OK, I'm still struggling with implementing balanced connections in a home audio environment.
I thought that one big benefit of using line level transformers in the home audio environment was to get galvanic isolation between input and output. Of course, there is also the benefit of improved CMRR of balanced interconnects.
Yet if I look at the balanced/unbalanced Bible, Jensen AN-003, the input and output grounds of the various examples therein are all interconnected. Is this because this is a pro audio application and the various devices being connected will have individual ground lift switches?
The reason I am struggling with this is that I would like to make a passive preamp with a line level transformer it in that will have both balanced and unbalanced outputs, and I am not sure how to wire the output side. Is it typical practice to add a switch to ground the negative side of the unbalanced connection when it is being used?
I thought that one big benefit of using line level transformers in the home audio environment was to get galvanic isolation between input and output. Of course, there is also the benefit of improved CMRR of balanced interconnects.
Yet if I look at the balanced/unbalanced Bible, Jensen AN-003, the input and output grounds of the various examples therein are all interconnected. Is this because this is a pro audio application and the various devices being connected will have individual ground lift switches?
The reason I am struggling with this is that I would like to make a passive preamp with a line level transformer it in that will have both balanced and unbalanced outputs, and I am not sure how to wire the output side. Is it typical practice to add a switch to ground the negative side of the unbalanced connection when it is being used?
The easy explanation would be that the cable shield carries all the possible currents flowing between the enclosures involved.
This would mean that, in a truly balanced implementation, input ground and output ground are the same potential - and both noninverting and inverting signal lines are isolated from this shield potential (via transformers or other means). A ground lift is not needed, although sometimes implemented due to compatibility reasons.
The application notes by Rane, particularly notes 110 ("Sound System Interconnection"), 151 ("Grounding and Shielding Audio Devices") and 165 ("Pin 1 Revisited"), are a recommended read up on standards, their implementation in the pro audio industry, and their reasoning.
Your best option for switching between balanced and unbalanced could be to simply ignore the inverting line in the unbalanced case. But as you've got the option, going fully balanced would be my recommendation.
This would mean that, in a truly balanced implementation, input ground and output ground are the same potential - and both noninverting and inverting signal lines are isolated from this shield potential (via transformers or other means). A ground lift is not needed, although sometimes implemented due to compatibility reasons.
The application notes by Rane, particularly notes 110 ("Sound System Interconnection"), 151 ("Grounding and Shielding Audio Devices") and 165 ("Pin 1 Revisited"), are a recommended read up on standards, their implementation in the pro audio industry, and their reasoning.
Your best option for switching between balanced and unbalanced could be to simply ignore the inverting line in the unbalanced case. But as you've got the option, going fully balanced would be my recommendation.
i once gave this cabling thing some thought and read that rane note ... then i got rid of all unbalanced connections in my system
i basically followed the philosophy from the rane note - go digital if you can, if you cant go digital go balanced.
now there is a total of 5 digital links in my system, 2 analog balanced links and 0 unbalanced analog ones.
actually of the 5 digital links 3 are balanced ( 110 ohm XLR ) hehe ... i prefer optical though
as far as grounding goes for balanced analog signal ( because we dont care about unbalanced - remember ? ) i believe the idea was that the shield should be tied to chassis ground on one side and not connected to anything on the other. alternatively neutrik implemented an XLR connector with a capacitively coupled ground which accomplishes the same thing - short circuit for RF and open circuit for DC.
now you might ask - if there is no DC connection for ground - where does the reference level come from ? well, i have no idea i guess the receiving end could use its own signal ground ? there is no need for ground really because you have both + and ( - ) so all that is really needed is a "shield" which only needs to be tied to anything at one end ...
on the other hand perhaps you could do a cable with something like a 1 kohm resistor in parallel with a capacitor ... then it would be some kind of a ground and a full functional shield at the same time.
by the way when i said the ground should be tied on one side i meant the source side. i guess this is one of those things that make XLR connector so special - its two sides are different and always in the right place. so if you tie the shield to chassis ground at the source side then the RF picked up by the shield will have a sink to go into but not reach the "destination" side of the cable.
of course THE problem is that your source, destination and cable are not made by the same people so even if what i said made any sense it doesn't mean you can count on stuff to be implemented that way ...
AAAAARGH ! ! !
so i guess your question is - what is the convention ? i have no idea ... but i think that Rane note discusses that ...
honestly the whole thing is too complicated for my brain, especially at 3.45 am ...
i basically followed the philosophy from the rane note - go digital if you can, if you cant go digital go balanced.
now there is a total of 5 digital links in my system, 2 analog balanced links and 0 unbalanced analog ones.
actually of the 5 digital links 3 are balanced ( 110 ohm XLR ) hehe ... i prefer optical though
as far as grounding goes for balanced analog signal ( because we dont care about unbalanced - remember ? ) i believe the idea was that the shield should be tied to chassis ground on one side and not connected to anything on the other. alternatively neutrik implemented an XLR connector with a capacitively coupled ground which accomplishes the same thing - short circuit for RF and open circuit for DC.
now you might ask - if there is no DC connection for ground - where does the reference level come from ? well, i have no idea
i guess the receiving end could use its own signal ground ? there is no need for ground really because you have both + and ( - ) so all that is really needed is a "shield" which only needs to be tied to anything at one end ...on the other hand perhaps you could do a cable with something like a 1 kohm resistor in parallel with a capacitor ... then it would be some kind of a ground and a full functional shield at the same time.
by the way when i said the ground should be tied on one side i meant the source side. i guess this is one of those things that make XLR connector so special - its two sides are different and always in the right place. so if you tie the shield to chassis ground at the source side then the RF picked up by the shield will have a sink to go into but not reach the "destination" side of the cable.
of course THE problem is that your source, destination and cable are not made by the same people so even if what i said made any sense it doesn't mean you can count on stuff to be implemented that way ...
AAAAARGH ! ! !
so i guess your question is - what is the convention ? i have no idea ... but i think that Rane note discusses that ...
honestly the whole thing is too complicated for my brain, especially at 3.45 am ...
Thanks for that contribution, Vasyachkin!
I also learned a lot just from those notes.
I would like to add that it really makes a difference what distances we are talking about.
Connecting the cable shield to pin one of the XLR connector would be the right attempt - on both sides, if you can.
A metal enclosure would have to be powered three-prong, a non-conductive enclosure would only require two-prong power connection. The connection between XLR pin one and enclosure (third prong: safety conductor) would then depend on how the balanced input is implemented.
A truely symmetrical transformer input and output wouldn't require a shield connection for signal conduction, so a ground lift (on one side) is feasible but not required.
For a really long disctance, say, from a stage to a FOH mixing desk in a concert hall, a ground lift would really be recommended, though. The potential difference between the safety conductors in the power outlets (and thus metal enclosures of gear) can be fairly large. That's why professional gear is best fitted with transformer in- and output.
BTW, this goes for analog and digital, as we're talking electricity (as such).
Cheers,
Sebastian.
I also learned a lot just from those notes.
I would like to add that it really makes a difference what distances we are talking about.
Connecting the cable shield to pin one of the XLR connector would be the right attempt - on both sides, if you can.
A metal enclosure would have to be powered three-prong, a non-conductive enclosure would only require two-prong power connection. The connection between XLR pin one and enclosure (third prong: safety conductor) would then depend on how the balanced input is implemented.
A truely symmetrical transformer input and output wouldn't require a shield connection for signal conduction, so a ground lift (on one side) is feasible but not required.
For a really long disctance, say, from a stage to a FOH mixing desk in a concert hall, a ground lift would really be recommended, though. The potential difference between the safety conductors in the power outlets (and thus metal enclosures of gear) can be fairly large. That's why professional gear is best fitted with transformer in- and output.
BTW, this goes for analog and digital, as we're talking electricity (as such).
Cheers,
Sebastian.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.