Here is a link to a circuit diagram that I traced from a Luna Six.
It is too rough,sorry, but may be legible.
https://app.box.com/s/df96ofgvhh8bfybkt1n3tw474mblfu8y
Looking at bottom of the Unit
Adjustment Potentiometers: Left to right P1, P2, P3, P4, P5
P1 (10k) is battery test current adjust.
P5 ( 20 k) is high illumination coarse adjust. (P5 is used when the nd filter covers the CdS LDR)
P2 ( 75 k) is low illumination coarse adjust. (P2 is used when the CdS LDR is directly exposed via the lens)
P3 ( 10k) is low illumination meter shunt adjust
P4 ( 10k) is high illumination meter shunt adjust
P3 and P4 need to be set near mid point.
The circuit is crude, just a passive resistive divider that approximates
the log function as a reciprocal resistance function.
The meter scale graduations are non linear to more closely approximate to the logarithm.
The meter scale graduations on the one here are divided into 1 ~ 22 in 2 ranges
and the circular slide rule converts those to EV from EV-5 to EV+17.5
In bright sun about EV 14 or so, with the dome on, (incident mode) and ND filter on, the LDR resistance is 520 Ohm.
Dark resistance is about 20 Megohm.
There was no compensation for battery voltage or tempco of the LDR.
The D'Arsonval galvanometer is of 25 uA fsd and internal resistance 2600 Ohm.
The meter needs to be upward and horizontal when taking a reading, as it has a counter weight which swings in a clamp.
The original battery was two by PX13, each with a stable voltage of 1.35 Volt over life and temperature.
The battery current is in range 10 ~ 60 uA, which is not enough to drive a small signal Scottky diode toward saturation,
so there would be accuracy problems by using the often suggested schottky diode dropper in series with modern cells.
If the plastic rear case of the unit is lifted away from the printed circuit board,
the rocker switch pivots come adrift and re-assembly is tricky due to the springs. ( I suppose there was a service jig)