The TriStar Regenerative Receiver
As a further evolution of the differential regenerative receiver the TriStar design improves a number of issues. Careful consideration and testing points toward using a high impedance AM detector and asymmetrical collector currents for better stability and control.
The TriStar circuit
L1, C2 and C3 are the main frequency determining components in the circuit. The ratio of C1 to C2 is important as it creates an impedance transformation from the high impedance resonant circuit to the relatively low input impedance of Q1. The 10 to 1 ratio is enough to prevent damping of the resonant circuit and provide maximum selectivity. You can alter the values of L1 and C3 to change the frequency range of the receiver. The transistor Q1 works in differential mode versus both Q2 and Q3. When the collector current of Q1 is increasing the collector currents of both Q2 and Q3 are decreasing. When the collector current of Q1 is decreasing the collector currents of both Q2 and Q3 are increasing. Also the emitters of both Q2 and Q3 present a low impedance (of a few ohms) to the emitter of Q1 via both C8 and C9.
This allows Q1 to have much higher gain that if it was working through the emitter resistor R1 alone. In fact the low impedance created is not exactly constant with signal level giving the circuit a compressive open loop gain characteristic. This is vital to smooth control of regeneration. In particular the asymmetry in the collector currents of Q1 and Q3 allows for very smooth control of regeneration and helps stabilize the circuit greatly. L2 is the so called tickler coil which provides positive feedback (regeneration) when wound on the same former as L1. If L2 is connected the wrong way around it will instead cause negative feedback and the radio will not work. The exact number of turns required for L2 is best determined by experimentation. The amount of regeneration is controlled by the voltage at the base of Q2. This can be adjusted by the potentiometers R4 and R5 which are the coarse and fine regeneration controls. L3 is a radio frequency choke. The PNP transistor Q4 is arranged as a square law AM detector. Both the collector and emitter resistors R10 and R11 have an unusually high value of 1 Meg. The reason for this is to reduce any signal leakage back to the highly sensitive regenerative circuit by operating the detector circuit at the lowest possible current. This also makes the detector circuit much more sensitive that if it was operated at higher current levels. If a substantial amount of signal leaks back to the regenerative circuit from the AM detector there is a characteristic motor-boating oscillation whose frequency is controlled by C12. C12 should be a multilayer ceramic or film capacitor that can bypass RF signals effectively. The JFET source follower J1 is used to transform the high impedance output of the detector to a more sensible value for driving an audio amplifier. It could be replaced by a high input impedance op-amp buffer circuit.
The circuit should be powered from a stabilized supply particularly if it shares the same DC power lines as any attached audio amplifier.
The Alternative TriStar receiver
Figure 2 shows the alternative TriStar receiver. R1 has been reduced to 270 Ohms giving Q1 a collector current of about 10mA. At such a high current Q1 has a transconductance of about 400 mA per volt (a rather high value). This allows you to increase the capacitor tap ratio of C1 and C2 from 10:1 to 47:1. The main consequence is that the effects of any base emitter capacitance change are suppressed to the extent of 2000:1 compared to 100:1 in the original circuit. Adjustments of the regeneration controls have little effect on frequency, there is virtually no locking to a received carrier frequency and selectivity is improved. The downside is that the higher tap ratio makes the circuit more linear. This narrows the the transition zone between the non-oscillating and oscillating state making the circuit more sensitive to supply line noise and stray pickup of hum etc.
The further refinements in circuit design and component values continues to provide marked improvements in the performance and stability of the differential regenerative receiver. Hopefully you will find this circuit useful and a good basis for both experimentation and creating practical radio receivers.