Is this really what you'd expect from an audio balun

[Bob F]

The schematic here seems completely wrong to me.
http://www.mcmelectronics.com/content/ProductData/Spec Sheets/50-7725.pdf

http://canada.newark.com/mcm-custom-audio/50-7725/audio-balun-rca-plug-rj45-shield/dp/96K0889

 

[Don Pearce]

The schematic here seems completely wrong to me.
http://www.mcmelectronics.com/content/ProductData/Spec Sheets/50-7725.pdf

http://canada.newark.com/mcm-custom-audio/50-7725/audio-balun-rca-plug-rj45-shield/dp/96K0889

Looks OK to me. It's configured as a series common mode stopper.
What's the problem?

d

 

[Bob F]

Looks OK to me. It's configured as a series common mode stopper.
What's the problem?

Will this design eliminate ground loop problems? My concept of baluns was that
they isolated the two ends by placing them on opposite windings of a
transformer. I am no expert, which is why I asked.

 

[Don Pearce]

Will this design eliminate ground loop problems? My concept of baluns was that
they isolated the two ends by placing them on opposite windings of a
transformer. I am no expert, which is why I asked.

That is a different thing. You need a transformer to kill a ground
loop entirely, but yes, provided there is enough inductance in the
coils this will still make a good job of getting rid of hum from a
ground loop.

The downside of the transformer method is that lower bass response is
usually somewhat compromised.

d

 

[Kevin McMurtrie]

Looks OK to me. It's configured as a series common mode stopper.
What's the problem?

Will this design eliminate ground loop problems? My concept of baluns was
that
they isolated the two ends by placing them on opposite windings of a
transformer. I am no expert, which is why I asked.[/QUOTE]

It will stop ground loop problems for the range of frequencies where the
transformers work. It's not clear what that is.

Transformers have a limited working range of frequencies. The ones you
posted have DC to MHz passthrough but limited common mode blocking. The
other wiring of isolators blocks DC to MHz common mode but only passes
through maybe 50Hz to 15 Khz; more or less depending on the quality.
Isolators can do weird things to the impedance too.

If this is pro-audio, the absolutely best fix is using balanced cables
with balanced connectors. The audio quality will be superior to any
other hack.

 

[Don Pearce]

If there's enough inductance to attenuate 60 (or 50) Hz from the power
line, why won't it do the same for similar frequencies in the desired
signal?

No, because of the way the windings are arranged their inductances
cancel each other out for the audio signal, but not for the induced
hum.

For any sort of well-designed and properly terminated transformer, the
lower 3dB point will be well below the frequency of any "musical note"
you'll ever want to pass through it. So, no, bass response won't be
compromised at all.

Effects strat becoming apparent well before you hit the 3dB point.

d

 

[Bob F]

No, because of the way the windings are arranged their inductances
cancel each other out for the audio signal, but not for the induced
hum.


Effects strat becoming apparent well before you hit the 3dB point.

d

 

[Bob F]

No, because of the way the windings are arranged their inductances
cancel each other out for the audio signal, but not for the induced
hum.

The windings being on a transformer arranged so that common mode noise cancels,
and the signal doesn't?

 

[William Sommerwerck]

I was going to get involved in this conversation, but have decided not to. All
I know is that the schematics don't make much sense (primarily because there
doesn't seem to be "proper" isolation between the bal and the un.

 

[Bob F]

I was going to get involved in this conversation, but have decided
not to. All I know is that the schematics don't make much sense
(primarily because there doesn't seem to be "proper" isolation
between the bal and the un.

Which was exactly why I posted the question. At the least, there is a DC(/LF)
path from one end to the other.

 

[William Sommerwerck]

I was going to get involved in this conversation, but have decided

Which was exactly why I posted the question. At the least, there
is a DC(/LF) path from one end to the other.

Which seems to kill the whole reason for baluns, does it not?

 

[Don Pearce]

The windings being on a transformer arranged so that common mode noise cancels,
and the signal doesn't?

That's it. The windings are wound in the same direction on the two
sides. That means that the forward and reverse signal currents on the
two windings are always in opposite directions magnetically. So the
signal doesn't see any net inductance.

d

 

[Don Pearce]

I was going to get involved in this conversation, but have decided not to. All
I know is that the schematics don't make much sense (primarily because there
doesn't seem to be "proper" isolation between the bal and the un.

Isolation is not implied in the function. As long as the signal on the
output is of equal amplitude and opposite phase on the two ports, the
job is done. This circuit satisfies that perfectly.

d

 

[Don Pearce]

Which seems to kill the whole reason for baluns, does it not?

Nope, nothing to do with it - unless you require performance down to
DC, which of course audio doesn't.

d

 

[Bob F]

That's it. The windings are wound in the same direction on the two
sides. That means that the forward and reverse signal currents on the
two windings are always in opposite directions magnetically. So the
signal doesn't see any net inductance.

Is this design going to avoid ground loop problems? Those signals would just be
on the one side, so would they be canceled?

 

[Don Pearce]

Is this design going to avoid ground loop problems? Those signals would just be
on the one side, so would they be canceled?

The ground loop is fixed by the inductance presenting a huge series
impedance to the hum current, which only passes along the ground wire.
There is no equal return current in the signal wire to cancel the
inductance.

d

 

[Don Pearce]

I'd sure like to see (and measure) those inductors. My "spidey sense"
(combined with the fact that they don't bother to spec the CMRR at 50 or
60 Hz.) tells me that they're probably way too physically small for that
to be the case.

Spidey sense is not always very useful.

Know anybody who can hear the difference *on actual program material*
between flat to 5 Hz and -3dB at 5 Hz? Because the little thumb-sized
not-at-all-special transformers I use in a homebrew groundloop killer
have that measured characteristic. It's just not hard to find decent
transformers for audio.

Isaac

Construction of these transformers doesn't have to follow the normal
rules. They don't need good linearity, so very high permeability
ferrite can be used for the core in order to get a high inductance
value. This is because the audio doesn't have to pass through the
ferrite, it bypasses it.

d

 

[William Sommerwerck]

Construction of these transformers doesn't have to follow the normal

rules. They don't need good linearity, so very high permeability
ferrite can be used for the core in order to get a high inductance
value. This is because the audio doesn't have to pass through the
ferrite, it bypasses it.

?????????????????????????????????????????

I know of no transformer type in which the signal "passes through" the core
material.

 

[Don Pearce]

?????????????????????????????????????????

I know of no transformer type in which the signal "passes through" the core
material.

Nit-picking the terminology? In a transformer, the core is used to
couple the signal from the primary to the secondary. In this sense the
signal "passes through" it. In the series balun the only signal that
interacts with the core is the unwanted common mode or single sided
signal. This is generally of such a low value that it has no chance of
causing sufficient field strength to cause non-linearity.

d

 

[Bob F]

Spidey sense is not always very useful.


Construction of these transformers doesn't have to follow the normal
rules. They don't need good linearity, so very high permeability
ferrite can be used for the core in order to get a high inductance
value. This is because the audio doesn't have to pass through the
ferrite, it bypasses it.

Doing a little research, I found the following page. It seems these baluns are
"current mode" baluns.

http://vk5ajl.com/projects/baluns.php