The SA612 transmit mixer output of puny amplitude and dirty spectrum needs amplification and filtering before driving the Class C final amplifier (Q6). A two-resonator filter, and two transistors (Q4 and Q5) perform this task. This note shows oscilloscope waveforms, and describes the linear amplifiers between mixer and final amp.

The mixer output drives two cascaded L-C bandpass filters that reject all but 40M signals. The filter's output is a clean 7MHz. sinewave, at the base of Q4. I have not included an oscilloscope trace for the following reason: Placing a 'scope probe here de-tunes T3. The probes' capacitance is enough to change T3's resonance to a lower frequency. This has the effect of decreasing the apparent amplitude. If you monitor the rig's power output, when you apply the probe, you'll see a marked decrease. You can also monitor DC current drawn from the +12v supply (while transmitting) and see the same effect.

However, at Q4's emitter, this "loading" effect is not observed. Q4 is called a "common-emitter" amplifier. It has a voltage gain of about one. So you might think that its not really an amplifier since its output is no bigger than its input. But Q4 has significant current gain. Another way of looking at it: as an amplifier, it has high input resistance, and low output resistance. In any case, it does have power gain. The oscilloscope loading at the emitter isn't noticeable because this is a much lower impedance point than the base - so the probe's capacitance has less loading effect.

The emitter of Q4 (red) drives a 500-ohm trimpot. The trimpot in turn drives Q5's base (green). Q5 is a common-emitter linear amplifier, having both current gain and voltage gain. Its emitter voltage is almost as large as the base voltage. And the emitter is always about 0.6 volts lower than the base, so its running linearly, or nearly so.

Why are the base and emitter voltages not sinusoidal? My scope is likely picking up some radiated junk from the final amp. In the vicinity of such large signals, its sometimes wise not to trust exactly what you see on a 'scope. A short ground connection can often help here. The 5" ground-clip produced a much grundgier waveshape due to ground-loop pickup. Traces shown are with a 1/2" probe groundpath.

The peak-to-peak emitter voltage (blue) of about 0.5v p-p appears across R27(10 ohm). This would translate to an AC current of 50mA p-p. But now let's find the average emitter current of Q5 (Ie):

This 23.8mA standing DC current almost supports the 25ma peak current of the AC waveform. So Q5 operates linearly for nearly all of the AC cycle. Dave has designed Q5 well. It is operating at a point very close to current saturation, and very close to voltage saturation at the collector. See that Q5's collector voltage (red) swings down very close to its base voltage of 2.4v. Operating like this extracts maximum power available from Q5. 

Q5's collector load is the transformer T4. T4 translates the very small base impedance of Q6 to a much higher impedance at Q5 collector. T4's turns ratio is 8t primary and 1t secondary. This should also reflect in the voltage ratio between Q5's collector voltage and Q6's base voltage (which is about 21:2 from the scope-trace above).  With a 50 ma. p-p collector current, and 21v p-p collector voltage, Q5 is seeing a load of about 420 ohms. Power out is about 130mW RMS.

You might wonder at T4's single-turn secondary winding - can it have enough impedance to effectively drive Q6's base? The AL winding factor for FT-37-43 cores says that there will be 0.42uH of inductance per t^2. In the case of a single turn, that's simply 0.42uH. At 7 MHz., the impedance is 18.5 ohms.  T4's turns ratio of 8:1 translates Q5's collector load of 420 ohms down to 420/64 or 6.5 ohms. Compare T4's inherent winding impedance of 18.5 ohms with the real load of 6.5 ohms. Since the transformer impedance is a good deal higher than the real load, T4 doesn't drain much RF power. The conclusion is that one turn is OK when driving Q6.