This is a little circuit that I used to power the 1.5v clock from a 12v supply. R1, D1/D2 and C1 form a 1.4v shunt regulator. R1 and D3 are just for reverse polarity protection. R1's value is not critical.
As an aside...I wondered how accurate these clocks were compared to the line operated clocks (NIST traceable over long periods of time). Because I never have to reset my battery operated clocks, I was pretty confident, but to check I set up two electronic counters. One driven off the 60Hz line through a divide-by-60 circuit, and the other though my diode-OR circuit from the gutted clock, for two 1-PPS signals, The gutted clock ran slight off, by a few thousand counts over a couple of days. ??? Though I haven't gone back and tested it, I think gutting the clock circuit, removing the motor coil, and cutting away some of the metal frame may have shifted the crystal oscillator circuit frequency by a slight change in capacitance. I need to do this again, but without any change to the clock.
Ken