What are you talking about, clock, 12 volts?
This thing'll run from a single 9-volt battery. All that stuff is internal to the DMM.
I did not realize that your main power source is only 9V.
The power supply that I confirmed will work will only produce about 1V less than the input clock max at no load.
With a 3.2mA load it will barely get to 7V output with no margin as the battery goes lower.
You can explore the original circuit and re-introduce an inverter to drive the bottom 100nF (instead of grounding it). But you'll have to get rid of the glitches it may produce. Possibly by choosing a faster inverter (74HC04 or 74HC14) or more capacitance (vs. 10uF) or a parallel fast 10nF (or so) cap to remove spikes. You'll have to play with it a bit while visualizing on an oscilloscope.
Alternatively, for another single chip solution, you can choose a (CMOS) 556 timer as the clock (first half being the clock with the other half acting as inverter). See
here for the EDN article,
simulation here and snap attached. You
will need a 12V zener to make sure you don't exceed the LCD panel meter power max. In the simulation you can vary the 556 power supply from 7-10V via slider on right margin.
You'll probably be better served with a properly spec'd isolated boost DC-DC module that can tolerate inputs going down to 7V or as high as 9.5V while producing a fixed 12V output.