Oh me oh my. First most audio professionals do not consider 10K ohm as high
impedance. Inputs such as used on the older radios, Heath's, Collins and
the like have their input Z in the order of several megs. As to signal
levels there are three basic groups: mike level, line level, and high level
or speaker level. None are compatible with the other which says that one
should not be connected to the other one. I'll not dive into the topic of
source impedance vs. load impedance and the like.
Now, as to running a broadcast audio chain into the Pegasus, first a
broadcast audio chain will have a source impedance of 10 to 25 ohms which is
ultimately designed to operate into a load Z of 500 ohms or higher. This is
done such that when multiple loads are desired to be connected across a line
Z output such as the output of a broadcast chain, then the loads should be
of bridging impedance typically in the order of 10K ohms or so. This allows
up to 10 loads to be safely connected across one source with out loading or
affecting the source output level. In all cases we are dealing with line
level outputs connected to line level inputs.
Next issue, a broadcast chain output is line level usually +4 dBM to +8 dBM
nominal level reference 1 mW into 600 ohms. The Pegasus input mike level is
designed to operate typically in the -60 dBM to -40 dBM range. The point
being, one can not run line level signals into a mike input without great
risk of overdriving the first audio stage. Usually the gain control is
after the first stage and the adjustment of such will not serve to solve the
problem as related to first stage overdrive.
Solution: build an unbalanced T network with about 45 dB attenuation for
600 ohms input to 600 ohms output and terminate the output in 600 ohms an
then feed that direct to the mike input of the Pegasus. That is equivalent
of using a bridging input across a terminated output. Perfectly correct
technically. No matching transformer is needed and only 4 resistors
required. For the T network the input and output series R values are
calculated at 593 ohms and the shunt to ground is 6.7 ohms. Use 1/4 watt
values and the closest standard 5% value resistors will work just fine.
Alternatively: one can build a 3 resistor PI network designed to operate
between 600 ohms and 10K ohms and provide some 45 dB attenuation. In this
case use a 628 ohm input R to ground, a 217787 series R, and a 10480 shunt R
to ground on the output side. Feed this direct to the input of the Pegasus.
Values are calculated and the nearest standard value 1/4 w carbon R will
work just fine. Again, no transformer is needed.
As to transformers, a matching transformer is just that, one designed to
transfer maximum energy from one impedance to another. In the case of
feeding a line level signal to a mike input this is exactly what is NOT
needed. Some method of attenuation is required because of the difference in
levels between the two devices of 40 to 60 dB. If one believes that a
transformer is needed for ground loop isolation then a 600 ohm to 600 ohm
transformer should be used between the output of the broadcast chain and the
input of the resistive attenuator. This places the transformer at the
higher signal level and thus minimized the chance of hum introduction at the
transformer.
73
Bob, K4TAX
I've been attempting to run my broadcast audio chain into my Pegasus
with limited success. I'd been informed, variously, that the mike input
to the Ten Tec rig was low impedance - balanced (which is what I need),
high impedance unbalanced, low impedance unbalanced and high impedance
balanced :-). The manual I have is devoid of information and I have no
schematic.
So, I called Gary, N4CJX, at Ten Tec -- the "Peg tech". He says the
mike input is high impedance (about 10K) unbalanced -- which is the
usual amateur practice (strangely). He's also sending me a schematic
(yippee!).
Anyway, armed with that knowledge, I'll employ a matching transformer
and pad and all should be well -- unless I run into more RF incursions.
I just thought I'd pass along the information in case anyone else is
curious.
|