Geiger counters made by the US government for civilian use in the event of a nuclear war with the USSR or China are now readily available for purchase as surplus on sites like eBay and sites that cater to "Preppers."
Because even the last models of these geiger counters are 50 years old, any unit bought as surplus will require some overhaul before it can be trusted to work. There are model specific rebuild kits available from eBay and other places. Since you are here on Instructables, I'm assuming you want to do as much work as you feel competent to do instead of buying a reworked model.
The probe voltage should be checked first. All models of CDV-700's use a nominal 900 dc volt geiger tube. The current supplied is very very low, but if you don't feel comfortable with that voltage, then you should not do maintenance on the unit. The same high voltage exists on the circuits inside as well
In the photo above the probe voltage is measured by inserting a 1 billion ohm (1 Gigohm) 1% resistor into the #1 probe socket and grounding that through a microamp meter to the shell. There is no other way to accurately measure the probe voltage without expensive lab equipment. The meter must be capable of measuring to 10 picoamps (0.01 microamps). The meter shown was bought at Home Depot for $40. Lionels 6b's have a weak high voltage circuit and less than 150 million ohms load will noticeably affect the voltage. When using a 1 Gohm resistor, 0.97 microamps, as shown, means you have 970 volts on the probe. Between 900 and 920 volts is ideal, but 970v on a Lionel is acceptable. A Vic should not be operated with the voltage above 920 because of the way it's audio circuit is made. Overvolting a Vic 6b will destroy a component that cannot be replaced.
If a Vic or Lionel probe voltage is too high, the corotron has failed. The corotron is the glass tube pictured. Any corotron with the Rad trefoil has failed because of age. Those used a radium pit with 22 year half-life. Less than 1/8 of the original radium is left by reason of age no matter how much or little use the unit saw. The solid state replacement shown next to the glass corotron is simply a string of 300v zener diodes turned backwards with one more diode added to make the string come close to 910 volts. The tolerance on most 300v zeners is 5%, so three are strung and the string measured, then one lesser value zener is added to make the proper drop.
The last models of CDV-700 were numbers 6, 6a and 6b. Almost all the units available surplus are Victoreen 6a and 6b, and the Lionel 6b. The ENI model should be avoided because the circuit was marginal even when new. ENI's use 4 D-cells, Vic's use 4, Lionels use 2. Anton Labs became Lionel, so Antons are at least one model older and use 5 D-cells.
Technology has changed so much in 50 years that you have to make certain substitutions when doing maintenance. Alkaline cells didn't exist when the CDV-700 was made, batteries were carbon-zinc chemistry. Carbon-zinc was lighter, had far less capacity and had a paper shell. Modern alkaline and "heavy-duty" manganese chemistry cells are cased in a steel can that is also the anode. The metal battery clips in the CDV should be wrapped in friction or electrical tape so the clip doesn't short out one battery. Because of the way both Vic and Lionel battery holders are made and wired, one battery will short out if the clip is tight enough to touch both batteries at the same time. It's also a good practice to get some AA to D cell plastic case converters and simply use AA alkaline batteries instead of D cells. AA alkaline cells weigh and have the same capacity as the original carbon zinc cells.