Audio baluns for sound card input?

[John Woodgate]

I read in sci.electronics.design that J M Noeding <[email protected]>

remember when they started local radio in this area (like BRMB or Radio
City 199), they had no understanding for unbalanced equipment and
balanced telphone lines,we got a lot of complaints. Transformers cured
the problem. I don't understand what is so difficult about
transformers, you find them in allmost any sorts of modems

I hope you are not correct in that assertion. The local commercial
stations had to meet ITC requirements, and they were not crude.

 

[John Woodgate]

I read in sci.electronics.design that Michael A. Terrell

How does this remove the hum? one input of the op amp will be
grounded because it is an unbalanced source.

The differential stage still has common-mode rejection, and the hum is
common-mode. This technique is widely used for video on coax, less so
for audio.

 

[Pooh Bear]

I read in sci.electronics.design that Pooh Bear


But those have very poor CMRR for radiated RF! Just what is NOT wanted
in this case (if what the OP thinks is the cause is indeed the cause).

I doubt if the car stereo chip the previous poster suggested will be any
better than a standard audio grade op-amp in differential mode (
especially since you can use close tolerance Rs with the discrete solution
).

I'm unclear if the problem has been definitively resolved as being RF
though. Heard of plenty of similar scenarios where it's basically 50/60
Hz.

Graham

 

[Pooh Bear]

I don't understand what is so difficult about transformers, you find them
in allmost any sorts of modems

Transformers with *respectable audio performance* are neither especially
cheap nor common. Not much demand for them these days either.

Graham

 

[Pooh Bear]

I read in sci.electronics.design that Michael A. Terrell


The differential stage still has common-mode rejection, and the hum is
common-mode. This technique is widely used for video on coax, less so
for audio.

Funny you should mention that John. Back in 1989/90 I designed a simple
differential video buffer for EMI studios to solve just such a problem.
They used bucketloads of them. I'm not aware if there was an equivalent
'off the shelf' product ( certainly at a sensible price ) that could do the
job.

Graham

 

[Pooh Bear]

I don''t know what you wanted me to look at but I looked at the EPR11
schematic, which is for a 2-wire and ... no Y-cap.

Note my other post with link to
http://www.powerint.com/images/schematic/di43.gif from a quick
applications solutions page.

It's the Y cap from pri to sec that's the cause of the leakge current of
course, not actually the EMI filter.

Graham

 

[Pooh Bear]

It depends on the quality of the transformer. Where I used to work we
used a particular type/brand of these telephone transformers for
everything including hifi (roll-off above 20kHz). Also beware not to
saturate a transformer.

LF will do any saturating that needs to be done. ;-)

Graham

 

[Pooh Bear]

I also pointed out commercial "DI Boxes" sold in music stores at high prices.

No they don't ! Depends what you mean by high I suppose.

Graham

 

[John Woodgate]

I read in sci.electronics.design that Pooh Bear

Note my other post with link to
http://www.powerint.com/images/schematic/di43.gif from a quick
applications solutions page.

It's the Y cap from pri to sec that's the cause of the leakge current
of course, not actually the EMI filter.

Yes, our messages overlapped. I'm not up on SMPS and I wonder what that
cap actually does, and whether that's the best way to do whatever it
does.

 

[Pooh Bear]

Hello Folks,

When someone at our church tried to record audio into a laptop this
caused a serious racket if line powered. Probably this is due to the
laptop switch mode power supply. Only running the laptop off its battery
produces a decent recording. Is there a simple balun transformer that
you could recommend?

I know the Muxlab (Montreal) versions which have great frequency
response. However, these are rather bulky devices that are ok for a
fixed installation. But this one needs to be small and portable even if
it's not 100% "hifi".

Ok, of course I could make one. But if there is a smaller of-the-shelf
version that would be better.

Regards, Joerg

To cut a long stort short.... Can you afford $30 ? It's small, portable and
does what you want.

Pro audio guys call it a DI box.

http://www.behringer.com/DI20/index.cfm?lang=ENG


Graham

 

[Jim Gregory]

Fred Bloggs wrote

Run the laptop off the battery and run a small jumper from laptop case
ground to the power supply safety ground while it is plugged in. If the
noise is present then a small isolation transformer, in conjunction with
screened twisted pair mic cable, should improve the situation and will be
the best you can do. If no noise is present with the jumper, then you are
wasting your time, the power supply is corrupting the sound card circuitry
itself - nothing you can do except go after that power supply to increase
attenuation of high frequency components by cutting cable and interposing
small box with multiple LC lowpass chain and ferrite chokes.

Joerg said his mixer o/p is unbal'd and multi-distributes from the one o/p,
so no buffering DA is in use. Nor do we know source Z. Imagine it's about
50r. So could protect mixer o/p distrib star node from accidental shorts
with, say, a 100r - 150r
resistor in series with *each* send core, and add one spare protecting way
for
the next probable add-on, or for testing, while he's at it.
There isn't a strong TV Tx mast close by, is there? Worth asking, too.

Yes, I now agree the PSU must be the offender and should be replaced by a
behaving, smooth one - they can't *all* emit large pulses on their DC lines.
(Or use a fully charged, high-capacity accumulator - and thus we would never
have read this which appears to have landed on a helpful but unlikely
newsgroup
anyway!).
If he nevertheless tries to use a low inductance AF 10k:10k (or a 10k:2 x
5k secs for
series/parallel config) xformer for i/p isolation, it must be canned +
connected to source ground
and preferably have an Electrostatic screen wire/tag which also *should* be
tied to the screen of
unbal cable from the mixer - as it is a good ground. Even smallish ones can
handle +12dBU at 30Hz.
But I don't know how rare or expensive these are across the Pond.

We still don't know what levels come out of the mixer port. If average
operating level is still too high, add a 10k to 22k log preset across the
secondary,
with its wiper and low side to "send" to laptop line i/p jack.

 

[Pooh Bear]

I read in sci.electronics.design that Pooh Bear


Yes, our messages overlapped. I'm not up on SMPS and I wonder what that
cap actually does, and whether that's the best way to do whatever it
does.

I'm trying to remember myself now !

It's explained in the ANs somewhere. Every SMPS I've seen ( albeit 3 wire
ones ) has a Y cap there, so it seems to be accepted practice.

Graham

 

[Joerg]

Hello Michael,

I have several hundred transformers and most have poor frequency
response for this application. I also pointed out commercial "DI Boxes"
sold in music stores at high prices.

Essentially the passive DI boxes are like that Radio Shack solution,
probably much better in frequency response though. Money is, of course,
an issue at church.

If you ever need a decent quality audio xfmr check these out:

http://www.muxlab.com/products/ve_avd_analog_balun.html

Regards, Joerg

 

[Joerg]

Hello Jim,

Assuming your mixer and other apparata are earthed/grounded to a reputable
ground,
try connecting audio screen to mains-operated (grounded) destins *only* at
receiving ends, so dissing them (o/c) at mixer send end, ie, use cores only
which you say are commoned.
But keep the screen for the laptop feed as unbalanced and it can then be
experimentally isolated with a 10k:10k i/p transformer at *receiving* end.

Yes, some day I think we have to tear it all down and rig it up again,
mainly because the system grew over time. But for now we have to leave
it as is and some cable connections are behind walls. So at this point
we are going to try what works reasonably well without tearing the whole
thing apart. Some noise is tolerable because it is mainly voice (the
sermon) that is going to be recorded and not music.

Regards, Joerg

 

[Joerg]

Hello Jim,

Or why not just change the switched-mode(?) power supply to something really
repectable to activate the laptop?
From your description of its filthy waveforms superimposed, it sounds
verboten to deliver decent clean DC outputs.

That's an option and actually one I did contemplate here in the lab.
It's all brand name stuff but when I placed an inkjet/scanner combo
there it blew my mind. The noise that its little power brick made was so
bad that I had to give it its own outlet with some big #43 toroids in there.

Regards, Joerg

 

[Joerg]

Hello Graham,

From experience or are you conjecturing ?

From sad experience. The first laptops were all two wire and resonably
quiet. Ok, some noise on the AM radio but that was it. The "new and
improved" laptops all came with three wire and sometimes when I measure
subtle stuff in the lab I have to unplug the PS. Never had to do that
with the old Wangs and Compaqs.

Their PS has to crank more power though because these are "new and
improved systems", meaning you need 256MB RAM, 2GHz clock and stuff to
write 'hello world' these days. But the noise still increased
disproportionately.

Regards, Joerg

 

[Joerg]

Hello Ban,

..... Another
possibility is the use of a wireless link with line inputs. These would
eliminate the need for galvanic connection and solve the A/D conversion as
well.

That's what I always wanted to have. But there doesn't seem to be enough
of a market since you can't buy anything decent around here for wireless
audio xmit. Except the devices X10 offers but they are bulky, need a
power supply each etc. Not very practical, too much cabling to tangle up.

Regards, Joerg

 

[Joerg]

Hello John,

It sounds as if it is faulty. Does it do the same thing when used
elsewhere with a different signal source? If so, It IS faulty.

Maybe, but it's not my laptop so I'll have to ask. Some "modern" power
supplies seem weird to me. One in a new printer works but produces lots
of EMI and makes inconsistent hissing noises. Like a whining bearing.
Dell said that's "normal".

Regards, Joerg

 

[Rich Grise]

Hello Rich,


An audio transformer can cut down on common mode noise currents quite a
bit. Just hook an EMC current clamp onto the DC line from the power
supply into the laptop. Then look at the analyzer when you connect
anything else to the laptop. You can even see it on the scope and the
picture ain't pretty. This stuff propagates onto almost any cable that
connects to the laptop.

So, it's not so much that the power supply itself is providing the
noise, just that the ground is so long that it makes a nice high
impedance path for the common mode noise to excite?

Makes a lot of sense, when you put it that way. Transformers it is!


Thanks!
Rich

 

[Pooh Bear]

Hello John,


Maybe, but it's not my laptop so I'll have to ask. Some "modern" power
supplies seem weird to me. One in a new printer works but produces lots
of EMI and makes inconsistent hissing noises. Like a whining bearing.
Dell said that's "normal".

I got one of my protoype switchers to *hiss*. It seemed to happen as it
transitioned from non-continuous to continuous mode. I fitted a new core
and the noise went away IIRC. Probably due to indifferent assembly of the
transformer.

Graham