Since the 1.610 MHz received signal was noisy, a 2.5 mH RF choke was connected across the output - to no avail. Then, on an impulse, it was connected across the OA90 diode. The result was an enormous increase in signal strength.
Later, tuning up with the converter inadvertently switched off, an equally strong signal was found on 1.830 MHz.
A real efficient tripler had been stumbled upon and the circuit drastically reduced!
So out went the perfboard, leaving only the diode and the choke connected to the sockets.
The available enclosure was too big to start with.
It now appears gigantic.