455kHz IF Transformers

[Bill Bowden]

There are usually 4 RF transformers in a AM radio with colors of Red
for the
LO oscillator, Yellow for the mixer or 1st IF, White for the 2nd IF,
and Black
for the 3rd IF. I notice the impedance of these transformers ranges
from
50K : 500 for the 1st stage (yellow), 30K : 500 for the second
(white), and
20K : 5K for the 3rd (black) stage.

I think I understand the 500 ohm secondary impedance which matches
the
input impedance of the next stage, but I can't figure out why the
primary
impedance is different for the 3 stages?

The 5K secondary impedance of the black (3rd stage) is probably
higher
to produce a higher voltage for the detector.

What exactly does a primary impedance of 20K mean and how would
that be setup with a bipolar transistor?

-Bill

 

[[email protected]]

There are usually 4 RF transformers in a AM radio with colors of Red
for the
LO oscillator, Yellow for the mixer or 1st IF, White for the 2nd IF,
and Black

I haven't seen a radio built like that in decades. Are you referring
to a particular model you're restoring?

 

[Tony Williams]

There are usually 4 RF transformers in a AM radio with colors of
Red for the LO oscillator, Yellow for the mixer or 1st IF, White
for the 2nd IF, and Black for the 3rd IF. I notice the impedance
of these transformers ranges from 50K : 500 for the 1st stage
(yellow), 30K : 500 for the second (white), and 20K : 5K for the
3rd (black) stage.

I have a 1960 app note describing the design of an
early AM receiver using germanium transistors.
This was a two band receiver, 540-1640KHz and
155-280KHz.

They used a pair of 28K:800ohm transformers and a
detector transformer whose impedances were not
specified, merely described as 1.85:1 turns ratio.

28k:800ohm was the 470KHz output/input impedances
of the particular transistor used, (Mullard OC45,
running at about 1mAdc).

They did note that the output impedance of the mixer
stage in the 540-1640KHz band generally exceeded
35k at 470KHz, but they used the 28k:800 transformer
simply for economy. So that might explain your 50k
mixer output transformer.

The impedances around the detector are not specified,
but the OC45 is running at a higher current so this
might explain an apparent lower output impedance.

 

[Ancient_Hacker]

I think I understand the 500 ohm secondary impedance which matches
the
input impedance of the next stage, but I can't figure out why the
primary
impedance is different for the 3 stages?

Each stage runs at a different collector current, resulting in a
different optimum collector impedance. In those days they tried to
get maximum gain out of each stage, so they took extra care, even if
it required different transformers for each stage.

As it turns out, the exact impedances are not all that critical, as
power transfer only varies as the square-root of the mismatch, and
those old transistors had very loose specs to begin with.

 

[Robert Baer]

I haven't seen a radio built like that in decades. Are you referring
to a particular model you're restoring?

The "standard" (tube) configuration was known as the all-american
five, meaning 5 tubes used counting the rectifier for the B+.
4 RF transformers has been common for tube and transistor radios, but
the impedances were different as necessary to match lo-Z transistors.

 

[Bill Bowden]

Each stage runs at a different collector current, resulting in a
different optimum collector impedance. In those days they tried to
get maximum gain out of each stage, so they took extra care, even if
it required different transformers for each stage.

As it turns out, the exact impedances are not all that critical, as
power transfer only varies as the square-root of the mismatch, and
those old transistors had very loose specs to begin with.

I'm not sure how to calculate the output impedance of the transistor,
but I think it's related to the supply voltage and bias current. And I
see these same transformers used in a variety of radios running on
different battery voltages.

So, if the supply voltage is say 6 volts, the bias current for the 50K
transformer would be around 6/50000 = 120uA

Seems a little low and the transistor gain may be higher at greater
bias currents?

Suppose we use a higher bias current to obtain a higher transistor
gain. How would that effect operation?

-Bill