelinefeed]

Fair enough. The capacitor is the cake icing, and only really needed if you live under a BBC Radio 2 transmitter!

But in these mobile phone infested days,
surely everyone has the equivalent of a Radio 2 transmitter in their pocket!
I have to switch my phone off if I'm recording otherwise the amount of interference is atrocious.

[Pangloss]

Hugh - so you could actually combine both methods then?

That is to say connect both the balanced screen and cold to the unbalanced ground via resistors at the unbalanced end?

Or even, make a little connection box with the necessary low noise resistors and caps inside? Maybe a variable resistor for the impedance matching?

[ramthelinefeed]

And still my question remains:
why can I not buy pseudo-balanced leads in shops? :/

[Hugh Robjohns]

But in these mobile phone infested days,
surely everyone has the equivalent of a Radio 2 transmitter in their pocket!
I have to switch my phone off if I'm recording otherwise the amount of interference is atrocious.

The solution is to turn the phone off!

But yes, if phone interefence is an issue ensuring the integrity of the cable screen is important. However, phone interference doesn't only get into cables -- it can also get directly into the equipment (particularly in legacy products)... in which case screened cables won't help at all!

hugh

[Hugh Robjohns]

Hugh - so you could actually combine both methods then?

No. One method is properly impedance balanced and the other isn't.

That is to say connect both the balanced screen and cold to the unbalanced ground via resistors at the unbalanced end?

Ah... I see what you mean. These are two separate techniques that are trying to do different things. So they can be used independently.

The balanced input device's cable screen should ideally be connected to the unbalanced output's ground through a resistor to prevent the cable screen acting as an aerial -- and it can be made even more efficient by shunting the resistor with a small value capacitor as previosuly discussed.

Connecting the balanced input cold terminal to the unbalanced device's ground through a resistor is only sensible and worthwhile if you have the means of determining the appropriate value to provide true impedance balancing... which few people would be able to do. And if it is not perfectly matched there is absolutely no advantage over not bothering to install the resitor. It is then more sensible to connect the balanced device's cold input directly to the unbalanced device's ground.

Or even, make a little connection box with the necessary low noise resistors and caps inside? Maybe a variable resistor for the impedance matching?

Making a connection box to provide the pseudo-balanced interface is a good idea ad would allow you to use normal balanced cables from the box to the balanced input device, instead of having pseudo-balanced cables in the system which might get confused with proper balanced cables and cause problems elsewhere.

But there is absolutely no point in having a variable resistor for the cold side because you have no means of determining the correct value.

Precise impedance balancing is necessary for common-mode rejection of interference. But if the cable screen is adequate there won't be any significant interference in the first place. And even if there is, the fact that we are working with line level signals means that it is unlikely to become audible in most practical situations anyway.

Far more important here is the removal of a ground loop, and the simple pseudo-balanced cable approach is a very effective and practical solution. Trying to generate a properly impedance balanced output in this kind of situation is not trivial and not worth the bother.

Hugh

[Pangloss]

Hello again Hugh,

I think I remember in that balancing article you recently linked to (A question of balance) that the screen is useful for defence against capacitively induced electric fields (from power cables etc.?) but less so against electromagnetic interference.

Is that correct or did I dream that bit up?

[Hugh Robjohns]

And still my question remains:
why can I not buy pseudo-balanced leads in shops? :/

Good question. I have raised the issue with a couple of cable manufacturers in the past but none seem interested...

Hugh

[Pangloss]

Thanks again Hugh.

The reason I mentioned a variable resistor is that I thought it might be possible in some cases to tune it by ear. I.e. with the volume way up, tweak it until interference is minimal.

As you say though - not worth the bother.

I will continue smugly as before I reckon. (I might even make some music one of these days).

[Dunewar]

That box seems like a very good idea! I just might make one to conect my system with a proper balanced snake to my computer. Great idea!

But how is that different from the cleanbox?

[Dunewar]

Oh and one more : I'm having ground loop trouble connecting two balanced connections (Mbox2 Pro to SM pro audio IN5), probably because one of the two isn't properly grounded internally, I suppose.
Will a pseudo balanced cable work in this case?