Line Drivers and Interfacing
Most balanced line drivers are one of a few variations of cross-
coupled mixed feedback circuits. You know, the ones where you can
plug in balanced, or unbalanced pin 2 or pin 3 grounded gear and are
supposed to allow fool-proof transformer-like interfacing. Two
problems - most use positive feedback at just under oscillation levels
which may mis-behave with reactive loads (like long cables). The
second problem is that this type of output stage generally reacts to
differences in the input's impedance balance and changes the source
impedance balance, which can mess up common mode rejection. In
fact, even if the next device may have a great common mode
rejection ratio spec like 80 dB across most of the audio spectrum, the
biggest requirement is that the source impedance also be balanced
or equal on both legs. For example, if the source impedance of one
leg is 100 ohms and the other is 105 ohms (5% resistors) the common
mode rejection could end up being 30 dB. But cross-coupled outputs
can amplify an impedance imbalance and make a bad thing worse.
The Mini takes a different approach, which uses that I/O switch on
the back panel. Output impedance remains constant and near ideally
balanced and about 30 ohms in all situations. As balanced outputs,
they will reach about +30 dBm, and in fact will drive 50 ohms or
headphones all day long.
The I/O switch and line driver circuit came about from a little email
challenge from George Massenburg, while discussing the issues and
problems of typical cross-coupled output circuits, the desirability of
an ideal transformer-like output, but avoiding the transformers. We
came up with a number of alternatives and this one used in the Mini
Massive was one of the best. The only downside was that it does
require the switch, which makes it less than brainless and automatic.
But the switch allows yet another benefit, which is the possibilty for
true consumer or semi-pro "-10 dBu" interfacing, and this becomes
the third position on the toggle. Yet another possible benefit is that
the switch may provide the means for a user to reduce headroom and
use premature clipping as an effect.
Most gear in most studios has balanced inputs and outputs. The
Mini Massive happliy accepts both balanced or unbalanced sources
and either XLR or phone plugs can be used - easy, no problem. If the
source is consumer level (or semi-pro) "-10 unbalanced" (probably
uses RCA phono jacks) then you may need a phono to phone plug
adapter, and you should select "-10 dBv" on the back panel to ensure
the best signal to noise performance. Most synths and electronic
instruments should also be used with the "-10 dBv" setting unless
the signal is particularly hot and you hear distortion. In that case try
the "+4 unbalanced" or "+4 balanced".
For most situations the default setting will be "+4 balanced" which
is that toggle in tthe middle setting. This setting has the best
headroom too, because it drives the signal on two legs so there is
another 6 dB of drive available, which in this case maxes out at a
respe30 dBm.
A small percentage of gear has unbalanced inputs and some mastering
studios are wired unbalanced (and some even remove input
transformers or extra circuitry used for balancing, towards a goal of
minimalist electronic path). For these rare situations the "+4
unbalanced" setting should prove ideal. So if the default "when in
doubt" setting for the toggle is "+4 balanced" and you patch an
unbalanced piece of gear after the Mini, then what happens is that
the basic level seems 6 dB too low. Then swith to "+4 unbalanced".
And then there is the transformer output option, which remains the
most forgiving and automatic method to interface given that the
ouput is floating and either leg can be grounded without a loss of
headroom and with less groundloop difficulties. Technicalities
aside, the transformer also provides a bit of fashionable transformer
warmth and richness or fatness. The switch associated with the
transformer option, then allows transformerless direct outputs, the
transformer, or a slightly exagerated transformer color labelled
vintage because it biases the transformer similarly to vintage British
console circuits. 3 sounds, your choice.
We may be to blame for suggesting that transformers were the
biggest factor for the desirable colors of vintage gear but you should
also be aware of potential downsides. Most importantly typical
transformers have frequency responces of 5 - 10 Hz to 50-100 kHz
and while one might assume that this should be inaudible, in reality,
it is accompanied with occasionally audible phase shift which might
diminish some tightness and impact or effortless treble extension.
Always a trade-off - warm and phat or tight and deep, smooth or
incredibly fast. The effects are generally subtle and depend both on
the how much exteme lows or highs are present in the sound and how
important accurate imaging is for this source. Most people are not
super-sensitive to phase and time smear issues and generally report
differences in imaging and depth. Actually, in the case of the Mini
Massive, simply adding the transformer was almost inaudible
because it is a pretty good transformer, and we had to exagerate its
characteristics to provide the expected audible difference and colors.
It is to be expected that the better the part or circuit the less that it will
affect the sound and more transparent it has to be. However, it is
generally easy to provide less than optimum drive conditions (if
transparent is considered optimum) for a transformer, or put another
way, easy to set up conditions where the transformer color becomes
subtle or significantly audible. Alternatively we could have designed
a less than ideal transformer with particular sonic character, but we
chose to use the same transformer that the Massive Passive uses to
maintain some of the same flavor and desirability and drive it with
similar characteristics, given the fan-base of the Massivo.
Which brings up another possible issue with transformers. Because
their characteristics depend on both the source and destination, there
may be conditions that cause unexpected results. In the case of an
output transformer, what it feeds may change its frequency response
slightly. Input transformers are generally much more of a wild card
that way, because a transformer is even more dependent on the
source, which may be anything - we don't know what you will plug
into it. Anyways for a typical line output transformer, we tend to
expect, that a heavily resistive load like 600 ohms may reduce both
the extreme highs and level slightly. A very long cable may reduce
highs. A very capacitive cable may affect the frequency response as
if one bell boosted a few dB at 30 or 50 kHz. It happens rarely but
about once a year we get a call reporting some unusual symptom
which ends up being caused by an unusually long cable or strange
piece of gear. This may be expected with any transformer coupled
gear. At least with the Mini Massive, it is easily solved by switching
to "transformerless".
The I/O toggle can be considered to be before the "transformer
option" toggle and does affect the drive to the transformer which
means, 95% of the time you'll still want the I/O toggle set to the
middle position "+4 balanced" for maximum headroom. The
exception would be with "-10 dBv" gear or instruments. However,
one doesn't need to worry about the 6 dB loss feeding unbalanced
gear with the transformer in use.
25