the MMIC at 50 mA due to the very low DC drop in that
choke. Even though I am covering the same bandwidth with
the presented signal generator, that method would not
work. The reason: bandwidth ratio.
The higher speed generator has a ratio of 1.15:1, and
one choke would present a constant load over that range.
My generator has a ratio of 500:1 (0.3-150 MHz). No single
choke could even come close to presenting a constant load
over that kind of range. So, I am committed to a resistive
load for consistency over the complete range.
Doing some math, I determined that the optimum value
of approximately 392 ohms is as low as I would want to go.
Carrying this further, 392 ohms/50 mA = 19.6V and adding
another 4. 4 volts across the MMIC results in the necessity
for a 24 VDC supply. Wasteful, yes, but that is one of the
prices you pay for a wide bandwith ratio amplifier. The
manufacturer highly recommends that you do not use a
constant current supply in this circuit.
Figure 2 is the control circuit and front panel controls
which are pretty simple and straightforward. IC1a drives the
MC1648 AGC pin 5 directly for AM modulation. Being a
follower type of configuration, its amplitude will remain
constant despite changing loads due to the AGC section of
the band switch presenting different loads depending on its
location. Diodes D1 and D2 clamp the input signal circuit
for a maximum signal level of 1.4 volts P-P. This is for the
protection of over-driving P5 of the MC1648.
R1 limits the current during periods of heavy clamping.
The external AM signal is fed into the mod input jack to P1
level control and through modulation switch S2 when in the
AM position. For FM modulation, the signal is brought in
through the same path other than the S2 switch position.
From there, it goes into IC1b’s summing point. When no
modulation is desired, S2 is set at the CW position
(Continuous Wave) and mod input capacitors C2 and C7
are grounded to keep these lines quiet.
IC1b has several functions: It sums all the voltage inputs
of the coarse tune control, the fine-tune control, and the FM
mod input (when used). Op-amp summers are nice for this
type of adding as there is absolutely no interaction between
the inputs feeding it. As can be seen, the tuning controls are
well filtered to keep this line as quiet as possible. The output
of IC1b requires a 100 ohm resistor R8 in series with it due
to the necessity of driving a large filter capacitance when
entering the Vt connection on the RF board. Op-amps can
become unstable, driving large cap loads; R8 is the cure for
this. Since the tuning span Vt has to cover a range of - 6.1
volts to +0.6 volts, an offset in the op-amp output of +0.6V
is needed for its starting point. This is accomplished with R9,
R10, R11, and D3 at its positive input.
D3 provides a convenient source of 0.6V, but also
performs one other function. It has a PN junction drop of 2
mV/degree C which is the exact opposite to the tank circuit
varactor diode Vd1, Vd2. This helps to stabilize the oscillator
frequency in that respect (ambient temperature) by putting
a slight shift in Vt vs. temperature. It’s not perfect, but helps.
The power supply shown in Figure 3 is pretty
straightforward and does not need much explanation. It
might look like overkill, but was the best I could come up
with considering the variety of voltage sources required.
When it comes to op-amp design, I am a big fan of split
supplies. The 5V supply has the capability of supplying
much more current than what is needed for the RF deck,
but I beefed it up to cover the counter and prescaler
options. It could also be used to supply a rear panel jack for
powering add-on outboard circuits. Also, I split up the 5V
through two different regulators to minimize interference on
the supply lines feeding the analog portion (RF deck) and
the digital portion (counter, etc.).
Before I get too far into this section, I will not go into a
lot of details in all aspects of construction as it would take
too much magazine space to discuss everything involved.
Fear not, though, for those that desire to build this
generator, I have a huge packet of info that I can email to
you. This will include full screen pictures, actual size drilling
templates, detailed layouts, construction tips, some artwork,
and lots of technical data.
There are just two basic subassemblies that will be
installed in this chassis: the RF deck and the power supply
deck. The control circuitry is so minimal, I built it up on a
small printed circuit board (PCB) and screwed it down to
the power supply board. This simplifies assembly and testing
before each unit is permanently installed.
The RF deck shown in Figure 4 was built on a 2” x 2-
1/2” single-sided PCB using typical RF prototype
construction. The MC1648 and R1, Q1, Q2, and Q3 are the
only components mounted on the laminate side of the
board. Be sure to use a socket for the IC. Since only eight
pins are used in this 14-pin chip, seven pins can be removed
before installing it. These are conveniently located as every
other pin and makes soldering underside components a lot
easier. After drilling all the lead through holes, all but the
ground lead holes are reamed out with a larger drill of
about 5/32” to give clearance from the lead to copper foil.
The GALI- 55 will require two small islands for input and
output connections. The SMV1404 varactor is quite small
and despite SMD soldering, I found an easy way to install
I first cut a piece of plastic laminate (formica, etc.) to
3/8” x 1/4” and then tack the varactor to the laminate with
a spot of super glue. Make sure the varactor is positioned
laying on its back with its “feet” pointing up as this makes
final soldering easier. This assembly is one of the last
components I install, and again a dab of glue helps to
adhere it to the PCB. When all leads that go to this device
are soldered to their proper nodes, cut the free end to the
exact length; then, a quick tack solder to the chip pins
June 2014 29