Hi Turk 182,
When you connect the centre tap of the transformer to the ground of the signal source, you rely on the matching between the two primaries for common-mode rejection. If they are perfectly equal, for common-mode signals, the magnetic field caused by the current through one primary cancels the equal but opposite field from the other primary. If they are not entirely equal, they only partly cancel.
With the centre tap not connected, common-mode signals don't cause any primary current to flow in the first place, so you don't need to worry about how well something cancels. Hence my preference for not connecting it.
However, I haven't taken the parasitic capacitance from primary to secondary into consideration. Suppose that the transformer has a big parasitic capacitance from primary to secondary, that the signal source has a high common-mode output impedance, that a common-mode interfering current somehow enters the balanced signal path and that the capacitance from primary to secondary is actually the main crosstalk path to the unbalanced output. In that case, you may be right, because grounding the centre tap then reduces the common-mode impedance and effectively shorts the common-mode interference.
I don't think this is applicable, though. The transformer has some sort of shield, presumably a capacitive shield (Faraday shield) between primaries and secondaries that eliminates the capacitive coupling. Besides, I doubt if the DAC has a high common-mode output impedance, although I'm only guessing that. So my preference would still be not to connect pins 2-3 of the primaries to pin 1 of the XLR.
Best regards,
Marcel
When you connect the centre tap of the transformer to the ground of the signal source, you rely on the matching between the two primaries for common-mode rejection. If they are perfectly equal, for common-mode signals, the magnetic field caused by the current through one primary cancels the equal but opposite field from the other primary. If they are not entirely equal, they only partly cancel.
With the centre tap not connected, common-mode signals don't cause any primary current to flow in the first place, so you don't need to worry about how well something cancels. Hence my preference for not connecting it.
However, I haven't taken the parasitic capacitance from primary to secondary into consideration. Suppose that the transformer has a big parasitic capacitance from primary to secondary, that the signal source has a high common-mode output impedance, that a common-mode interfering current somehow enters the balanced signal path and that the capacitance from primary to secondary is actually the main crosstalk path to the unbalanced output. In that case, you may be right, because grounding the centre tap then reduces the common-mode impedance and effectively shorts the common-mode interference.
I don't think this is applicable, though. The transformer has some sort of shield, presumably a capacitive shield (Faraday shield) between primaries and secondaries that eliminates the capacitive coupling. Besides, I doubt if the DAC has a high common-mode output impedance, although I'm only guessing that. So my preference would still be not to connect pins 2-3 of the primaries to pin 1 of the XLR.
Best regards,
Marcel
Last edited:
this sort of reaffirms my level/lack of understanding.
with respect to the grounding the case and x-former is it better to link the ground to the source (input) or load(output)?
in my use of repeat coils primarily to resolve ground loops in touring PA and commercial installs i may have been misguided in the practice of grounding similar balanced to unbalanced circuits.
apology in advance if i seem to be thread jacking but i just want to insure correctness of information which i hope is useful to the OP.
with respect to the grounding the case and x-former is it better to link the ground to the source (input) or load(output)?
in my use of repeat coils primarily to resolve ground loops in touring PA and commercial installs i may have been misguided in the practice of grounding similar balanced to unbalanced circuits.
apology in advance if i seem to be thread jacking but i just want to insure correctness of information which i hope is useful to the OP.
Last edited:
Hi Turk 182,
Practice is usually more reliable than theory. You wouldn't believe how long a group of experienced electronics engineers can talk about optimal grounding strategies before finally choosing a solution that seems optimal in theory, but doesn't work well in reality (-;
If you look at repeat coil application schematics from Jensen, for example on the first page of the JT-11SSP-7MPC datasheet, they also typically leave the centre tap floating. Is it your experience that this doesn't work well? Or is it just your experience that the solution with the centre tap connected works well enough?
Regards,
Marcel
Practice is usually more reliable than theory. You wouldn't believe how long a group of experienced electronics engineers can talk about optimal grounding strategies before finally choosing a solution that seems optimal in theory, but doesn't work well in reality (-;
If you look at repeat coil application schematics from Jensen, for example on the first page of the JT-11SSP-7MPC datasheet, they also typically leave the centre tap floating. Is it your experience that this doesn't work well? Or is it just your experience that the solution with the centre tap connected works well enough?
Regards,
Marcel
Hi,
Ground is ground and for the balanced XLR input
it should always be attached to the centre tap.
There is no need to connect the input to ground at
all, whist the output has to referenced to ground.
rgds, sreten.
Ground is ground and for the balanced XLR input
it should always be attached to the centre tap.
There is no need to connect the input to ground at
all, whist the output has to referenced to ground.
rgds, sreten.
With an unbalanced output, I wouldn't want to have any interfering voltage between the Faraday shield and the secondaries, so the Faraday shield has to link to the ground of the unbalanced output (if the grounds are separated (because of ground loop issues)). I would also tend to connect the enclosure to the ground of the unbalanced output, just to have better RF shielding at the output.
Why? Why not leave the centre tap floating, considering that the signal voltage is between pins 2 and 3 of the XLR?
Marcel
as a standard practice it seems to help signal to noise ratio as well as decrease distortion(subtle difference in signal quality)
most touring tech's/soundguy's have Pin 1 lift adapters for the express purpose of resolving ground loops but these are a blessing and a curse misuse of these can result in causing system instability(oscillation)in those cases capacitive (AC coupling) of signal grounds is needed.
which brings me 'round to sreten's comment "ground is ground" sorry if grounding was that easy we collectively wouldn't all be losing brain cells trying to resolve grounding issues and the noise by-products that can result from just "grounding it" so i'll wait to see what your answer to Marcel's question in post #26 is.
as a standard practice it seems to help signal to noise ratio as well as decrease distortion(subtle difference in signal quality)
most touring tech's/soundguy's have Pin 1 lift adapters for the express purpose of resolving ground loops but these are a blessing and a curse misuse of these can result in causing system instability(oscillation)in those cases capacitive (AC coupling) of signal grounds is needed.
which brings me 'round to sreten's comment "ground is ground" sorry if grounding was that easy we collectively wouldn't all be losing brain cells trying to resolve grounding issues and the noise by-products that can result from just "grounding it" so i'll wait to see what your answer to Marcel's question in post #26 is.
None of the transformers in these application notes even have a centre tap (or at least they are not drawn in the schematics in the application notes).
Are you referring to the practice of connecting pin 1 to the centre of the primary or of connecting pin 1 to the enclosure?
In the first case, is this with or without Faraday shields in the transformer?
In the second case, I couldn't agree with you more. Connecting pin 1 to the enclosure when there is no need to break a ground loop has the advantage that you get one continuous, uninterrupted shield from the source through the cable to the transformer box through the next cable to the load. If there is a need to break the connection, it is indeed good practice to provide a capacitive connection for RF shielding (and to prevent excessive ground inductance that might cause instability).
Connecting the centre tap to ground on an input transformer is HORRIBLE practise as it basically drops the common mode impedance to near zero.
Now consider that for good CMRR the recever common mode impedance should be as high as possible to minimise the effect of imbalance in the wiring and source impedances and the full awfulness of the idea becomes apparent.
What makes transformers so useful is that they have very high common mode impedance, by grounding the centre tap you loose all of that.
Remember 'balanced' refers to the two signal legs having the same impedance to ground, it says nothing about symetrical voltages or anything of that nature, think of a balanced interface pair as a wheatstone bridge to see how all this really works.
Best practise is to tie pin one directly to chassis as close to the connector as possible, connect the transformer across pins 2&3 and connect the internal electronics reference plane to chassis at one point (If the transformer has an interwinding screen this should probably go to chassis as close to the input as possible).
This keeps the current in the screen away from the electronics and makes the screen an extension of the boxes it links.
There is a word for a screen lifted at one end, it is aerial.
73 Dan.
Now consider that for good CMRR the recever common mode impedance should be as high as possible to minimise the effect of imbalance in the wiring and source impedances and the full awfulness of the idea becomes apparent.
What makes transformers so useful is that they have very high common mode impedance, by grounding the centre tap you loose all of that.
Remember 'balanced' refers to the two signal legs having the same impedance to ground, it says nothing about symetrical voltages or anything of that nature, think of a balanced interface pair as a wheatstone bridge to see how all this really works.
Best practise is to tie pin one directly to chassis as close to the connector as possible, connect the transformer across pins 2&3 and connect the internal electronics reference plane to chassis at one point (If the transformer has an interwinding screen this should probably go to chassis as close to the input as possible).
This keeps the current in the screen away from the electronics and makes the screen an extension of the boxes it links.
There is a word for a screen lifted at one end, it is aerial.
73 Dan.
probably old practice from outdated TV repairman's book
thank you for clearing up that misconception
well i've obviously done things wrong in the past.
which leaves me wondering why despite it being wrong doing so did produce a positive result?
i wish i currently had an assortment of x-formers to replicate what i've done,all the balanced stuff i have kicking around at the moment is op-amp based.
i really need to sort out my knowledge base on this subject before i upgrade my 24/4 snake from passive split to x-former isolated for the monitor tail.
can anyone help me clear up some potential erroneous info: are primaries of balanced input transformer like the Jensen JT 11 not center tapped to ground internally?
which leaves me wondering why despite it being wrong doing so did produce a positive result?
i wish i currently had an assortment of x-formers to replicate what i've done,all the balanced stuff i have kicking around at the moment is op-amp based.
i really need to sort out my knowledge base on this subject before i upgrade my 24/4 snake from passive split to x-former isolated for the monitor tail.
can anyone help me clear up some potential erroneous info: are primaries of balanced input transformer like the Jensen JT 11 not center tapped to ground internally?
Maybe your transformers had well-matched primaries and the common-mode signal and/or hum level at their input was small, in which case connecting the centre tap would not do much harm.
Well, I'm confused 🙂
I have contacted the manufacturer for my transformers to find out if there is a faraday shield or some such in it. All it has is a drain wire connected directly to the iron of the transformer and 8 terminals is what I see.
Once I get more info, maybe we can figure out what is best for my needs.
Thanks to everyone for responding. I learned allot, but maybe trial and error of all configurations will get me to the best conclusion.
I have contacted the manufacturer for my transformers to find out if there is a faraday shield or some such in it. All it has is a drain wire connected directly to the iron of the transformer and 8 terminals is what I see.
Once I get more info, maybe we can figure out what is best for my needs.
Thanks to everyone for responding. I learned allot, but maybe trial and error of all configurations will get me to the best conclusion.
Last edited:
Ok, I have got word that my transformers have no shield, just space, and a drain wire from the core of the transformer.
I was also informed that the grounding is pretty much what will work best in the situation. The manufacturer also recommends running the balanced input ground to the center tap of primary side.
So, it looks like I have some playing around to do.
I was also informed that the grounding is pretty much what will work best in the situation. The manufacturer also recommends running the balanced input ground to the center tap of primary side.
So, it looks like I have some playing around to do.
Got a little playing done today.
I could not really tell much difference in sound running the input ground to center tap or float.
Will try running input ground grounded close to input jack next.
I could not really tell much difference in sound running the input ground to center tap or float.
Will try running input ground grounded close to input jack next.
sorry for confounding the whole subject.
i've just spent some time rereading Bill Whitlock's white papers on the Jensen website.
i'm also waiting for a reply from an old colleague as to the what and why of the methods we used.
the rack panels we were assembling where receiving long lines close to 1000 feet i'm not sure if that was the reason or quite what so hopefully i'll hear from my old friend soon perhaps he can shed some light on this or set me straight if i'm wrong.
i've just spent some time rereading Bill Whitlock's white papers on the Jensen website.
i'm also waiting for a reply from an old colleague as to the what and why of the methods we used.
the rack panels we were assembling where receiving long lines close to 1000 feet i'm not sure if that was the reason or quite what so hopefully i'll hear from my old friend soon perhaps he can shed some light on this or set me straight if i'm wrong.
No problem. I believe this is something that needed to be discussed as there is just no real solid info about it out there that I can find.
I added some switches to the enclosure so I could toggle all possible options, and the way that sounds best is the attachment in post #10 of this thread. This provided the best transparency for my system. I must also add that it sounds incredible. These transformers are crazy good, and very expensive. However, I do not regret purchasing them at all. No detriment at all to the sound, and only improvements were realized.
This configuration may not translate for every system and need. Here, it works like a freakin charm. 🙂
I added some switches to the enclosure so I could toggle all possible options, and the way that sounds best is the attachment in post #10 of this thread. This provided the best transparency for my system. I must also add that it sounds incredible. These transformers are crazy good, and very expensive. However, I do not regret purchasing them at all. No detriment at all to the sound, and only improvements were realized.
This configuration may not translate for every system and need. Here, it works like a freakin charm. 🙂
Last edited:
- Status
- Not open for further replies.