Masthead amplifier uses discrete components

I have used with good results the MAR-6 masthead amplifier.
Mostly as RF signal amplification on VHF yagi antenna that I use to track weak satellite signals.

I am located in a inner city, surburban area and swamped by RF signals covering a wide spectrum:
In use, the MAR-6 can sometimes mix these strong RF signals and self-oscillate (AKA 'taking off').
I think that mixing/intermod' occurs in the masthead amp'
possibly caused by the MAR-6 operating into a non-linear part of its characteristic curve.

Self-oscillation can be countered by adjustment of the supply voltage - I use a series connected 1KΩ "pot" in the power supply.
By adjusting the pot, and listening for clean audio from receiver, self-regeneration is prevented.

As an experiment; I have fabricated a discrete component "MAR-6 equivalent" to
Joćo Kolar De Marco, PY2WM's (SK) original design.
This will give the flexibility to vary component values and hopefully lessen the effect of mixing strong RF signals

PY2WM original circuit

Two-stage circuit using bipolar transistors.
First stage emitter follower: < unity gain, buffer, low impedance output
Base-bias uses low ohmage 470Ω resistors,
which should present as a lowish input impedance (50Ω)
Second is a gain stage, beta of BFR93A transistors is 90
I obtained BFR93A from china, very cheap Ft 6GHz, 12V, 30mA
However SOT23 package, SMC (surface mount component) very small!

Re-draft and Vero layout

Test prototype with circuit measurements...

From the prototype and measurement of voltages/currents:
BFR93A within makers specs' - should be linear type operation (class A?)