On 7 Jul 2003 20:02:40 -0700, (Peter B.) wrote:

clipped cause you wrote it and know what it says

Reading between the lines, I'm guessing that you've
extracted these numbers from a spec sheet and have
removed the parts about "at such and such frequencies".

Transformers (or any other comparably simple device)
cannot have a fixed phase shift in their passband.
The spec's you're seeing help to define their
response near the ends of the passband.

IOW, not to worry.


Chris Hornbeck,
guyville{at}aristotle{dot}net
question Authority

Peter B. wrote:
I'm looking to buy an 8ch mic splitter at a local audio shop. The guys
there were kind enough to let me know what was inside. I called the
transformer company up and requested a data sheet on the transformer.
Today I got the sheet in the mail. In searching around for information
on mic splitter transformers (and other transformers for that matter)
I'm noticing that there are both positive and negative values for
phase shift. What is this all about?

Sometimes the phase shift makes high frequencies lead. Sometimes it makes
low frequencies lead.

Generally the numbers on the data sheet don't apply unless you have exactly
the rated source and load impedances that they did the measurements with.

I don't know much about the subject but I'm guessing that this phase
shift shouldn't matter because it's so near the beginning of the
recording chain and common to all inputs. Nothing 'unshifted' is
destructively summing with what I'm putting in as long as all my mic
inputs are shifted by the same amount. Right or wrong? Partial credit?

Well, it's still group delay and it's still affecting the sound.

How much would it matter if multiple microphones are picking up the
same sonic events (through bleed) and some are transformer phase
shifted by say... -14 degrees and the other mics are on transformers
shifted by +1 degree? Will the summed result make a big difference or
will it be drasticlly overshadowed by the accoustical time delays
between mics in the room?

Yes, but it'll still be audible, even with one mike in mono. It's still
something altering the signal, and in many case the phase shift in the
transformer is actually a side effect of ringing or of frequency response
variations.

Now if all the mics were perfectly matched and could occupy the same
space at the same time picking up the same source equally what would
happen with different transformers?

They will all sound totally different, because they all color the sound
substantially in different ways.

For these examples lets assume that all properties of the transformers
are equal except phase shift.

That's totally impossible. Phase shift is only a symptom of more dramatic
differences.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

"Peter B." wrote in message
om...
Hello,
I'm looking to buy an 8ch mic splitter at a local audio shop. The guys
there were kind enough to let me know what was inside. I called the
transformer company up and requested a data sheet on the transformer.
Today I got the sheet in the mail. In searching around for information
on mic splitter transformers (and other transformers for that matter)
I'm noticing that there are both positive and negative values for
phase shift. What is this all about?


Welcome to the real world of electronics. Phase shift can be either positive
or negative. In general,its positve (leading) at low frequencies and
negative (lagging) at high frequencies. This is not a peculiarity of
transformers, but is in fact common to just about all pieces of analog audio
gear.

I don't know much about the subject but I'm guessing that this phase
shift shouldn't matter because it's so near the beginning of the
recording chain and common to all inputs.

That's a good thought, but its not that simple.

Nothing 'unshifted' is
destructively summing with what I'm putting in as long as all my mic
inputs are shifted by the same amount. Right or wrong? Partial credit?

Substantial partial credit for being mostly right.

Phase shift has its most audible effects under the conditions you mention -
when shifted and unshifted signals are summed. However phase shift, if
extreme enough can in and of itself be audible. One high quality transformer
probably won't do it, but not all applications involve just one high quality
transformer.

How much would it matter if multiple microphones are picking up the
same sonic events (through bleed) and some are transformer phase
shifted by say... -14 degrees and the other mics are on transformers
shifted by +1 degree?

At this point you could lose points for focusing on transformer phase shift
when it's very likely that acoustical phase shift will be a far more
significant source of problems. As you pointed out earlier, relative phase
shift is a problem, but you need to know that phase shift caused by
microphones being at various distances from the same source is a far more
audible and pervasive problem in the real world.

Consider a piano string, say the A string vibrating at 440 Hz. At 440 Hz, a
wavelength is about 30 inches. This corresponds to 360 degrees of phase
shift. So the phase shift at the relatively low frequency of 440 Hz due to
the speed of sound is about 12 degrees per inch. If you pick this sound up
with two microphones, one an inch closer than the other, there will be 12
degrees of phase shift. Look up the phase shift in your transformer at 440
Hz. I'll bet its far less than 12 degrees. Then consider your observation
that both mics pass through the same type of transformer so the relative
phase shift due to the transformer is zero. Note that there's nothing you
can do about phase shift due to the speed of sound except move the mics
around. Then consider that 440 Hz is a relatively low frequency, and that
phase shift due to the speed of sound goes up proportional to frequency. At
4400 Hz, where the ear is most sensitive, phase shift due to the speed of
sound is 10 times as much or 120 degrees per inch!

Will the summed result make a big difference or
will it be drastically overshadowed by the acoustical time delays
between mics in the room?

Your high score will be intact if you discern the right answer to this
question, which is "yes".

Now if all the mics were perfectly matched and could occupy the same
space at the same time picking up the same source equally what would
happen with different transformers?

The phase shift would be different, and there would be some cancellation of
otherwise identical signals due to phase shift. This cancellation would be
frequency-dependent and is likely to be damaging to sound quality because it
will change the timbre of the music.

For these examples lets assume that all properties of the transformers
are equal except phase shift.

Aside from the highly significant cancellation effects due to different
phase shifts, there are also some potentially audible effects due to phase
shift being applied to just a single signal. While the ear can't discern
phase shift directly at higher frequencies, say those above 500-1000 Hz, it
can discern phase shift at lower frequencies like 32 Hz. Virtually every
piece of analog audio equipment has phase shift at low frequencies, and we
use long chains of this equipment which add more and more phase shift at low
frequencies. Remember that transformers are hardly unique when they add
phase shift - just about all analog equipment does it, too - transformers or
not.

Welcome to the real world of electronics. Phase shift can be either positive
or negative. In general,its positve (leading) at low frequencies and
negative (lagging) at high frequencies.


How does this 'deviation from linear phase' in a transformer change
with the amplitude of the incoming signal? With a stronger low
frequency input signal does low end phase shift lead more than at
lower levels?


Thanks for the good posts everyone.
Peter

"Peter B." wrote in message
om
Welcome to the real world of electronics. Phase shift can be either
positive or negative. In general,its positive (leading) at low
frequencies and negative (lagging) at high frequencies.


How does this 'deviation from linear phase' in a transformer change
with the amplitude of the incoming signal?

Depends on the transformer.

With a stronger low
frequency input signal does low end phase shift lead more than at
lower levels?

Yes, if the transformer lacks sufficient power-handling capacity.

If the transformer lacks power handling capacity, its core will saturate,
decreasing mutual inductance. The decreased mutual inductance will
effectively increase the phase shift at low frequencies.