[sdiy] Transistor buffer stage with differential output

[Jim Addie]

First it's important to understand the purpose a balanced line 
system.  It's simply to increase immunity to noise picked up by the 
cable through inductive or capacitive coupling from a noise source. 
Using a shielded cable reduces capacitive coupling, so you're really 
just dealing with inductive coupling from a stray AC magnetic field. 
In many cases, there simply isn't a high level stray noise source to 
begin with, such as in the mountains of Chile for example.

Before I spent a lot of time and effort to establish a balanced 
system, I'd want to try out the mics connected straight unbalanced. 
Because the source impedance of the mics are already fairly low (1K), 
they create a sort of noise attenuator inherently, which might just 
be all you need.  There are thousands of mics connected unbalanced in 
the world in environments repeat with hum fields that work just fine. 
Rig the mics up and give them a try first.  I'll bet they are just 
fine the way they are.  I've used similar mics with 25 feet of cable 
attached, no noise problems.  Most camcorders are unbalanced, and 
people plug in long mic cords all the time with no problems.

Building a mic balancing circuit with transistors is not trivial. 
You have noise issues to deal with, as well as power consumption, 
distortion, and coupling capacitors.  A transformer designed for mic 
level is not very expensive, and would be my first choice - after 
just leaving things well enough alone and going for unbalanced. 
Second, I might consider just building a local preamp and running 
line level down the cables.  There are several self-contained 
microphone preamp chips around that are pretty quiet.

There is a point to balanced cable, but it is only an issue around 
hum fields and other cable, particularly power cables or 
data-carrying cables where there is a potential high level noise 
source.  Then, the real magic happens at the input side, where 
common-mode rejection takes place.  Not all balanced inputs are alike 
either, with common-mode rejection ratios being anywhere from 20dB in 
a poorly designed circuit (I did test one of these once!) to 70dB in 
a well designed one. All balanced input circuits, active or 
transformer, also have varying rejection ratios over frequency, with 
the high frequency ratios usually being much worse than low, and some 
non-existent at RF frequencies.

Just give it a try. I've used mic cartridges like this plugged 
directly into my Sony portable DAT recorder, which supplies them with 
power too, and they are just fine.

Jim

PS, if "VR-techniques" means "virtual reality", wouldn't you want to 
record binaurally rather that 4 channel surround?


>Hi,
>
>I'm planning a trip to the desert near the mountains on the border between
>Argentina and Chile. There I will help some students with their master
>thesis, to rebuild an old Indian city using modern VR-techniques. One thing
>I'm going to do is to record sounds in the environment, so the final show
>will be more alive than just images.
>
>My idea is to use four of these microphones:
>http://www.elfa.se/elfa/produkter/se/2013813.htm
>
>I will place them in each direction and record surround sound. Using this
>technique I can use all four microphones to record A B stereo, X-Y stereo,
>Blumlein and so on...
>
>The rather simple microphone I'm planning to use, uses up to +10V power
>supply and I'm going to go for +9V so I can use regular batteries.
>
>I also want to have balanced cables... Does anyone have any schematic or a
>suggestion how to make a good balanced output from these microphones? I
>know that some microphones have a transistor output instead of expensive
>transformers, but how do I do a good transistor buffer stage with
>differential output? Can anyone help me out?
>
>Best regards, and have pleasant weekend!
>
>/Niklas