Hola EA1FC !
To understand the purpose of this constant current source, first you will see that the rectifier diodes and the capacitors provide a negative supply (more negative than the emitter of power transistor).
In combination with the JFET it behaves like a negative current source that can pull the base current more down as the GND output terminal.
It gives the possibility for Q02, as driver,? and the power transistor on heat sink to conduct at its maximum current (close to saturation)
By this way the voltage drop on the power transistor is minimized. ...and the voltage drop on teh regulation system too !
The voltage regulation is done by the transistor Q03 and the zener as reference.
To get the best voltage stability, the zener is supplied? by the resistor R02. Since the output voltage is constant, the zener current is constant, and zener voltage too !
But to start the power supply the transistor Q03 has to conduct ! This is not possible without the JFET current source because Q03 is referenced at the emitter of Q02.
The presence of capacitor after the current source is not a schematic error. I saw that on the PCB !
I don't know why the engineer did that ? This is odd !
Based on my experience as an electronic engineer, I was faced with solving instability in closed loop systems.
And having such delay by capacitor in a loop circuit can cause instability !
The transistor of the FT-980 PA are 28V models. Supplied at 24 gives a little margin.
An important remark is to have at least a 100-220 ?F on the supply line of a PA to damp current variations in SSB or pulsed CW that could cause instability of the regulated voltage.
I had also a tantalum capacitor in short on the bias of the PA ! I changed all of them by 105°C aluminium.
Having a variac to supply the FT-980 is possible !
I had the idea to remove some turns on the transformer, but there was too much risk of damage to the transformer...
I changed also the trip point of the fan controler.... Now my FT-980 is perfect !
73 F5RCT